Atoll 3.3.2 Technical Reference Guide Radio

Version 3.3.
Technical Reference Guide

for Radio Networks
AT332_TRR_E0
AT332_TRR_E0
Atoll 3.3.2 Technical Reference Guide for Radio Networks

Release: AT332_TRR_E0 (October 2016)
Copyright 1997-2016 Forsk. All Rights Reserved.
Published by:
Forsk
7 rue des Briquetiers
31700 Blagnac, France
Tel: +33 562 747 210
Fax: +33 562 747 211
The software described in this document is provided under a licence agreement. The software may only be used or copied under the terms and conditions of the licence agreement.
No part of the contents of this document may be reproduced or transmitted in any form or by any means without written permission from the publisher.
The product or brand names mentioned in this document are trademarks or registered trademarks of their respective registering parties.
Third party services that are not part of Atoll are governed by the terms and conditions of their respective providers, which are subject to change without notice.
The publisher has taken care in the preparation of this document, but makes no expressed or implied warranty of any kind and assumes no responsibility for errors or omissions.
No liability is assumed for incidental or consequential damages in connection with or arising out of the use of the information contained herein.
AT332_TRR_E0

Table of Contents
Table of Contents
Atoll 3.3.2 Technical Reference Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21

1
Antennas and Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
1.1
1.1.1
1.1.2
1.1.3
Antenna Attenuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Calculation of Azimuth and Tilt Angles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Antenna Pattern 3D Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Additional Electrical Downtilt Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.2
Antenna Pattern Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.3
Power Received From Secondary Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.4
1.4.1
1.4.2
Transmitter Radio Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

GSM Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.5
1.5.1
1.5.1.1
1.5.1.2
1.5.1.3
1.5.1.4
1.5.2
1.5.2.1
1.5.2.2
1.5.3
1.5.3.1
1.5.3.2
Repeaters and Remote Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Total Gain Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Repeater Noise Figure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Appendix: Carrier Power and Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
GSM Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
EIRP Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Donor-side Parameter Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Azimuth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Mechanical Downtilt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
1.6
1.6.1
1.6.1.1
1.6.1.2
1.6.1.3
1.6.1.4
1.6.2
1.6.3
1.6.4
Beamforming Smart Antenna Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Downlink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Uplink Beamforming. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Uplink Beamforming and Interference Cancellation (MMSE). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Downlink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Uplink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Uplink Beamforming and Interference Cancellation (MMSE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.7
Grid-of-Beams Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
1.8
Adaptive Beam Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
1.9
Statistical Smart Antenna Gain Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Radio Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57
2.1
2.1.1
2.1.2
2.1.2.1
2.1.2.2
2.1.3
2.1.4
Path Loss Calculation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

Ground Altitude Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Clutter Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Clutter Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Clutter Heights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Geographic Profile Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Resolution of the Extracted Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
2.2
List of Default Propagation Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.3
2.3.1
2.3.2
Okumura-Hata and Cost-Hata Propagation Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Hata Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Corrections to the Hata Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Table of Contents
2016 Forsk. All Rights Reserved.
2.3.3
Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
2.4
2.4.1
2.4.2
2.4.3
ITU 529-3 Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

ITU 529-3 Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
Corrections to the ITU 529-3 Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
2.5
2.5.1
2.5.2
2.5.2.1
2.5.2.2
2.5.2.3
2.5.2.4
2.5.2.5
2.5.2.6
2.5.2.7
2.5.3
2.5.3.1
2.5.3.2
2.5.4
Standard Propagation Model (SPM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65

SPM Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65
Visibility and Distance Between Transmitter and Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
Effective Transmitter Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
Effective Receiver Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69
Correction for Hilly Regions in Case of LOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70
Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70
Losses due to Clutter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70
Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71
Automatic Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
General Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
Sample Values for SPM Path Loss Formula Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73
Unmasked Path Loss Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74
2.6
2.6.1
2.6.2
WLL Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75

WLL Path Loss Formula. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75
2.7
2.7.1
2.7.2
ITU-R P.526-5 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76

2.8
2.8.1
2.8.2
ITU-R P.370-7 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76

2.9
2.9.1
2.9.2
2.9.3
Erceg-Greenstein (SUI) Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78

SUI Terrain Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78
Erceg-Greenstein (SUI) Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79
2.10
2.10.1
2.10.1.1
2.10.1.2
2.10.1.3
2.10.1.4
2.10.1.5
2.10.1.6
ITU-R P.1546-2 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80

Step 1: Determination of Graphs to be Used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80
Step 2: Calculation of Maximum Field Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
Step 3: Determination of Transmitter Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
Step 4: Interpolation/Extrapolation of Field Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
Step 5: Calculation of Correction Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83
Step 6: Calculation of Path Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84
2.11
Sakagami Extended Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84
2.12
Free Space Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86
2.13
2.13.1
2.13.2
2.13.3
2.13.4
2.13.5
Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86
Knife-edge Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86
3 Knife-edge Deygout Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87
Epstein-Peterson Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
Deygout Method with Correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
Millington Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
2.14
2.14.1
2.14.1.1
2.14.1.2
2.14.2
2.14.2.1
2.14.2.1.1
2.14.2.1.2
2.14.2.2
2.14.2.2.1
2.14.2.2.2
Shadow Fading Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90

Shadowing Margin Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95
Shadowing Margin Calculation in Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96
Shadowing Margin Calculation in Monte-Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97
Macro-Diversity Gains Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
Uplink Macro-Diversity Gain Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
Shadowing Error PDF (n Signals). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
Uplink Macro-Diversity Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Downlink Macro-Diversity Gain Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Shadowing Error PDF (n Signals). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Downlink Macro-Diversity Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
AT332_TRR_E0

Table of Contents
2.15
2.15.1
2.15.2
2.15.3
2.15.3.1
2.15.3.2
Path Loss Matrices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

Calculation Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Validity of Path Loss Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Path Loss Tuning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Transmitter Path Loss Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Repeater Path Loss Tuning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
2.16
2.16.1
2.16.1.1
2.16.1.2
2.16.1.3
2.16.2
2.16.2.1
2.16.2.2
2.16.2.3
2.16.3
2.16.3.1
2.16.3.2
2.16.3.3
2.16.3.4
2.16.3.5
2.16.3.6
2.16.4
2.16.4.1
2.16.4.2
2.16.5
2.16.5.1
2.16.5.2
2.16.5.3
2.16.5.4
File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

Path Loss Matrix File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Pathloss.dbf File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Pathloss.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
LOS File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Path Loss Tuning File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Pathloss.dbf File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Pathloss.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
PTS File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Interference Matrix File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
CLC Format (One Value per Line). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
CLC File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
DCT File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
IM0 Format (One Histogram per Line) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
IM1 Format (One Value per Line, TX Name Repeated) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
IM2 Format (Co- and Adjacent-channel Probabilities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
"Per Transmitter" Prediction File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
<per_transmitter_prediction>.dbf File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
<per_transmitter_prediction>.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Coverage Prediction Export and Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Filtering Coverage Predictions at Export. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Smoothing Coverage Predictions at Export . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Examples of Prediction Export Filtering and Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Coverage Prediction Reports Over Focus/Computation Zones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
GSM GPRS EDGE Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125
3.1
3.1.1
3.1.2
3.1.3
3.1.3.1
3.1.3.2
3.1.4
3.1.4.1
3.1.4.1.1
3.1.4.1.2
3.1.4.1.3
3.1.4.1.4
3.1.4.1.5
3.1.4.1.6
3.1.4.1.7
3.1.4.1.8
3.1.4.2
3.1.4.2.1
3.1.4.2.2
Signal Level Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

DL Signal Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
UL Signal Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Signal Level-based DL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
DL Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Best Signal Level and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Second Best Signal Level and an Overlap Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Best Signal Level per HCS Layer and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Second Best Signal Level per HCS Layer and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
HCS Servers and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Highest Priority HCS Server and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Best Idle Mode Reselection Criterion (C2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Coverage Display. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.2
3.2.1
3.2.2
3.2.3
3.2.3.1
3.2.3.2
3.2.3.2.1
3.2.3.2.2
3.2.3.3
3.2.3.3.1
3.2.3.3.2
Interference-based DL Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

DL Carrier-to-Interference Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Interference-based DL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Service Area Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Coverage Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Interference Condition Satisfied by At Least One TRX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Interference Condition Satisfied by The Worst TRX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Coverage Display. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.3
3.3.1
GPRS/EDGE Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

Coding Scheme Selection and Throughput Calculation Without Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

Table of Contents
3.3.1.1
3.3.1.2
3.3.1.3
3.3.2
3.3.2.1
3.3.2.2
3.3.2.3
3.3.3
3.3.4
3.3.5
3.3.5.1
3.3.5.1.1
3.3.5.1.2
3.3.5.1.3
3.3.5.1.4
3.3.5.1.5
3.3.5.1.6
3.3.5.1.7
3.3.5.1.8
3.3.5.2
3.3.5.2.1
3.3.5.2.2
Calculations Based on C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculations Based on C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculations Based on C/(I+N) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coding Scheme Selection and Throughput Calculation With Ideal Link Adaptation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculations Based on C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculations Based on C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Application Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
BLER Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
GPRS/EDGE Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Signal Level and an Overlap Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Second Best Signal Level and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Signal Level per HCS Layer and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Second Best Signal Level per HCS Layer and an Overlap Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HCS Servers and an Overlap Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Highest Priority HCS Server and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Idle Mode Reselection Criterion (C2). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
137
137
137
138
138
138
139
139
140
140
140
140
140
141
141
141
141
142
142
142
142
143
3.4
3.4.1
3.4.2
3.4.2.1
3.4.2.2
3.4.3
3.4.3.1
3.4.3.2
3.4.4
3.4.4.1
3.4.4.1.1
3.4.4.1.2
3.4.4.1.3
3.4.4.1.4
3.4.4.1.5
3.4.4.1.6
3.4.4.1.7
3.4.4.2
3.4.4.2.1
3.4.4.2.2
Codec Mode Selection and CQI Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Circuit Quality Indicator Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CQI Calculation Without Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculations Based on C/N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CQI Calculation With Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculations Based on C/N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Circuit Quality Indicators Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Signal Level per HCS Layer and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Second Best Signal Level per HCS Layer and an Overlap Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HCS Servers and an Overlap Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Highest Priority HCS Server and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
145
147
147
147
148
148
148
148
149
149
149
149
150
150
150
150
151
151
151
151
3.5
3.5.1
3.5.1.1
3.5.1.2
3.5.1.3
3.5.1.4
3.5.1.5
3.5.1.6
3.5.1.7
3.5.1.8
3.5.2
3.5.2.1
3.5.2.2
3.5.2.2.1
3.5.2.2.2
3.5.2.2.3
3.5.2.2.4
3.5.3
3.5.3.1
3.5.3.2
3.5.3.3
3.5.3.4
3.5.3.4.1
UL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DL Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
All Servers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Signal Level and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Second Best Signal Level and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Signal Level per HCS Layer and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Second Best Signal Level per HCS Layer and an Overlap Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HCS Servers and an Overlap Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Highest Priority HCS Server and an Overlap Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Idle Mode Reselection Criterion (C2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage by UL Signal Level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UL Signal Level (in dBm, dBV, dBV/m) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best UL Signal Level (in dBm, dBV, dBV/m). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UL Total Losses (dB). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Minimum UL Total Losses (dB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage by UL C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UL C/I Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
152
152
152
152
153
153
153
154
154
155
155
155
155
155
155
156
156
156
156
156
156
156
156
AT332_TRR_E0

Table of Contents
3.5.3.4.2
3.5.3.4.3
3.5.4
3.5.4.1
3.5.4.2
3.5.4.3
3.5.5
3.5.5.1
3.5.5.2
Max C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

Min C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Coverage by UL Coding Schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Coding Scheme Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Coverage by UL Codec Modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Codec Mode Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.6
3.6.1
3.6.1.1
3.6.1.1.1
3.6.1.1.2
3.6.1.2
3.6.1.2.1
3.6.1.2.2
3.6.1.3
3.6.1.3.1
3.6.1.3.2
3.6.2
3.6.2.1
3.6.2.1.1
3.6.2.1.2
3.6.2.1.3
3.6.2.2
3.6.2.2.1
3.6.2.2.2
3.6.2.2.3
Traffic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Traffic Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Circuit Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Packet Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Concentric Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Calculation of the Traffic Demand per Subcell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Concentric Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Concentric Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
3.7
3.7.1
3.7.1.1
3.7.1.2
3.7.1.2.1
3.7.1.2.2
3.7.1.2.3
3.7.2
3.7.2.1
3.7.2.1.1
3.7.2.1.2
3.7.2.2
3.7.2.2.1
3.7.2.2.2
3.7.2.2.3
3.7.2.2.4
3.7.2.2.5
3.7.2.2.6
Network Dimensioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Dimensioning Models and Quality Graphs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Circuit Switched Traffic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Blocking Probability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Network Dimensioning Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Network Dimensioning Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Network Dimensioning Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Step 1: Timeslots Required for CS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Step 2: TRXs Required for CS Traffic and Dedicated PS Timeslots . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Step 3: Effective CS Blocking, Effective CS Traffic Overflow and Served CS Traffic . . . . . . . . . . . . . . . 178
Step 4: TRXs to Add for PS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Step 5: Served PS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Step 6: Total Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
3.8
3.8.1
3.8.1.1
3.8.1.2
3.8.1.3
3.8.2
3.8.2.1
3.8.2.1.1
3.8.2.1.2
3.8.2.1.3
3.8.2.1.4
3.8.2.1.5
3.8.2.1.6
3.8.2.2
3.8.2.2.1
3.8.2.2.2
3.8.2.2.3
3.8.2.2.4
Key Performance Indicators Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

Circuit Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Erlang B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Erlang C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Served Circuit Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Case 1: Total Traffic Demand > Dedicated + Shared Timeslots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Packet Switched Traffic Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Throughput Reduction Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Blocking Probability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Served Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Case 2: Total Traffic Demand < Dedicated + Shared Timeslots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Packet Switched Traffic Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Throughput Reduction Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Table of Contents
3.8.2.2.5
3.8.2.2.6
Blocking Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Served Packet Switched Traffic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.9
3.9.1
3.9.1.1
3.9.1.2
3.9.1.3
3.9.1.4
3.9.1.5
3.9.1.6
3.9.1.7
3.9.1.8
3.9.1.9
Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Radio Resource Management in GSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
GSM Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Servers Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Codec Mode Assignment and DL Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coding Scheme Assignment, Throughput Evaluation and DL Power Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Subcell Traffic Loads Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Half-Rate Traffic Ratio Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DL Power Control Gain Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DTX DL Gain Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
GSM Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
184
184
184
187
187
188
189
189
189
190
190
3.10
3.10.1
3.10.2
3.10.3
3.10.4
Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Neighbour Allocation for a Group of Transmitters or One Transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Neighbour Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix: Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
191
191
193
194
195
3.11
3.11.1
3.11.1.1
3.11.1.2
3.11.1.2.1
3.11.1.2.2
3.11.1.2.3
3.11.2
3.11.2.1
3.11.2.2
3.11.2.3
3.11.3
3.11.3.1
3.11.3.2
3.11.3.3
3.11.3.4
3.11.3.4.1
3.11.3.4.2
3.11.3.4.3
3.11.3.5
AFP Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The AFP Cost Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cost Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cost Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Separation Violation Cost Component. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Cost Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
I_DIV, F_DIV and Other Advanced Cost Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The AFP Blocked Traffic Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of New Traffic Loads Including Blocked Traffic Loads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Recalculation of CS and PS From Traffic Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Testing the Blocked Cost Using Traffic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Using Interferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cumulative Density Function of C/I Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Precise Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Precise Interference Distribution Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Direct Availability of Precise Interference Distribution to the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . .
Efficient Calculation and Storage of Interference Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Robustness of the IM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Load and Interference Information Discrimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
195
195
196
198
198
199
201
202
202
203
204
204
204
204
205
205
205
206
206
206
UMTS HSPA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
4.1
4.1.1
4.1.2
4.1.2.1
4.1.2.2
4.1.3
4.1.3.1
4.1.3.1.1
4.1.3.1.2
4.1.3.1.3
4.1.3.2
4.1.3.2.1
4.1.3.2.2
General Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plot Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
209
209
209
209
210
210
210
210
210
211
211
211
211
4.2
4.2.1
4.2.2
4.2.3
4.2.4
4.2.5
Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ec/I0 Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DL Eb/Nt Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UL Eb/Nt Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
212
212
213
220
221
222
4.3
4.3.1
4.3.1.1
Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Generating a Realistic User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Simulations Based on User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
AT332_TRR_E0

Table of Contents
4.3.1.1.1
4.3.1.1.2
4.3.1.2
4.3.1.2.1
4.3.1.2.2
4.3.1.2.3
4.3.2
4.3.2.1
4.3.2.2
4.3.2.3
4.3.2.3.1
4.3.2.3.2
4.3.2.3.3
4.3.2.3.4
4.3.2.3.5
4.3.2.3.6
4.3.2.3.7
4.3.2.4
4.3.2.4.1
4.3.2.4.2
4.3.2.4.3
4.3.2.4.4
4.3.2.5
4.3.3
4.3.3.1
4.3.3.2
4.3.3.2.1
4.3.3.2.2
4.3.3.2.3
4.3.3.2.4
4.3.4
4.3.4.1
4.3.4.2
4.3.4.2.1
4.3.4.2.2
4.3.4.2.3
4.3.4.3
4.3.4.3.1
4.3.4.3.2
4.3.4.4
4.3.4.5
4.3.4.6
Circuit Switched Service (i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224

Packet Switched Service (j). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Throughputs in Uplink and Downlink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Total Number of Users (All Activity Statuses) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Number of Users per Activity Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Power Control Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
R99 Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
HSDPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
HSDPA Power Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users . . . . . . . . . . . . . . . . . 236
HSDPA Bearer Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Fast Link Adaptation Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
MIMO Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Scheduling Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Dual-Cell HSDPA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
HSUPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Admission Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
HSUPA Bearer Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Noise Rise Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Radio Resource Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Convergence Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
R99 Related Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
HSPA Related Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Statistics Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Mobiles Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Cells Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Sites Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Admission Control in the R99 Part . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Resources Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
OVSF Codes Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Channel Elements Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Iub Backhaul Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Downlink Load Factor Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Downlink Load Factor per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Downlink Load Factor per Mobile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Uplink Load Factor Due to One User . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Inter-carrier Power Sharing Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Best Serving Cell Determination in Monte Carlo Simulations - Old Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
4.4
4.4.1
4.4.2
4.4.2.1
4.4.2.2
4.4.2.3
4.4.2.4
4.4.2.5
4.4.3
4.4.3.1
4.4.3.1.1
4.4.3.1.2
4.4.3.2
4.4.3.2.1
4.4.3.2.2
4.4.3.3
4.4.3.3.1
4.4.3.3.2
4.4.3.4
4.4.3.4.1
4.4.3.4.2
4.4.3.5
4.4.3.5.1
4.4.3.5.2
UMTS HSPA Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

Best Serving Cell and Active Set Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
Point Analysis - AS Analysis Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
Bar Graph and Pilot Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
Downlink R99 Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Uplink R99 Sub-Menu. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
HSDPA Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
HSUPA Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
Pilot Quality Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
Prediction Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
Study Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
Downlink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
Uplink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Downlink Total Noise Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
HSDPA Prediction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302

Table of Contents
4.4.3.6
4.4.3.6.1
4.4.3.6.2
4.4.3.6.3
HSUPA Prediction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Prediction Study Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
307
307
308
308
4.5
4.5.1
4.5.2
4.5.3
4.5.3.1
4.5.3.2
4.5.4
4.5.4.1
Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Neighbour Allocation for a Group of Transmitters or One Transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Importance of Intra-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Importance of Inter-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
310
310
314
314
314
316
317
317
4.6
4.6.1
4.6.1.1
4.6.1.2
4.6.1.2.1
4.6.1.2.2
4.6.1.3
4.6.1.3.1
4.6.1.3.2
4.6.1.3.3
4.6.2
4.6.2.1
4.6.2.1.1
4.6.2.1.2
4.6.2.1.3
4.6.2.1.4
4.6.2.2
Primary Scrambling Code Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Options and Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Multi-Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Transmitter Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Site Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Allocation Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Allocation Strategies and Use a Maximum of Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Strategy: Clustered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Strategy: Distributed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Strategy: One Cluster per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Strategy: Distributed per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Allocate Carriers Identically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
317
318
318
320
320
321
321
321
324
324
324
324
325
326
326
326
327
4.7
4.7.1
4.7.2
4.7.2.1
4.7.2.2
4.7.2.3
4.7.2.3.1
Automatic GSM-UMTS Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Algorithm Based on Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
328
328
328
328
329
331
331
CDMA2000 Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
5.1
5.1.1
5.1.2
5.1.2.1
5.1.2.2
5.1.3
5.1.3.1
5.1.3.1.1
5.1.3.1.2
5.1.3.1.3
5.1.3.2
5.1.3.2.1
5.1.3.2.2
General Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plot Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
335
335
336
336
336
336
336
336
336
337
337
337
337
5.2
5.2.1
5.2.1.1
5.2.1.2
5.2.1.3
5.2.1.4
5.2.1.5
5.2.2
5.2.2.1
5.2.2.2
5.2.2.3
5.2.2.4
Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Parameters Used for CDMA2000 1xRTT Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ec/I0 Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Parameters Used for CDMA2000 1xEV-DO Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ec/I0 and Ec/Nt Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
338
338
338
343
344
345
346
348
348
352
353
354
10
AT332_TRR_E0

Table of Contents
5.3
Active Set Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
5.4
5.4.1
5.4.1.1
5.4.1.1.1
5.4.1.1.2
5.4.1.2
5.4.1.3
5.4.2
5.4.2.1
5.4.2.1.1
5.4.2.1.2
5.4.2.1.3
5.4.2.2
5.4.2.2.1
5.4.2.2.2
5.4.2.2.3
5.4.3
5.4.3.1
5.4.3.2
5.4.3.2.1
5.4.3.2.2
5.4.3.3
5.4.3.3.1
5.4.3.3.2
5.4.3.4
5.4.3.5
Simulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
Generating a Realistic User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
Number of Users, User Activity Status and User Throughput. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
Simulations Based on User Profile Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
Transition Flags for 1xEV-DO Rev.0 User Throughputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
User Geographical Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
Network Regulation Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
CDMA2000 1xRTT Power Control Simulation Algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
Presentation of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
Convergence Criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
CDMA2000 1xEV-DO Power/Data Rate Control Simulation Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Presentation of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Convergence Criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Admission Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Resources Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Walsh Code Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Channel Element Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Downlink Load Factor Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
Downlink Load Factor per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
Downlink Load Factor per Mobile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Best Server Determination in Monte Carlo Simulations - Old Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Radio Bearer Allocation Algorithm for Multi-carrier EVDO Rev.B - Old Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
5.5
5.5.1
5.5.1.1
5.5.1.2
5.5.1.2.1
5.5.1.2.2
5.5.1.3
5.5.1.3.1
5.5.1.3.2
5.5.2
5.5.2.1
5.5.2.2
5.5.2.2.1
5.5.2.2.2
5.5.2.3
5.5.2.3.1
5.5.2.3.2
5.5.2.4
5.5.2.4.1
5.5.2.4.2
CDMA2000 Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390

Point Analysis: The AS Analysis Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
Bar Graph and Pilot Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
Downlink Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
Uplink Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
Pilot Quality Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
Downlink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Uplink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Downlink Total Noise Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
Analysis on the Best Carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
Analysis on a Specific Carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
5.6
5.6.1
5.6.2
5.6.3
5.6.3.1
5.6.3.2
5.6.4
5.6.4.1
Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421

Neighbour Allocation for all Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
Neighbour Allocation for a Group of Transmitters or One Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Importance of Intra-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Importance of Inter-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
5.7
5.7.1
5.7.1.1
5.7.1.2
5.7.1.2.1
5.7.1.2.2
5.7.1.2.3
5.7.1.3
5.7.1.3.1
PN Offset Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428

Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
Options and Constraints. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
Allocation Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
Multi-Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Difference between Adjacent and Distributed PN-Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
11

Table of Contents
5.7.1.3.2
5.7.1.3.3
5.7.2
5.7.2.1
5.7.2.2
5.7.2.3
5.8
5.8.1
5.8.2
5.8.2.1
5.8.2.2
5.8.2.3
Transmitter Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Site Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Allocation Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Strategy: PN Offset per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Strategy: Adjacent PN-Clusters Per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Strategy: Distributed PN-Clusters Per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
433
434
434
434
435
435
Automatic GSM-CDMA Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Algorithm Based on Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
436
436
436
437
437
440
LTE Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
6.1
Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
6.2
6.2.1
6.2.2
6.2.3
6.2.4
6.2.5
6.2.6
6.2.7
6.2.8
6.2.9
6.2.10
6.2.11
6.2.12
6.2.13
6.2.14
6.2.15
6.2.16
6.2.17
6.2.18
6.2.19
6.2.20
Calculation Quick Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Downlink Transmission Powers Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Co- and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Rise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Downlink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Uplink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Downlink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Uplink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation . . . . . . .
Scheduling and Radio Resource Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
449
449
451
452
455
455
459
460
461
462
462
463
463
463
464
465
465
465
466
467
469
6.3
6.3.1
6.3.1.1
6.3.1.2
6.3.1.3
6.3.1.4
6.3.2
6.3.2.1
6.3.2.2
6.3.2.3
6.3.2.4
6.3.3
6.3.4
6.3.4.1
6.3.4.1.1
6.3.4.1.2
6.3.4.2
Available Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Profile View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reception View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Details View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Downlink Signal Level Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Effective Signal Analysis Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/(I+N)-based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cell Identifier Collision Zones Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculations on Subscriber Lists. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simulations Based on User Profile Traffic Maps and Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . .
Simulations Based on Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
470
470
470
470
470
470
471
471
472
473
475
476
476
476
477
478
479
6.4
6.4.1
6.4.2
6.4.2.1
6.4.2.2
6.4.2.3
6.4.2.4
6.4.3
6.4.3.1
6.4.3.2
Calculation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Downlink Transmission Power Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conversion From Channel Numbers to Start and End Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Co-Channel Overlap Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adjacent Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Total Overlap Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Subframe Pattern Collision Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Subframe Pattern Normalisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Determination of Effective Subframe Patterns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
485
485
493
494
495
496
496
497
497
498
12
AT332_TRR_E0
6.4.3.3
6.4.4
6.4.4.1
6.4.4.2
6.4.4.3
6.4.4.4
6.4.4.5
6.4.4.6
6.4.4.7
6.4.4.8
6.4.4.8.1
6.4.4.8.2
6.4.4.9
6.4.4.10
6.4.5
6.4.6
6.4.6.1
6.4.6.1.1
6.4.6.1.2
6.4.6.2
6.4.6.2.1
6.4.6.2.2
6.4.6.3
6.4.7
6.4.7.1
6.4.7.2
6.5
6.5.1
6.5.2
6.5.3
6.5.3.1
6.5.3.2
6.5.3.3
6.5.4
6.5.4.1
6.5.4.2
6.5.4.3
6.5.5
6.5.5.1
6.5.5.2
6.5.5.3
6.5.6
6.5.6.1
6.5.6.2

Table of Contents
Calculation of Subframe Collision Probabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499

Signal Level and Signal Quality Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
Noise Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
C/(I+N) and Bearer Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
Interference Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
Interfering Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
Noise Rise Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
C/(I+N) and Bearer Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
Best Server Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
Calculation of Downlink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
Calculation of Uplink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
Calculation UE Capacities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
Calculation of Downlink UE Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
Calculation of Uplink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation . . . . . . 551
Scheduling and Radio Resource Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
Scheduling and Radio Resource Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
Automatic Planning Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
Automatic Neighbour Planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
Automatic Inter-technology Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
Automatic Frequency Planning Using the AFP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573
Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573
Cost Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574
AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
Automatic Physical Cell ID Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
Automatic PRACH RSI Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582
Interference Matrix Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582
Distance Importance Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582
3GPP Multi-RAT Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .587
7.1
Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
7.2
7.2.1
7.2.2
Multi-RAT Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587

User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
7.3
Multi-RAT Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
3GPP2 Multi-RAT Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .593
8.1
Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
8.2
8.2.1
8.2.2
Multi-RAT Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593

User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
8.3
Multi-RAT Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
TD-SCDMA Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .599

13

Table of Contents
9.1
9.1.1
9.1.2
9.1.3
9.1.4
9.1.5
9.1.6
9.1.7
Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
P-CCPCH Eb/Nt and C/I Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DwPCH C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DL TCH Eb/Nt and C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UL TCH Eb/Nt and C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HSDPA Dynamic Power Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
599
599
604
604
605
605
606
606
9.2
9.2.1
9.2.1.1
9.2.1.2
9.2.2
9.2.2.1
9.2.2.2
9.2.2.2.1
9.2.2.2.2
9.2.2.3
9.2.2.4
9.2.2.5
9.2.2.5.1
9.2.2.5.2
9.2.2.6
9.2.2.6.1
9.2.2.6.2
9.2.2.7
9.2.2.7.1
9.2.2.7.2
9.2.2.8
Signal Level Based Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
RSCP Based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
P-CCPCH RSCP Coverage Prediction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Server P-CCPCH Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
P-CCPCH Pollution Analysis Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DwPCH RSCP Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UpPCH RSCP Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Baton Handover Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Scrambling Code Interference Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
606
606
607
607
607
607
608
608
608
609
609
609
609
610
610
610
610
611
611
611
611
9.3
9.3.1
9.3.1.1
9.3.1.1.1
9.3.1.1.2
9.3.1.2
9.3.1.2.1
9.3.1.2.2
9.3.1.2.3
9.3.2
9.3.2.1
9.3.2.2
9.3.2.2.1
9.3.2.2.2
9.3.2.2.3
9.3.2.2.4
9.3.2.2.5
9.3.2.2.6
9.3.2.2.7
9.3.2.3
9.3.2.3.1
9.3.2.3.2
9.3.2.3.3
9.3.2.3.4
9.3.2.3.5
9.3.2.4
Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Generating a Realistic User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simulations Based on User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Circuit Switched Service (i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Packet Switched Service (j) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Throughputs in Uplink and Downlink. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Total Number of Users (All Activity Statuses) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Number of Users per Activity Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Power Control Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Algorithm Initialisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
R99 Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Determination of Mis Best Server (SBS(Mi)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dynamic Channel Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Uplink Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Downlink Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Uplink Signals Update . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Downlink Signals Update. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Control of Radio Resource Limits (Downlink Traffic Power and Uplink Load) . . . . . . . . . . . . . . . . . .
HSDPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HSDPA Power Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Connection Status and Number of HSDPA Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HSDPA Admission Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HSDPA Dynamic Channel Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ressource Unit Saturation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Convergence Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
612
612
612
613
613
616
616
617
617
617
618
618
618
619
621
623
625
625
625
626
626
628
628
629
629
629
9.4
9.4.1
9.4.2
9.4.3
9.4.4
9.4.5
9.4.6
9.4.7
9.4.8
9.4.9
TD-SCDMA Prediction Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

P-CCPCH Reception Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DwPCH Reception Analysis (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Downlink TCH RSCP Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Uplink TCH RSCP Coverage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Downlink Total Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Downlink Service Area Analysis (Eb/Nt) or (C/I). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Uplink Service Area Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Effective Service Area Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cell to Cell Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
630
630
632
633
634
635
635
637
639
640
14
AT332_TRR_E0
9.4.10
9.4.11

Table of Contents
UpPCH Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641

HSDPA Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
9.5
9.5.1
9.5.1.1
9.5.1.2
9.5.1.3
9.5.1.4
9.5.1.5
9.5.2
9.5.3
9.5.4
9.5.4.1
9.5.4.1.1
9.5.4.1.2
9.5.4.2
9.5.4.2.1
9.5.4.2.2
9.5.4.2.3
Smart Antenna Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642

Modelling in Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
Grid of Beams Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
Adaptive Beam Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
Statistical Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645
Beamforming Smart Antenna Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645
3rd Party Smart Antenna Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
Construction of the Geographic Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
Modelling in Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
HSDPA Quality and Throughput Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
Fast Link Adaptation Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
CQI Based on P-CCPCH Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
CQI Based on HS-PDSCH Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652
Coverage Prediction Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
Colour per CQI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
Colour per Peak Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654
Colour per HS-PDSCH Ec/Nt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654
9.6
9.6.1
N-Frequency Mode and Carrier Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654

Automatic Carrier Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654
9.7
9.7.1
9.7.2
9.7.3
9.7.4
Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655

Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
Neighbour Allocation for a Group of Transmitters or One Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
Appendix: Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
9.8
9.8.1
9.8.1.1
9.8.1.2
9.8.1.3
9.8.1.3.1
9.8.1.3.2
9.8.1.4
9.8.1.4.1
9.8.1.4.2
9.8.1.4.3
9.8.2
9.8.2.1
9.8.2.1.1
9.8.2.1.2
9.8.2.1.3
9.8.2.1.4
9.8.2.2
Scrambling Code Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660

Allocation Constraints and Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
Allocation Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
Allocation Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
Multi-Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
Transmitter Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
Site Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
Scrambling Code Allocation Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
Strategy: Clustered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
Strategy: Distributed per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
Strategy: One SYNC_DL Code per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
Strategy: Distributed per Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
Multi Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
9.9
9.9.1
9.9.1.1
9.9.1.2
9.9.1.3
9.9.1.3.1
Automatic GSM/TD-SCDMA Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670

Algorithm Based on Distance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
10
WiMAX BWA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .677
10.1
Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
10.2
10.2.1
10.2.2
10.2.3
10.2.4
10.2.5
10.2.6
10.2.7
Calculation Quick Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682

Co- and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
Preamble Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
Preamble Noise Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
Preamble Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
Preamble C/N Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
Preamble C/(I+N) Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Traffic and Pilot Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
15

Table of Contents
10.2.8
10.2.9
10.2.10
10.2.11
10.2.12
10.2.13
10.2.14
10.2.15
10.2.16
10.2.17
10.2.18
10.2.19
10.2.20
Traffic and Pilot Noise Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Noise Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation . . . . . . .
User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
684
685
686
686
687
687
687
688
688
688
689
691
693
10.3
10.3.1
10.3.1.1
10.3.1.2
10.3.1.3
10.3.1.4
10.3.2
10.3.2.1
10.3.2.2
10.3.2.3
10.3.2.4
10.3.3
10.3.4
10.3.4.1
10.3.4.1.1
10.3.4.1.2
10.3.4.2
Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Profile View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reception View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Details View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Preamble Signal Level Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cell Identifier Collision Zones Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simulations Based on User Profile Traffic Maps and Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . .
Simulations Based on Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
694
694
694
694
694
694
695
695
696
697
699
699
700
700
700
702
703
10.4
10.4.1
10.4.1.1
10.4.1.2
10.4.1.3
10.4.1.4
10.4.1.5
10.4.2
10.4.2.1
10.4.2.2
10.4.2.3
10.4.2.4
10.4.2.5
10.4.3
10.4.4
10.4.5
10.4.6
10.4.6.1
10.4.6.2
10.4.6.3
10.4.6.3.1
10.4.6.3.2
10.4.6.4
10.4.6.5
10.4.6.6
10.4.6.7
10.4.6.8
10.4.6.8.1
10.4.6.8.2
10.4.6.9
10.4.6.10
10.4.7
10.4.7.1
10.4.7.1.1
10.4.7.1.2
Conversion From Channel Numbers to Start and End Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Co-Channel Overlap Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adjacent Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDD TDD Overlap Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Preamble Signal Level and Quality Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Preamble Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Preamble Noise Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Preamble Interference Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Preamble C/N Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Preamble C/(I+N) Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Server Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Service Area Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Permutation Zone Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot Signal Level and Quality Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot Interference Signal Levels Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Effective Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic and Pilot C/(I+N) and Bearer Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Interference Signal Levels Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Rise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic C/(I+N) and Bearer Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Sampling Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Symbol Duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
706
706
707
708
709
709
710
711
711
712
714
715
716
717
717
718
719
719
721
722
722
726
729
731
733
734
735
735
737
738
740
744
744
744
744
16
AT332_TRR_E0
10.4.7.1.3
10.4.7.1.4
10.4.7.2
10.4.8
10.4.8.1
10.4.8.2
10.5
10.5.1
10.5.2
10.5.3
10.5.3.1
10.5.3.2
10.5.3.3
10.5.4
10.5.4.1
10.5.4.2
10.5.4.3
10.5.5
10.5.5.1
10.5.5.2
10.5.5.3
10.5.6
10.5.6.1
10.5.6.2
11

Table of Contents
Calculation of Total Cell Resources - TDD Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745

Calculation of Total Cell Resources - FDD Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746
Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-User Throughput Calculation . . . . . . 747
Scheduling and Radio Resource Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761
Automatic Frequency Planning Using the AFP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
Automatic Preamble Index Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
Automatic Zone PermBase Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
Interference Matrix Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
Distance Importance Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 778
Wi-Fi Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .783
11.1
Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
11.2
11.2.1
11.2.2
11.2.3
11.2.4
11.2.5
11.2.6
11.2.7
11.2.8
11.2.9
11.2.10
11.2.11
11.2.12
11.2.13
11.2.14
11.2.15
Calculation Quick Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786

Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
C/N Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
C/N Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
Calculation of Total Cell Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
Channel Throughput, Cell Capacity, and Per-user Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790
User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792
11.3
11.3.1
11.3.1.1
11.3.1.2
11.3.1.3
11.3.2
11.3.2.1
11.3.2.2
11.3.2.3
11.3.3
11.3.4
11.3.4.1
11.3.4.1.1
11.3.4.1.2
11.3.4.2
Available Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793

Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
Profile View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
Reception View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
Interference View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
Signal Level Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
Effective Signal Analysis Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
C/(I+N)-based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
Calculations on Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798
Simulations Based on User Profile Traffic Maps and Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . . 798
Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801
11.4
11.4.1
11.4.1.1
11.4.1.2
11.4.1.3
Calculation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 803

Conversion From Channel Numbers to Start and End Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804
Co-Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 805
Adjacent Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
17

Table of Contents
11.4.1.4
11.4.2
11.4.2.1
11.4.2.2
11.4.2.3
11.4.2.4
11.4.2.5
11.4.2.6
11.4.2.7
11.4.2.8
11.4.2.8.1
11.4.2.8.2
11.4.2.9
11.4.2.10
11.4.3
11.4.4
11.4.5
11.4.5.1
11.4.5.2
11.4.6
11.4.6.1
11.4.6.2
11.5
11.5.1
11.5.2
11.5.3
11.5.3.1
11.5.3.2
11.5.3.3
11.5.4
11.5.4.1
11.5.4.2
12

Signal Level and Quality Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/(I+N) and Bearer Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Signal Levels Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Rise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/(I+N) and Bearer Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Best Server Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Service Area Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Channel Throughput, Cell Capacity, and Per-user Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Scheduling and Radio Resource Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
806
807
807
808
809
810
812
813
814
815
815
815
816
818
819
820
820
820
821
824
825
829
Automatic Planning Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Automatic Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Automatic Inter-technology Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Automatic Frequency Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cost Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Matrix Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Distance Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
830
830
834
837
837
838
839
839
839
839
LPWA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
12.1
Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
12.2
12.2.1
12.2.2
12.2.3
12.2.4
12.2.5
12.2.6
12.2.7
12.2.8
12.2.9
12.2.10
Calculation Quick Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Channel Throughput, Cell Capacity, and Per-user Throughput Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
846
846
846
846
846
847
847
847
847
847
848
12.3
12.3.1
12.3.1.1
12.3.1.2
12.3.1.3
12.3.2
12.3.2.1
12.3.2.2
12.3.2.3
12.3.3
Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Profile View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reception View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Signal Level Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
849
849
849
849
849
850
850
851
852
853
12.4
12.4.1
12.4.1.1
12.4.1.2
12.4.1.3
12.4.1.4
12.4.1.5
Signal Level and Quality Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C/(I+N) and Bearer Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
854
854
854
855
855
856
857
18
AT332_TRR_E0

Table of Contents
12.4.1.6
12.4.1.7
12.4.1.8
12.4.1.9
12.4.2
12.4.3
12.4.4
12.4.4.1
Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859

Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860
C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861
C/(I+N) and Bearer Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862
Best Server Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 864
Service Area Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 864
Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
Channel Throughput, Cell Capacity, and Per-user Throughput Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
12.5
12.5.1
12.5.2
13

ACP Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .879
13.1
13.1.1
13.1.1.1
13.1.1.2
13.1.1.3
13.1.2
13.1.2.1
13.1.2.2
13.1.2.3
13.1.2.4
13.1.2.5
13.1.2.6
13.1.2.7
13.1.2.8
13.1.3
Objectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879
Quality Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879
Definition and Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879
Progressive Thresholds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
Target Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
ACP Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
GSM Quality Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
UMTS Quality Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
CDMA2000 Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
LTE Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
WiMAX Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882
Quality Indicator Parameters and Reference Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 883
Advanced Objective Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 883
Cost Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 883
Atoll and ACP Predictions Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 883
13.2
13.2.1
13.2.2
13.2.2.1
13.2.2.2
13.2.2.3
13.2.2.4
13.2.2.5
13.2.3
13.2.4
Quality Predictions and Antenna Masking Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884

Optimised Masking Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884
Antenna Masking Methods for Non-Native Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885
Basic Masking Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885
Improved Masking Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885
Full Path Loss Masking Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885
Antenna Correction Masking Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Antenna Masking and Repeaters, Remote Antennas, Secondary Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
CrossWave Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Aster Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
13.3
13.3.1
13.3.1.1
13.3.1.2
13.3.1.3
Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Configuring an Optimisation Setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Antenna Setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887
Additional Electrical Tilt (AEDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887
Relative Electrical Tilt Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888
13.4
13.4.1
13.4.2
Multi-RAT and Co-planning Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888

Multi-RAT and Co-planning Modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888
Technology Layer Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889
13.5
13.5.1
13.5.2
13.5.3
13.5.4
13.5.5
13.5.5.1
13.5.5.2
13.5.6
13.5.7
13.5.8
13.5.9
13.5.10
13.5.10.1
13.5.10.2
Optimisation Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889

Search Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 890
Tuning Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 890
Sorting Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
Principles of the Pre-optimisation of Activated Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
Global Score Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892
Search Algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892
Tuning Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892
Weighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893
Controlling the Optimisation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893
Implementation Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893
Memory Usage and Optimisation Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895
Internal Data Management and Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895
Memory Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895
Disk Space Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895
19

Table of Contents
13.6
13.6.1
13.6.2
13.6.2.1
13.6.2.2
13.6.2.3
13.6.2.4
13.6.2.5
13.6.2.6
13.6.3
13.6.3.1
13.6.3.2
13.6.4
13.6.4.1
13.6.4.2
13.6.5
Load Balancing Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Principle Used in ACP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Optimisation Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Capture for Load Balancing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cell Capacity Load Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Load Balancing Score Function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Load Quality Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Captured Traffic Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Introduction of Load Balancing as a Quality Indicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Quality Figures Used for Graphs and Statistics Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Load Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Average Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Optimisation Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Load Balancing Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Impact on the Global Score Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
895
895
896
896
896
897
897
898
898
899
899
899
900
900
900
901
13.7
13.7.1
13.7.2
13.7.3
13.7.3.1
13.7.3.2
13.7.4
13.7.5
13.7.6
Throughput Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Optimisation Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Difference between the RLC Peak Rate and Throughput Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Traffic Capture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Basic Configuration without Atoll Traffic Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Advanced Configuration with Atoll Traffic Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Throughput Score Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Impact on the Global Score Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Optimisation Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
901
902
902
902
902
902
903
903
904
13.8
13.8.1
13.8.1.1
13.8.1.2
13.8.1.3
13.8.1.4
13.8.1.5
13.8.2
13.8.3
EMF Exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Concepts of ACP EMF Exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Propagation Classes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Terrain Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Distribution of Evaluation Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Contribution of Transmitter Power to EMF Exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Worst-case Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Workflow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
EMF Exposure Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
904
904
904
905
905
905
905
905
906
13.9
Shadowing Margin and Indoor Coverage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907
13.10
13.10.1
13.10.2
13.10.3
13.10.4
13.10.5
13.10.5.1
13.10.5.2
13.10.5.3
13.10.5.4
13.10.5.5
13.10.6
Multi-Storey Optimisation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Path Loss Calculation and Data Caching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Pixel Weighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Concepts of ACP EMF Exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Propagation Classes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Terrain Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Distribution of Evaluation Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Contribution of Transmitter Power to EMF Exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Worst-case Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Workflow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13.11
ACP Software Data Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 910
20
907
907
908
908
908
909
909
909
909
910
910
910

Introduction
AT332_TRR_E0
Atoll 3.3.2 Technical Reference Guide

This Technical Reference Guide is aimed at radio network engineers with an advanced knowledge of Atoll and radio network
planning. It provides detailed information about the inner workings and formulas that are implemented by Atoll.
About Atoll
Atoll is a 64-bit multi-technology wireless network design and optimisation platform. Atoll is open, scalable, flexible, and
supports wireless operators throughout the network life cycle, from initial design to densification and optimisation.
Atoll includes integrated single RAN multiple RAT network design capabilities for both 3GPP (GSM/UMTS/LTE) and 3GPP2
(CDMA/LTE) technology streams. It provides operators and vendors with a powerful native 64-bit framework for designing
and optimising current and future integrated multi-technology networks. Atoll supports multi-technology HetNets, small cell
planning, and Wi-Fi offloading.
Atolls integration and automation features help operators smoothly automate planning and optimisation processes through
flexible scripting and SOA-based mechanisms. Atoll supports a wide range of implementation scenarios, from standalone to
enterprise-wide server-based configurations.
If you are interested in learning more about Atoll, please contact your Forsk representative to inquire about our training solutions.
About Forsk
Forsk is an independent company providing radio planning and optimisation software solutions to the wireless industry since
1987.
In 1997, Forsk released the first version of Atoll, its flagship radio planning software. Since then, Atoll has evolved to become
a comprehensive radio planning and optimisation platform and, with more than 7000 installed licenses worldwide, has
reached the leading position on the global market. Atoll combines engineering and automation functions that enable operators to smoothly and gradually implement SON processes within their organisation.
Today, Forsk is a global supplier with over 450 customers in 120 countries and strategic partnerships with major players in the
industry. Forsk distributes and supports Atoll directly from offices and technical support centres in France, USA, and China as
well as through a worldwide network of distributors and partners.
Since the first release of Atoll, Forsk has been known for its capability to deliver tailored and turn-key radio planning and optimisation environments based on Atoll.
To help operators streamline their radio planning and optimisation processes, Forsk provides a complete range of implementation services, including integration with existing IT infrastructure, automation, as well as data migration, installation, and
training services.
Getting Help
The online help system that is installed with Atoll is designed to give you quick access to the information you need to use the
product effectively. It contains the same material as the Atoll 3.3.2 User Manual.
You can browse the online help from the Contents view, the Index view, or you can use the built-in Search feature.
You can also download manuals from the Forsk web site at:
http://www.forsk.com/MyForskAccount/
Printing Help Topics

You can print individual topics or chapters from the online help.
To print help topics or chapters:
1. In Atoll, click Help > Help Topics.
2. In the Contents tab, expand the table of contents.
3. Right-click the section or topic that you want to print and click Print. The Print Topics dialog box appears.
4. In the Print Topics dialog box, select what you want to print:
If you want to print a single topic, select Print the selected topic.
If you want to print an entire section, including all topics and sections in that section, select Print the selected
heading and all subtopics.
21

Introduction
5. Click OK.
About Atoll Documentation

The following PDF manuals are available for Atoll and Atoll Microwave and can be downloaded from the Forsk web site at:
Atoll User Manual

Atoll Administrator Manual
Atoll Data Structure Reference Guide
Atoll Technical Reference Guide
Atoll Task Automation Guide
Atoll Model Calibration Guide
To read PDF manuals, download Adobe Reader from the Adobe web site at:
http://get.adobe.com/reader/
Hardcopy manuals are also available. For more information, contact to your Forsk representative.
Contacting Technical Support

Forsk provides global technical support for its products and services. To contact the Forsk support team, visit the My Forsk
web site at:
Alternatively, depending on your geographic location, contact one of the following support teams:
Forsk Head Office

For regions other than North and Central America and China, contact the Forsk Head Office support team:
Tel.: +33 562 747 225

Fax: +33 562 747 211
Email: [email protected]
Opening Hours: Monday to Friday 9.00 am to 6.00 pm (GMT +1:00)
Forsk US
For North and Central America, contact the Forsk US support team:
Tel.: 1-888-GO-ATOLL (1-888-462-8655)

Fax: 1-312-674-4822
Opening Hours: Monday to Friday 8.00 am to 8.00 pm (Eastern Standard Time)
Forsk China
For China, contact the Forsk China support team:
Tel: +86 20 8557 0016

Fax: +86 20 8553 8285
Opening Hours: Monday to Friday 9.00am to 5.30pm (GMT+08:00) Beijing, Chongqing, Hong Kong, Urumqi.
22
Chapter 1
Antennas and
Equipment
This chapter covers the following topics:
"Antenna Attenuation" on page 25
"Antenna Pattern Smoothing" on page 27
"Power Received From Secondary Antennas" on page 29
"Transmitter Radio Equipment" on page 30
"Repeaters and Remote Antennas" on page 32
"Beamforming Smart Antenna Models" on page 43
"Grid-of-Beams Smart Antenna Model" on page 51
"Adaptive Beam Smart Antenna Model" on page 52
"Statistical Smart Antenna Gain Model" on page 53

2016 Forsk. All Rights Reserved
24

Chapter 1: Antennas and Equipment
AT332_TRR_E0
1 Antennas and Equipment

1.1 Antenna Attenuation
To determine the transmitter antenna attenuation, Atoll calculates the accurate azimuth and tilt angles and performs 3D
interpolation of the horizontal and vertical patterns.
1.1.1 Calculation of Azimuth and Tilt Angles

From the direction of the transmitter antenna and the receiver position relative to the transmitter, Atoll determines the
receiver position relative to the direction of the transmitter antenna (i.e. the direction of the transmitter-receiver path in the
transmitter antenna coordinate system).
aTx and eTx are respectively the transmitter (Tx) antenna azimuth and tilt in the coordinate system S 0 x y z .
aRx and eRx are respectively the azimuth and tilt of the receiver (Rx) in the coordinate system S 0 x y z .
d is the distance between the transmitter (Tx) and the receiver (Rx).
Figure 1.1: Azimuth and Tilt Computation

In the coordinate system S 0 x y z , the receiver coordinates are:
x Rx
cos e Rx sin a Rx d
y Rx =
cos e Rx cos a Rx d
z Rx
(1)
sin e Rx d
Let az and el respectively be the azimuth and tilt of the receiver in the transmitter antenna coordinate system S Tx x'' y'' z'' .
These angles describe the direction of the transmitter-receiver path in the transmitter antenna coordinate system. Therefore,
the receiver coordinates in S Tx x'' y'' z'' are:
x'' Rx
y'' Rx =
z'' Rx
cos el sin az d
cos el cos az d
sin el d
(2)
According to the figure above, we have the following relations:

x'
y' =
z'
cos a Tx sin a Tx 0
x
sin a Tx cos a Tx 0 y
z
0
0
1
(3)
and
25

1
0
0
x''
x'
y'' = 0 cos e Tx sin e Tx y'
z''
z'
0 sin e Tx cos e Tx
(4)
Therefore, the relation between the system S 0 x y z and the transmitter antenna system S Tx x'' y'' z'' is:
1
0
0
cos a Tx sin a Tx 0
x''
x
y'' = 0 cos e Tx sin e Tx sin a Tx cos a Tx 0 y
z''
z
0 sin e Tx cos e Tx
0
0
1
(5)
We get,
x''
y'' =
z''
cos a Tx
sin a Tx
0
x
cos e Tx sin a Tx cos e Tx cos a Tx sin e Tx y
z
sin e Tx sin a Tx sin e Tx cos a Tx cos e Tx
(6)
Then, substituting the receiver coordinates in the system S0 from Eq. (1) and the receiver coordinates in the system STx from
Eq. (2) in Eq. (6) leads to a system where two solutions are possible:
1st solution: If a Rx = a Tx , then az = 0 and el = eRx e Tx
2nd solution: If a Rx a Tx , then
1
az = atan ---------------------------------------------------------------------------------------cos e Tx
sin e Tx tan e Rx
----------------------------------- + ---------------------------------------------tan a Rx a Tx
sin a Rx a Tx
and
cos e Tx tan e Rx
sin e Tx
- + ----------------------------------------------
el = atan sin az ---------------------------------sin a Rx a Tx
tan a Rx a Tx
If sin az sin a Rx a Tx 0 , then az = az + 180
1.1.2 Antenna Pattern 3D Interpolation

The direction of transmitter-receiver path in the transmitter antenna coordinate system is given by angle values az and el.
Atoll considers these values to determine transmitter antenna attenuations in horizontal and vertical patterns.
It reads the following:
H(az): the attenuation in the horizontal pattern for the calculated azimuth angle az
H(a0): the attenuation in the horizontal pattern for the electrical azimuth angle a0
V(el): the attenuation in the vertical pattern for the calculated tilt angle el
V(e0): the attenuation in the vertical pattern for the electrical tilt angle e0
Then it calculates the antenna total attenuation, L antTx az el :

180 az a 0
az a 0
- H a 0 V el + ------------------- H 180 + a 0 V 180 el if |el| 90
L antTx az el = H az ---------------------------------180
180
Else: L antTx az el = V(el)
Atoll assumes that the horizontal and vertical patterns are cross-sections of a 3D pattern. If the electrical tilt is e0, the
horizontal pattern is a conical section with an elevation of e0 degrees off the horizontal plane, and if the electrical azimuth is
a0, the vertical pattern is a plane section with a rotation a0 degrees from the vertical plane, the description of the antenna
pattern must satisfy the following conditions:
H(a0)=V(e0)
and
H(180+a0)=V(180-e0)
For a0 = 0 and e0 = 0, the antenna pattern must satisfy the following:

H(0)=V(0)
26
and
H(180)=V(180)

AT332_TRR_E0
If the above constraints are satisfied, it implies that:
Interpolated horizontal and vertical patterns respectively fit in with the entered horizontal and vertical patterns, even
in case of electrical tilt, and
The contribution of both front and back parts of the vertical pattern are taken into account.
Otherwise, only the second point is guaranteed.
This interpolation is performed in dBs.

Angle values in formulas are stated in degrees.
This interpolation is not used with 3D antenna patterns.
1.1.3 Additional Electrical Downtilt Modelling

The additional electrical downtilt, AEDT, also referred to as remote electrical downtilt or REDT, introduces a conical
transformation of the 3D antenna pattern in the vertical axis. In order to take it into account, the vertical pattern is
transformed as follows:
V x = V x AEDT when x [ 90,90]
V x = V x + AEDT when x [90,270]
Where, the angle values are in degrees.
The vertical pattern transformation is represented below. The left picture shows the initial vertical pattern when there is no
electrical downtilt and the right one shows the vertical pattern transformation due to an electrical downtilt of 10.
Then, Atoll proceeds as explained in the previous section. It determines the antenna attenuation in the transformed vertical
pattern for the calculated tilt angle (V(el)) and applies the 3D interpolation formula in order to calculate the antenna total
attenuation, L antTx az el .
Figure 1.2: Vertical Pattern Transformation due to Electrical Downtilt
1.2 Antenna Pattern Smoothing

Empirical propagation models, like the Standard Propagation Model (SPM), require antenna pattern smoothing in the vertical
plane to simulate the effects of reflections and diffractions. Signal level predictions can be improved by smoothing the highattenuation points of the vertical pattern. You can smooth vertical as well as horizontal antenna patterns in Atoll.
The antenna pattern smoothing algorithm in Atoll first determines the peaks and nulls in the pattern within the smoothing
angle (ASmoothing) defined by the user. Peaks (P) are the lowest attenuation angles and nulls (N) are the highest attenuation
angles in the pattern. Then, it determines the nulls to be smoothed (NSmoothing) and their corresponding angles according to
the defined Peak-to-Null Deviation (DPeak-to-Null). DPeak-to-Null is the minimum difference of attenuation in dBs between two
peaks and a null between them. Finally, Atoll smoothes the pattern between 0 and the smoothing angle (ASmoothing) by
applying the smoothing factor (FSmoothing) defined by the user.
Lets take an example of an antenna pattern to be smoothed, as shown in Figure 1.3 on page 28. Let DPeak-to-Null be 10 dB,
ASmoothing = 90 degrees, and FSmoothing = 0.5.
27

Figure 1.3: Vertical Antenna Pattern

Atoll first determines the peaks and nulls in the part of the pattern to be smoothed by verifying the slopes of the pattern curve
at each angle.
Figure 1.4: Peaks and Nulls in the Antenna Pattern

Peaks (P) and Nulls (N)
Angle ()
Attenuation (dB)
0.1
15
33.5
21
13.2
30
37.6
38
16.9
49
32.2
67
15.6
Then, Atoll verifies whether the difference of attenuation at a given angle is DPeak-to-Null less than the before and after it. This
comparison determines the nulls to be smoothed (NSmoothing).
Nulls to be smoothed (NSmoothing)
28
Angle ()
Attenuation (dB)
15
33.5
30
37.6
49
32.2

AT332_TRR_E0
Once the nulls are known, Atoll applies the smoothing algorithm to all the attenuation values at all the angles between the
first peak, the null, and the last peak.
Smoothing Algorithm
For all nulls n N Smoothing surrounded by two peaks P1 and P2 at angles 1 and 2 ,
A 2 A 1
- i 1
A i Smoothed = A i F Smoothing A i A + ---------------------1 2 1
Where,
i is the angle in degrees from 1 to 2 incremented by 1 degree,
AAngle is the attenuation at any given angle which can be i, 1 or 2 , and
FSmoothing is the smoothing factor defined by the user.
1.3 Power Received From Secondary Antennas

When secondary antennas are installed on a transmitter, the signal level received from it is calculated as follows:
G ant m Tx
G ant i
---------------------P
1
Tx
Tx
i
P Tx X i ------------------L Tx
L Tx
i
------------------------------------------------------------------ +
-----------------------------------------
L ant i az i el i
L ant m az m el m
Tx
Tx
i
= -------------------------------------------------------------------------------------------------------------------------------- (not in dB1)

L model
P rec
Where,
PTx is the transmitter power (Ppilot in UMTS HSPA and CDMA2000, PP-CCPCH in TD-SCDMA, PPreamble in WiMAX, and PDLRS in
LTE),
i is the secondary antenna index,
xi is the percentage of power dedicated to the secondary antenna, i,
G ant m
Tx
is the gain of the main antenna installed on the transmitter,
LTx are transmitter losses (LTx=Ltotal-DL),

G ant i
Tx
is the gain of the secondary antenna, i, installed on the transmitter,
Lmodel is the path loss calculated by the propagation model,

L ant m az m el m is the attenuation due to main antenna pattern,
Tx
L ant i az i el i is the attenuation due to pattern of the secondary antenna, i.

Tx
The definition of angles, az and el, depends on the used calculation method.
1.
Method 1 (must be indicated in an Atoll.ini file):

azm: the difference between the receiver antenna azimuth and azimuth of the transmitter main antenna,
elm: the difference between the receiver antenna tilt and tilt of the transmitter main antenna,
azi : the difference between the receiver antenna azimuth and azimuth of the transmitter secondary antenna, i,
eli : the difference between the receiver antenna tilt and tilt of the transmitter secondary antenna, i,
Method 2 (default):
azm : the receiver azimuth in the coordinate system of the transmitter main antenna,
elm : the receiver tilt in the coordinate system of the transmitter main antenna,
azi : the receiver azimuth in the coordinate system of the transmitter secondary antenna, i,
eli : the receiver tilt in the coordinate system of the transmitter secondary antenna, i.
Formula cannot be directly calculated from components stated in dB and must be converted in linear values.
29

1.4 Transmitter Radio Equipment

Radio equipment such as TMA, feeder and BTS, are taken into account to evaluate:
Total UL and DL losses of transmitter ( L total UL L total DL ) and transmitter noise figure NF Tx in UMTS HSPA,
CDMA2000 1xRTT 1xEV-DO, TD-SCDMA, WiMAX, and LTE documents,

Transmitter total losses L Total in GSM GPRS EGPRS documents.
In Atoll, the transmitter-equipment pair is modelled a single entity. The entry to the BTS is considered the reference point
which is the location of the transmission/reception parameters.
Figure 1.5: Reference Point - Location of the Transmission/Reception parameters
According to the book Radio network planning and optimisation for UMTS by Laiho
J., Wacker A., Novosad T., the noise figure corresponds to the loss in case of passive
components. Therefore, feeder noise figure is equal to the cable uplink losses.
UL
NF Feeder = L Feeder
Loss and gain inputs specified in ATL documents must be positive values.
1.4.1 GSM Documents

Atoll calculates DL total losses as follows:
DL
DL
DL
DL
L Total DL = L TMA + L Feeder + L Misc + L BTS Conf

Where,
DL
L TMA is the TMA transmission loss.
L Feeder is the feeder transmission loss ( L Feeder = L Feeder I Feeder + L Connector , where L Feeder , I Feeder and L Connector
DL
DL
DL
DL
DL
DL
are respectively the feeder loss per metre, the transmission feeder length in metre and the connector transmission
loss).
DL
L Misc are the miscellaneous transmission losses.
L BTS Conf are the losses due to BTS configuration (BTS property).
DL
1.4.2 UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents

As the reference point is the BTS entry, the transmitter noise figure corresponds to the BTS noise figure. Therefore, we have
NF TX = NF BTS . Where NF BTS is the BTS noise figure.
Uplink Total Losses
Atoll calculates total UL losses as follows:
UL
UL
UL
UL
L Total UL = L Misc + L Feeder + L BTS Conf + NR Repeaters G Ant div G TMA

Where,
30
UL
L Misc are the miscellaneous reception losses (Transmitter property).

AT332_TRR_E0
UL
UL
UL
UL
UL
L Feeder are the feeder reception losses ( L Feeder = L Feeder I Feeder + L Connector , where L Feeder , I Feeder and
UL
L Connector are respectively the feeder loss per metre (Feeder property), the reception feeder length in metre
(Transmitter property) and the connector reception losses.
UL
G Ant div is the antenna diversity gain (Transmitter property). This gain does not exist in WiMAX and LTE documents.
NR Repeaters is the noise rise at transmitter due to repeaters. This parameter is taken into account only if the
UL
transmitter has active repeater(s). The noise rise at transmitter due to repeaters is calculated as follows:
NR Repeaters = 10 Log 1 +
------------------
NIM
Rp
r
For each active repeater ( k ), Atoll calculates a noise injection margin ( NIM Rp ). This is the difference between the
k
donor transmitter noise figure ( NF TX ) and the repeater noise figure received at the donor.
Rp k
NIM Rp = NF TX NF Rp + G amp L
r
k
TX Rp k
Where,
NF Rp is the repeater noise figure,

k
Rp k
G amp is the repeater amplification gain (repeater property),
For each active repeater ( k ), Atoll converts the noise injection margin ( NIM Rp ) to Watt. Then, it uses the values
TX R p k
are the losses between the donor transmitter and the repeater (repeater property).
k
to calculate the noise rise at the donor transmitter due to active repeaters ( NR Repeaters ).
G TMA is the gain due to TMA, which is calculated as follows:

WithoutTMA
G TMA = NF Composite
WithTMA
WithTMA
NF Composite
WithoutTMA
Where NF Composite and NF Composite
are the composite noise figures with and without TMA respectively.
Friis' equation is used to calculate the composite noise figure when there is a TMA.
WithTMA
NF Composite
NF Feeder
NF BTS
NF
-----------------------------------------TMA
------------------
10
10
10
10
1
10
1
+ ---------------------------------- + -----------------------------------------------
= 10 Log 10
UL
UL
UL
G TMA
G TMA
G Feeder
----------------------------------------------------
10
10
10
10
10
10
WithoutTMA
And, NF Composite
= NF BTS + NF Feeder
Where,
NF Feeder is the feeder noise figure.
NF TMA is the TMA noise figure.
NF BTS is the BTS noise figure.
G TMA is the TMA reception gain.
G Feeder is the feeder UL gain G Feeder = L Feeder .
L Feeder is the feeder reception loss ( L Feeder = L Feeder I Feeder + L Connector , where L Feeder , I Feeder and
UL
UL
UL
UL
UL
UL
UL
UL
UL
UL
L Connector are respectively the feeder loss per metre, the reception feeder length in metre and the connector
reception loss).
Downlink Total Losses
Atoll calculates total DL losses as follows.
31

DL
DL
DL

DL
L Total DL = L TMA + L Feeder + L Misc + L BTS Conf

Where,
DL
L TMA is the TMA transmission loss.
L Feeder is the feeder transmission loss ( L Feeder = L Feeder I Feeder + L Connector , where L Feeder , I Feeder and L Connector
DL
DL
DL
DL
DL
DL
are respectively the feeder loss per metre, the transmission feeder length in metre and the connector transmission
losses).
DL
L Misc are the miscellaneous transmission losses.
DL
1.5 Repeaters and Remote Antennas

A repeater receives, amplifies, and re-transmits the radiated or conducted RF carrier both in downlink and uplink. It has a
donor side and a server side. The donor side receives the signal from a donor (transmitter, repeater, or remote antenna), and
the server side amplifies and re-transmits the received signal. Repeaters increase the coverage area of their donors by retransmitting all the frequencies (TRXs in GSM, carriers in UMTS, CDMA2000 and TD-SCDMA, and channels in WiMAX and LTE
documents).
Donors and repeaters may be linked through:
Air:
Microwave Links:
Optical Fibre Links:
User-defined or calculated propagation losses

User-defined link losses
User-defined link losses
Remote antennas are antennas located far from the transmitters, at locations that would normally require long runs of feeder
cable. A remote antenna is connected to the base station with an optic fibre. Remote antennas allow you to ensure radio
coverage in an area without a new base station. In Atoll, remote antennas should be connected to base stations that do not
have any antennas. A remote antenna, as opposed to a repeater, does not have any equipment and therefore generates
neither amplification gain nor noise.
In UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE documents, Atoll calculates the signal level received from a repeater or a
remote antenna by determining the total downlink and uplink gains (described in "UMTS, CDMA2000, TD-SCDMA, WiMAX,
and LTE Documents" on page 32). In GSM documents, the received signal level from a repeater or a remote antenna is
calculated by determining the EIRP transmitted by the repeater or remote antenna (described in "GSM Documents" on
page 39).
The following sections describe how received signal levels, and the related downlink and uplink gains and EIRP, are calculated
from a repeater or remote antenna R with a donor D.
1.5.1 UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents

1.5.1.1 Signal Level Calculation
The received signal level (dBm) on a carrier ic from a donor D at a pixel/mobile Mi via a repeater or remote antenna R (see
Figure 1.6 on page 34) is calculated as follows:
R
R Mi
C DL ic = P DL ic + G Total L Path M Shadowing L Indoor + G
Mi
Mi
Mi
Mi
L Ant L Body L Misc DL
If a pixel/mobile Mi receives signals from the donor D and its repeater R, the total signal
D
strength is the sum of the two signals: C DL ic + C DL ic

The received signal level (dBm) from a pixel/mobile Mi at a donor D via a repeater or remote antenna R (see Figure 1.6 on
page 34) is calculated as follows:
Mi
Mi
R Mi
C UL = P UL + G Total L Path M Shadowing L Indoor + G
Mi
Mi
Mi
Mi
Here:
32
P DL ic is the downlink transmission power of a donor D on carrier ic.
P UL is the uplink transmission power of a pixel/mobile Mi.
Mi
L Ant L Body L Misc UL

AT332_TRR_E0
R
G Total is the total gain, user-defined or calculated as explained in "Total Gain Calculation" on page 34.
L Path is the path loss (dB) calculated as follows:
RM
RM
L Path = L Model + L Ant , with:
L Model is the path loss calculated using a propagation model.
L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the repeater or remote
antenna R.
M Shadowing is the shadowing margin.
L Indoor is the indoor loss.
Mi
Mi
is the terminal antenna gain for the pixel/mobile Mi.

is the terminal loss for the pixel/mobile Mi.
Mi
L Ant is the terminal antenna attenuation (from antenna patterns) calculated for the pixel/mobile Mi (available in
WiMAX and LTE only).
Mi
For calculating the useful signal level from the best serving cell, L Ant is determined in the
direction (H,V) = (0,0) from the antenna patterns of the antenna used by Mi. For
Mi
calculating the interfering signal level from any interferer, L Ant is determined in the
direction of the interfering cell from the antenna patterns of the antenna used by Mi,
while the antenna is pointed towards Mis best serving cell.
Mi
L Body is the body loss defined for the service used by the pixel/mobile Mi.
Mi
,G
Mi
Mi
Mi
, L Ant , and L Body are not used in all the calculations. For more information, see
the technology-specific chapters.

R
L Misc DL is the miscellaneous transmission losses defined for the repeater or remote antenna R.
L Misc UL is the miscellaneous reception losses defined for the repeater or remote antenna R.
33

Figure 1.6: UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE: Signal Level Calculation
1.5.1.2 Total Gain Calculation

The total gain is calculated from the donor transmitter reference point ( ) to the repeater or remote antenna reference point
( ) as follows:
Over-the-Air Repeaters
D
DR
R
G Total = L Total DL + G Ant L Model + G Donor Ant LDonor
RX Feeder
+ G Amp LCov
TX Feeder
+ G Cov Ant
Figure 1.7: Downlink Total Gain: Over-the-Air Repeaters
34

AT332_TRR_E0
Here:
D
L Total DL are the total downlink losses of the donor D.
G Ant is the gain of the antenna used at the donor D.
L Model is the path loss between the donor D and the repeater or remote antenna R. This can be user-defined or
DR
calculated using the selected propagation model. If you do not select a propagation model, the propagation losses
between the donor and the repeater or remote antenna are calculated using the ITU 526-5 propagation model.
R
G Donor Ant is the gain of the donor-side antenna used at the repeater or remote antenna R.
LDonor
are the donor-side reception feeder losses for the repeater or remote antenna R.
RX Feeder
R
G Amp is the amplifier gain of the repeater R. For remote antennas, this is 0.
LCov
are the coverage-side transmission feeder losses for the repeater or remote antenna R.
TX Feeder
G Cov Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R.
Secondary antennas are fully supported in the evaluation of the repeater gains.
Microwave Link Repeaters

DR
R
G Total = L MW + G Amp LCov
TX Feeder
+ G Cov Ant
Figure 1.8: Downlink Total Gain: Microwave Link Repeaters

Here:
DR
L MW are the user-defined microwave link losses between the donor D and the repeater or remote antenna R.
LCov
TX Feeder
Optical Fibre Link Repeaters and Remote Antennas

DR
R
G Total = L Fibre + G Amp LCov
TX Feeder
+ G Cov Ant
35

Figure 1.9: Downlink Total Gain: Optical Fibre Link Repeaters or Remote Antennas
Here:
DR
L Fibre are the user-defined optical fibre link losses between the donor D and the repeater or remote antenna R.
LCov
TX Feeder
Repeater Downlink Power Limitation

Atoll verifies that the downlink power after amplification is consistent with the repeater equipment limitation.
D
P DL ic + G Total P Max + G Cov Ant LCov
TX Feeder
Here:
P DL ic is the downlink transmission power of a donor D on carrier ic. When the donor has more than one cell, Atoll
considers the highest power.
R
G Total is the total downlink gain, user-defined or calculated as explained in "Total Gain Calculation" on page 34.
P Max is the maximum downlink power allowed by the equipment.
LCov
TX Feeder
1.5.1.3 Repeater Noise Figure

You can define and assign a repeater equipment to each repeater. In addition to the allowed ranges of gains and powers
allowed to each repeater, these equipment contain a noise figure which is applied to the repeater they are assigned to. This
noise figure has an impact on the donor total reception losses. For information, see "Transmitter Radio Equipment" on
page 30.
1.5.1.4 Appendix: Carrier Power and Interference Calculation

This section explains how Atoll calculates the received carrier power and interference when a transmitter has a connected
repeater.
A mobile receiver receives signal from the donor transmitter as well as its repeater. Similarly, the signal from the mobile is
received at the donor transmitter as well as its repeater. In practice, when a mobile receiver is in the vicinity of the donor
transmitter, the signal to/from the repeater would be very weak due to high pathloss between the repeater and the mobile
receiver. Similarly, when the mobile receiver is located in the vicinity of the repeater, the signal to/from the donor transmitter
would be very weak due to the same reason.
Atoll does not differentiate between the mobile receiver being in the transmitter coverage area or being in its repeater
coverage area. Atoll adds the signals received from the donor transmitter and its repeater to generate a combined pathloss
matrix that is associated with the donor transmitter and includes the effect of its repeater.
36

AT332_TRR_E0
Calculation of Total Path Loss

The total pathloss, L Total , is calculated by computing a downlink budget. If we take the case of a CDMA project, without
considering any shadowing margin or indoor loss, the power received from the donor transmitter, Txd on a carrier ic, at the
mobile receiver can be stated as (for a link over the air):
D
P Pilot ic G Ant
D
P Rec ic = ------------------------------------------D
D Mi
L Total DL L Path
Where,
D
P Rec ic is the carrier power received at the receiver from the donor transmitter on a carrier ic (in W)
D
P Pilot ic is the pilot power of the donor transmitter on the carrier ic (in W)
D
G Ant is the donor transmitter antenna gain.

D
L Total DL is the transmission feeder loss of the donor transmitter.

D Mi
L Path is the path loss between the donor transmitter and the mobile receiver.
Similarly, the power received at the mobile receiver from the repeater R is:
D
P Pilot ic G Total
R
P Rec ic = -------------------------------------------R Mi
L Path
Where,
R
P Rec ic is the carrier power received at the mobile receiver from the repeater on a carrier ic (in W)
D
P Pilot ic is the pilot power of the donor transmitter on the carrier ic (in W)
R
G Total is the total gain of repeater linked to a donor transmitter with an air link.
R Mi
L Path is the path loss between the repeater and the mobile receiver
So, the total carrier power received at the mobile receiver is:
D
R
G Ant
G Total
DR
R
R
D
P Rec ic = P Rec ic + P Rec ic = P Pilot ic ---------------------------------------------
- + ------------ Mi
R Mi
L DTotal DL L DPath
L Path
Since,
D
P Pilot ic G Ant
L Total = -----------------------------------------------DR
D
L Total DL P Rec ic
Therefore,
D
P Pilot ic G Ant
L Total = ----------------------------------------------------------------------------------------------------------------------------D
R
G Ant
G Total
D
D
L Total DL P Pilot ic ---------------------------------------------- + --------------
Mi
R Mi
L DTotal DL L DPath
L Path
Hence,
D
G Ant
L Total = --------------------------------------------------------------------------------------------------D
R
G Total
G Ant
D
- + ------------L Total DL ---------------------------------------------
Mi
R Mi
L DTotal DL L DPath
L Path
This total path loss depends on the location of the mobile receiver in realistic network scenarios. As a mobile in the donor
transmitter/repeater coverage area is likely to be far from the repeater/donor transmitter coverage area, the respective
pathloss value will be very large. This implies that we can study the two cases separately without influencing the results much.
Case 1: Receiver in Donor Transmitter Coverage Area
37


R
G Total
R Mi
- can be ignored. This implies that:
L Path is likely to be very high, so the term ------------R Mi
L Path
D Mi
L Total = L Path
Considering this total pathloss value, the total received power in the uplink and in the downlink can be stated as:
D
P Pilot ic G Ant
P Pilot ic G Ant
D
P Rec DL ic = ------------------------------------------- = --------------------------------------------D
D
D Mi
L Total DL L Total
L Total DL L Path
Mi
Mi
P Output ic G Ant
P Output ic G Ant
D
- = ---------------------------------------------P Rec UL ic = ---------------------------------------------D
D
D Mi
L Total UL L Total
L Total UL L Path
Where,
Mi
P Output ic is the transmitted power from the mobile terminal on the carrier ic (in W)
D
L Total UL is the reception feeder loss of the transmitter
Case 2: Receiver in Repeater Coverage Area

D
G Ant
D Mi
- can be ignored. This implies that:
L Path is likely to be very high, so the term --------------------------------------------D
D Mi
L Total DL L Path
D
G Ant
G Ant
L Total = --------------------------------------------------- = ---------------------------------------R
R
R
G Total L Total DL
D
G Total
-----------------------------------------
L Total DL ------------R Mi
Mi
L RPath
L Path
D
P Pilot ic G Ant
P Pilot ic G Total
D
- = -------------------------------------------P Rec DL ic = ------------------------------------------D
R Mi
L Total DL L Total
L Path
Mi
Mi
D
P Output ic G Total L Total DL
P Output ic G Ant
D
- ---------------------P Rec UL = ---------------------------------------------- = ------------------------------------------------R Mi
D
D
L Path
L Total UL
L Total UL L Total
Where,
Mi
P Output ic is the transmitted power from the mobile terminal (in W)

D
L Total UL is the reception feeder loss of the transmitter

Calculation of Eb/Nt Uplink
In the uplink, the quality level at the transmitter on a traffic channel is:
E b
C W
---= --- ---- N t UL
I R
Where,
C is the carrier power received from the mobile terminal (in W)
I is the total interference (in W)
W is the spreading bandwidth (Hz)
R is the effective service throughput in the uplink (bits/s)
(W/R is the service processing gain in the uplink)
C and I are both evaluated at the same reference point, which is the entry of BTS using the following formulas.
Mi
P Output G Ant
C = P Total UL = --------------------------------------D
L Total UL L Total
I = I Total + N 0
38

AT332_TRR_E0
Where,
I Total is the sum of the signals received from mobile terminals inside the same cell and those outside (in W)
N 0 is the transmitter equipment thermal noise (in W)
Therefore, for each mobile terminal Mi,
I Total =
Mi
Mi
P Output G Ant
--------------------------------------
L DTotal UL L Mi
Total
And,
D
N 0 = NF K T W
Where,
D
NF is the noise figure of the transmitter equipment at the reference point, i.e. the entry of the BTS
K is Boltzman constant
T is the ambient temperature (in K)
Hence
N 0 = NF
BTS
KTW
1.5.2 GSM Documents

1.5.2.1 Signal Level Calculation
The received signal level (dBm) on a TRX type tt from a donor D at a pixel/mobile Mi via a repeater or remote antenna R (see
Figure 1.10 on page 40) is calculated as follows:
R
R Mi
C DL tt = EIRP DL tt P tt L Path M Shadowing L Indoor + G
Mi
Mi
L Misc DL
If a pixel/mobile Mi receives signals from the donor D and its repeater R, the total signal
D
strength is the sum of the two signals: C DL tt + C DL tt

Here:
R
EIRP DL tt is the effective isotropic radiated power of the repeater or remote antenna R on the TRX type tt. It can be
user-defined or calculated as explained in "EIRP Calculation" on page 40.

the downlink transmission power of a donor D on carrier ic.
P tt is the power offset defined for the TRX type tt.
R Mi
R Mi
L Path = L Model + L Ant , with:
L Model is the path loss calculated using a propagation model.
L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the repeater or remote
antenna R.
M Shadowing is the shadowing margin.
L Indoor is the indoor loss.
is the terminal antenna gain for the pixel/mobile Mi.
R
L Misc DL
is the terminal loss for the pixel/mobile Mi.

is the miscellaneous transmission losses defined for the repeater or remote antenna R.
39

s
Figure 1.10: GSM: Signal Level Calculation
1.5.2.2 EIRP Calculation

D
The EIRP of a repeater or remote antenna R is calculated at the repeater or remote antenna reference point ( ) w. r. t. P DL
at the donor reference point ( ) as follows:
Over-the-Air Repeaters
R
DR
EIRP DL tt = P DL L Total DL + G Ant L Model + G Donor Ant LDonor
RX Feeder
Figure 1.11: EIRP: Over-the-Air Repeaters

Here:
40
P DL is the downlink transmission power of the donor D.
L Total DL are the total downlink losses of the donor D.
+ G Amp LCov
TX Feeder
+ G Cov Ant

AT332_TRR_E0
D
G Ant is the gain of the antenna used at the donor D.
L Model is the path loss between the donor D and the repeater or remote antenna R. This can be user-defined or
DR
calculated using the selected propagation model. If you do not select a propagation model, the propagation losses
between the donor and the repeater or remote antenna are calculated using the ITU 526-5 propagation model.
R
G Donor Ant is the gain of the donor-side antenna used at the repeater or remote antenna R.
LDonor
are the donor-side reception feeder losses for the repeater or remote antenna R.
RX Feeder
R
LCov
TX Feeder
Secondary antennas are fully supported in the evaluation of the repeater gains.
Microwave Link Repeaters

D
DR
R
EIRP DL tt = P DL L MW + G Amp LCov
TX Feeder
+ G Cov Ant
Figure 1.12: Downlink Total Gain: Microwave Link Repeaters

Here:
D
L MW are the user-defined microwave link losses between the donor D and the repeater or remote antenna R.
LCov
DR
R
TX Feeder
Optical Fibre Link Repeaters and Remote Antennas

D
DR
R
EIRP DL tt = P DL L Fibre + G Amp LCov
TX Feeder
+ G Cov Ant
41

Figure 1.13: Downlink Total Gain: Optical Fibre Link Repeaters or Remote Antennas
Here:
D
L Fibre are the user-defined optical fibre link losses between the donor D and the repeater or remote antenna R.
LCov
DR
R
TX Feeder
R
Repeater Downlink Power Limitation

Atoll verifies that the EIRP after amplification is consistent with the repeater equipment limitation.
R
EIRP DL tt P Max + G Cov Ant LCov
TX Feeder
Here:
R
EIRP DL tt is the effective isotropic radiated power of the repeater R on the TRX type tt.
P Max is the maximum downlink power allowed by the equipment.
LCov
TX Feeder
1.5.3 Donor-side Parameter Calculations

1.5.3.1 Azimuth
This is the angle at which the donor antenna is situated with respect to the North at the repeater or remote antenna. This
angle is measured clock-wise as shown in the figure below. It is the absolute horizontal angle at which the donor-side antenna
of the repeater should be pointed in order to be aligned with the donor antenna.
Figure 1.14: Angle from North (Azimuth)
1.5.3.2 Mechanical Downtilt

This is the tilt angle for the repeaters donor-side antenna, which ensures that it points towards the donor antenna in the
vertical plane. As a general rule, downtilt angles are considered positive and uptilt angles negative.
42

AT332_TRR_E0
Figure 1.15: Positive/Negative Mechanical Downtilt

Since this parameter depends on the difference of heights/altitudes between the donor transmitter and the repeater, it can
be automatically calculated in the repeaters Donor side properties. If the height/altitude of the antenna is modified, the
corresponding tilt angle can be found out and applied using the Calculate button.
Example
Figure 1.16: Tilt Angle Computation

The tilt angle repeaters donor-side antenna in the above figure would be:
R
H Donor Ant H Ant

R
-
T Donor Ant = atan ------------------------------------------DR
D
As obvious, this angle will be negative for uptilts and positive for downtilts of the antenna.
Here:
R
H Donor Ant is the height of the donor-side antenna of the repeater or remote antenna R.
H Ant is the height of the antenna of the donor D.
DR
is the distance between the antenna of the donor D and the antenna of the repeater or remote antenna R.
1.6 Beamforming Smart Antenna Models

Adaptive antenna systems use more than one antenna elements, along with smart signal processing, to locate and track
various types of signals, to dynamically minimize interference, and maximize useful signal reception. The signal processor
dynamically applies weights to each element of the adaptive antenna system to create array patterns in real-time.
Beamforming smart antennas dynamically create antenna patterns with a main beam pointed in the direction of the user
being served, i.e., the useful signal. Adaptive algorithms can also be used in order to minimize the interference received by
the cells. These algorithms are based on optimization methods such as the minimum mean square error method.
The following beamforming smart antenna models are available in Atoll. These smart antenna models support linear adaptive
array systems, such as the one shown in Figure 1.17 on page 44.
Optimum Beamformer: The Optimum Beamformer smart antenna model performs dynamic beamforming in
downlink as explained in "Downlink Beamforming" on page 46, and beamforming and interference cancellation in
uplink using the minimum mean square error algorithm as explained in "Uplink Beamforming and Interference
Cancellation (MMSE)" on page 49. Smart antenna results are later on used in coverage prediction calculations.
Conventional Beamformer: The Conventional Beamformer smart antenna model performs dynamic beamforming in
downlink and uplink as explained in "Downlink Beamforming" on page 46 and "Uplink Beamforming" on page 48,
respectively. Smart antenna results are later on used in coverage prediction calculations.
43

Figure 1.17: Linear Adaptive Antenna Array

In the following explanations, we assume:
There are a total of E SA elements in the adaptive antenna system.
is the angle of arrival for the useful signal.
is the angle at which we want to calculate the smart antenna gain.

d is the distance between two adjacent antenna elements.
1.6.1 Definitions and Formulas

The tables in the following subsections list the parameters and formulas used in beamforming smart antenna models.
1.6.1.1 Definitions
Name
Value
Unit
Description
E SA
Smart antenna model parameter
None
Number of smart antenna elements
Calculation parameter
Degrees
Angle of arrival for the useful signal
Calculation parameter
Degrees
Angle at which the smart antenna

effect is calculated
--- , where is the wavelength of the signal

2
Distance between two adjacent

antenna elements
1.6.1.2 Downlink Beamforming

Name
Value
Unit
Description
gn
Smart antenna model parameter
None
Gain of a single element
None
Steering vector for the direction of
None
Complex smart antenna weight
None
Array correlation matrix for a given

user direction
None
Smart antenna gain in any direction
1 e
2
j ------ d sin
... e
wn
e
44
2
j ------ 2d sin
T
2
j ------ E 1 d sin
SA
2
j ------ nd sin
j n sin
with d =
--2
H
S S
G SA
g n S R S = g n S S S S = g n E SA

AT332_TRR_E0
Name
Value
Unit
Description
None
Downlink array correlation matrix for

iteration k
None
Average downlink array correlation

matrix over a simulation (K
iterations)
pj Rj
Rk
j=1
K
--1-
K
R Avg
Rk
k=1
1.6.1.3 Uplink Beamforming

Name
Value
Unit
Description
S
----------E SA
None
Vector of ESA complex weights for

the conventional beamformer
None
Total noise correlation matrix
None
Thermal noise correlation matrix
None
Interference correlation matrix
RN
Rn + RI = n I +
pj Sj Sj
j=1
2
n I
Rn
J
pj Sj Sj
RI
j=1
H
PN
w RN w
Total uplink noise power
p w S S w = p E SA
Total power received from the

served user
CINR UL
p E SA
P
------ = -------------------------H
PN
w RN w
None
C/(I+N) in the uplink

(WiMAX)
Q UL
p E SA
P
------ = -------------------------H
PN
w RN w
None
Signal quality in the uplink

(TD-SCDMA)
G SA
E SA
None
Uplink smart antenna beamforming

gain in the direction of the served
user
Average noise correlation matrix
Uplink interference
None
Angular distribution of uplink noise

rise
SA
RN
--1-
K
Avg
RN k
k=1
I UL
w RN
Avg
w n
2
I UL + n
--------------------------2
n
NR UL
1.6.1.4 Uplink Beamforming and Interference Cancellation (MMSE)

Name
Value
Unit
Description
RN S
None
Vector of ESA complex weights for

the optimum beamformer
E SA
----------------------------H
1
S RN S
None
MMSE optimization constant
None
Total noise correlation matrix
None
Thermal noise correlation matrix
RN
Rn + RI =
2
n
I+
pj Sj Sj
j=1
Rn
n I
45

Name
Value
Unit
Description
None
Interference correlation matrix
Total uplink noise power (optimum

beamformer)
Total power received from the

served user (optimum beamformer)
pj Sj Sj
RI
j=1
P N
S RN S
p S RN S
CINR UL
P
H
P
1
------ = ------ = p S R N S
P
PN
N
None
C/(I+N) in the uplink

(WiMAX)
Q UL
P
H
P
1
------ = ------ = p S R N S
PN
P N
None
Signal quality in the uplink

(TD-SCDMA)
G SA
S I S = E SA
None
Uplink smart antenna beamforming

gain in the direction of the served
user
Average inverse noise correlation

matrix
Uplink interference
None
Angular distribution of uplink noise

rise
SA
--1-
K
1
RN
Avg
RN
1
k
k=1
E SA
2
------------------------------------ n
H
1
S RN
S
I UL
Avg
I UL + n
--------------------------2
n
NR UL
1.6.2 Downlink Beamforming
Figure 1.18: Downlink Beamforming

Beamforming dynamically creates a beam towards the served user. The smart antenna processor applies complex weights,
w n , to each antenna element in order to form a beam towards the served user. The magnitude of these complex weights is
set to 1. The beamforming is performed using only the phase of the complex weights. The steering vector, S , representing
the complex weights for forming a beam towards the served user, i.e., at the angle of arrival is given by:
S = 1 e
2
j ------ d sin
2
j ------ 2d sin
... e
T
2
j ------ E SA 1 d sin
Where T represents the transpose of a matrix.
46

AT332_TRR_E0
Therefore, the complex weight at any nth antenna element can be given by:
wn = e
2
j ------ nd sin
j n sin
In Atoll, d = --- , therefore, w n = e

.
2
The smart antenna gain in any direction can be given by:

H
G SA = g n S R S
Where H represents the Hilbert transform, which is the complex conjugate transpose of a matrix, g n is the gain of the nth
antenna element in the direction , and R is the array correlation matrix for a given user direction , given by:
H
R = S S
For the direction of the served user, i.e., , the smart antenna gain is calculated as follows:
H
G SA = g n S R S = g n S S S S = g n E SA
The smart antenna gain includes the gain of the beamforming as well as the gain of power combination.
The smart antenna gain in dB will be G SA = 10 Log G SA .
The smart antenna is able to form the beam only in the horizontal plane, therefore, the vertical pattern is assumed to remain
the same.
Power Combining Gain
Cell transmission power is fed to each antenna element of the smart antenna system. Since each element transmits the same
input power, this results in a gain due to power combination, i.e., the powers fed to each antenna element are combined for
transmission.
Additional Processing in Monte Carlo Simulations
During Monte Carlo simulations, Atoll calculates the smart antenna gains (array correlation matrix R ) for each served mobile
in a cells coverage area in each iteration. The sum of these array correlation matrices for all the users served in one iteration
k is calculated as follows:
J
Rk =
pj Rj
j=1
Where R k for any cell is the downlink array correlation matrix for iteration k, J is the number of served mobiles during the
iteration, p j is the EIRP transmitted towards the mobile j, and R j is the array correlation matrix for the mobile j.
Atoll calculates a moving average of the array correlation matrices calculated in each iteration. At the end of a simulation with
K iterations, the average downlink array correlation matrix for any cell is given by:
K
R Avg
1
= ---
K
Rk
k=1
47

1.6.3 Uplink Beamforming
Figure 1.19: Uplink Beamforming

Let w represent the vector of ESA complex weights for the beamformer. w is given by:
S
w = ----------E SA
Where S is the steering vector in the direction of the served user, .
The total noise received in the uplink, i.e., interference and thermal noise, is stored in a total noise correlation matrix, R N .
The total noise correlation matrix is the sum of the thermal noise correlation matrix R n , and the interference correlation
matrix R I , given by:
J
2
RN = Rn + RI = n I +
pj Sj Sj
j=1
J
Where R n =
2
n
I and R I =
pj Sj Sj
j=1
2
n
is the thermal noise power. I is the identity matrix. p j is the power received by one element of the smart antenna from
the jth interfering mobile. S j is the steering vector in the direction of the jth interfering mobile, . J is the total number of
interfering mobiles.
The total noise power, including thermal noise and interference from all uplink interferers, received by a cell is given by:
H
PN = w RN w
And, the total power received from the served user is given by:
H
P = p w S S w = p E SA
Where p is the power received by one element of the smart antenna from the served user.
In TD-SCDMA, the uplink signal quality is calculated by:
p E SA
P
SA
Q UL = ------ = -------------------------H
PN
w RN w
In WiMAX, the C/(I+N) in the uplink is then calculated by:
p E SA
P
CINR UL = ------ = -------------------------H
PN
w RN w
From the above equation, we can determine the uplink smart antenna beamforming gain in the direction of the served user,
which equals the number of smart antenna elements, i.e., G SA = E SA .
48

AT332_TRR_E0

The noise correlation matrix R N for each iteration k includes the effect of the matrix calculated for the previous iteration. The
result is the angular distribution of the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX), which is calculated from the
noise correlation matrix obtained at the end of the last iteration of a Monte Carlo simulation. This angular distribution of the
uplink load (TD-SCDMA) or the uplink noise rise (WiMAX) can be stored in the Cells table. The average of the noise correlation
matrices is calculated as follows:
K
RN
Avg
1
= ---
K
RN k
k=1
Where R N
Avg
is the average of the noise correlation matrices of all the iterations from k = 1 to K, and R N
is the noise
correlation matrix of the kth iteration.

The interference can be isolated from the thermal noise and can be calculated for any direction using the formula.
H
I UL = w R N
Avg
w n
Where I UL is the interfering signal in the direction , ESA is the number of smart antenna elements, S is the steering
2
vector in the direction , and n is the thermal noise power.

In TD-SCDMA, the uplink load is calculated from the average noise correlation matrix. In WiMAX, the angular distribution of
the uplink noise rise is given by:
2
I UL + n
NRUL = --------------------------2
n
1.6.4 Uplink Beamforming and Interference Cancellation (MMSE)

The optimum beamformer uses the Minimum Mean Square Error algorithm in the uplink in order to cancel interference. The
Minimum Mean Square Error algorithm optimizes the useful signal as well as maximizes the signal quality.
A simple null steering beamformer can cancel the interference from the most interfering E SA 1 interfering mobiles. The
optimum beamforming method used in Atoll overcomes this limitation. It calculates the optimum smart antenna weights
using the knowledge of directions and power levels of interference. These weights do not try to fully cancel E SA 1
interference signals, but rather try to reduce the overall received interference as much as possible.
Figure 1.20: Uplink Adaptive Algorithm

represent the vector of ESA complex weights for the beamformer. w
is given by:
Let w
= R N1 S
w
49

Where S is the steering vector in the direction of the served user, . , which is a constant value for a given useful signal
that optimizes the beamformer weights. It is given by the equation:
E SA
= ----------------------------H
1
S RN S
1
R N is the inverse of the total noise correlation matrix. The total noise correlation matrix is the sum of the thermal noise
correlation matrix R n , and the interference correlation matrix R I , given by:
J
RN = Rn + RI =
2
n
I+
pj Sj Sj
j=1
J
Where R n =
2
n
pj Sj Sj
I and R I =
j=1
2
n is the thermal noise power. I is the identity matrix. p j is the power received by one element of the smart antenna from
the jth interfering mobile. S j is the steering vector in the direction of the jth interfering mobile, . J is the total number of
interfering mobiles.
The total noise power, including thermal noise and interference from all uplink interferers, received by a cell is given by:
H
2
1
P N = S R N S
And, the total power received from the served user is given by:
2
H
2
1
P = p S R N S
Where p is the power received by one element of the smart antenna from the served user.
In TD-SCDMA, the uplink signal quality is calculated by:
H
P
SA
1
Q UL = ------ = p S R N S
P N
In WiMAX, the C/(I+N) in the uplink is then calculated by:

H
P
1
CINR UL = ------ = p S R N S
P N
From the above equation, we can determine the uplink smart antenna beamforming gain in the direction of the served user.
RSCP TCH UL (TD-SCDMA) or C UL (WiMAX) can be calculated from the above equation by considering the interference and
1
noise to be null, i.e., R N = I . This gives:

H
In TD-SCDMA, RSCP TCH UL = p S I S = p E SA

H
In WiMAX, C UL = p S I S = p E SA
From the above equation, the uplink smart antenna beamforming gain equals the number of smart antenna elements, i.e.,
G SA = E SA .
1
The inverse noise correlation matrix R N for each iteration k includes the effect of the matrix calculated for the previous
iteration. Hence, Atoll is able to calculate an average of the smart antenna interference-cancellation effect. The result is the
angular distribution of the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX), which is calculated from the inverse of
the noise correlation matrix obtained at the end of the last iteration of a Monte Carlo simulation. This angular distribution of
the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX) can be stored in the Cells table. The average of the inverse noise
correlation matrices is calculated as follows:
K
1
RN
Avg
1
= ---
K
RN
k=1
50

AT332_TRR_E0
1
Where R N
Avg
is the average of the inverse noise correlation matrices of all the iterations from k = 1 to K, and R N
is the
inverse noise correlation matrix of the kth iteration.

The interference can be isolated from the thermal noise and can be calculated for any direction using the formula.
E SA
2
I UL = ------------------------------------ n
H
1
S RN
S
Avg
Where I UL is the interfering signal in the direction , ESA is the number of smart antenna elements, S is the steering
2
vector in the direction , and n is the thermal noise power.

In TD-SCDMA, the uplink load is calculated from the average inverse noise correlation matrix. In WiMAX, the angular
distribution of the uplink noise rise is given by:
2
I UL + n
NRUL = --------------------------2
n
1.7 Grid-of-Beams Smart Antenna Model

A grid-of-beams smart antenna, called GOB, consists of more than one directional antenna pattern (beam) in different
directions. Each beam of a GOB has a different azimuth so that the GOB as a whole covers an entire sector. During the
simulations, Atoll determines the most suitable beam from the GOB for each user served by the smart antenna. The most
suitable beam (best beam) is the one which provides the highest gain towards the served user:
BeamBest = Beam
H
H
V
Max G Beam L Beam L Beam
V
Where G Beam , L Beam , and L Beam are the gains, horizontal, and vertical attenuations of the beams of the GOB. In words, the
best beam is the one among all the beams of a GOB that has the highest difference between gain, and horizontal and vertical
SA
SA
SA
SA
attenuations. The gains and losses of the GOB ( G DL , G UL , L DL , and L UL ) are determined from the selected best beam.
The following example shows how Atoll calculates the GOB gains and losses.
Example:
Let us assume a GOB with 5 beams that have the same vertical patterns, and whose horizontal patterns are pointed towards
different directions as shown in the figure below:
Figure 1.21: Grid Of Beams Modelling

Let us assume that all the beams and the main antenna have the same 18 dBi gain, and the vertical attenuation at the user
location is 15 dB, which is also the same for all the beams because we assume that the vertical patterns are the same.
51

If the user is located at = 70 azimuth, as shown in the figure below, Atoll determines the best beam, which has the
highest gain towards , as follows:
Horizontal
Vertical
Attenuation (dB) Attenuation (dB)
G Beam L Beam L Beam
Total Gain
(dB)
15
18 - 60 - 15
-57
60
15
18 - 60 - 15
-57
18
2.21
15
18 - 2.21- 15
0.79
-30
18
60
15
18 - 60 - 15
-57
-60
18
60
15
18 - 60 - 15
-57
Beam
Gain (dBi)
18
60
30
18
60
Transmitter
Centre of the pixel where the served user is
located
Angle between the user and the transmitter

azimuth
Figure 1.22: GOB Modelling - Determination of the Best Beam

In our example, the total gain of the beam at 60 is the highest. Therefore this beam is selected as the best beam.
If this beam has been selected in the downlink,
SA
SA
G DL = 18 dB and L DL = L Beam + L Beam = 17,21 dB

If this beam has been selected in the uplink,
SA
SA
G UL = 18 dB and L UL = L Beam + L Beam = 17,21 dB
1.8 Adaptive Beam Smart Antenna Model

An adaptive beam smart antenna is capable of steering a given antenna pattern towards the direction of the served signal. In
Atoll, this is modelled using a single antenna pattern, called a beam because of its highly directional shape. During the
simulations, this adaptive beam is oriented in the direction of each served user in order to model the effect of the smart
antenna.
SA
SA
SA
The adaptive beam gains ( G DL and G UL ) are the antenna gains defined for the beam, and the adaptive beam losses ( L DL and
SA
L UL ) are the horizontal and vertical pattern attenuations L Beam + L Beam towards the user direction.
The following example shows how Atoll calculates the adaptive beam gains and losses.
Example:
Let us assume an adaptive beam smart antenna selected for a transmitter along with a main antenna. Let us assume that the
adaptive beam and the main antenna have the same 18 dBi gain, and the vertical attenuation at the user location is 15 dB.
If the user is located at = 60 azimuth, as shown in the figure below:
52

AT332_TRR_E0
Transmitter
located

azimuth
Figure 1.23: Adaptive Beam Modelling - Determination of the Best Beam

If the adaptive beam smart antenna is selected in the downlink, the gain and losses of the adaptive beam at are:
SA
SA
G DL = 18 dB and L DL = L Beam + L Beam = 15 dB

If the adaptive beam smart antenna is selected in the uplink, the gain and losses of the adaptive beam at are:
SA
SA
G UL = 18 dB and L UL = L Beam + L Beam = 15 dB

H
In fact, as the ideal beam steering algorithm steers the beam towards the served user, L Beam = 0 . These values are used in
interference calculation to determine the downlink interfering signal due to transmission towards the served user, as well as
for calculating the uplink interfering signals received at transmitter when decoding signal received from the served user.
1.9 Statistical Smart Antenna Gain Model

A statistical modelling approach is also available in Atoll which can be used to model the effect of smart antennas through C/
I gains. You can create smart antenna equipment in Atoll based on the statistical approach by providing C/I gains and their
cumulative probabilities for different spreading angles, Spread .
You can assign a spreading angle to each clutter class in your document. Atoll reads the clutter class in which the served user
is located to determine the spreading angle. Different clutter types have different spreading effects on the propagation of
radio waves. Urban and dense urban clutter types introduce more multipath and spread the signal at a wider angle than an
open or rural clutte type.
Once you have assigned the spreading angles to clutter classes, you can enter the C/I gains and their cumulative probabilities
for each spreading angle, in the smart antenna equipment based on the statistical model. For each smart antenna equipment
based on statistical modelling, you can set a probability threshold, TProb
SA
To find the smart antenna gain, Atoll determines the clutter class of the served user, it reads the spreading angle from the
clutter class properties, it reads the probability threshold from the smart antenna properties, and reads the smart antenna C/
I gain defined for the Probability = 1 TProb
SA
corresponding to the spreading angle.
The following example shows how Atoll calculates the statistical C/I gains and losses.
Example:
Let us assume that the served user is located at a an urban clutter class with Spread = 10 . The smart antenna equipment
SA
SA
has TProb = 80 % . Atoll will read the smart antenna C/I gain G for Prob = 20 % . If a gain for the exact probability
value of 20% is not defined, Atoll linearly interpolates the gain value from the two surrounding values.
If G
SA
Prob = 19%
= 4,6298 dB and G
SA
Prob = 20,4%
= 4,7196 dB , then G
SA
Prob = 20%
= 4,6941 dB
The smart antenna gains are the same for uplink and downlink. Their are no losses for this type of smart antenna equipment.
Negative values of C/I gains are considered as losses.
53

54
Chapter 2
Radio Propagation
"Path Loss Calculation Prerequisites" on page 57
"List of Default Propagation Models" on page 62
"Okumura-Hata and Cost-Hata Propagation Models" on

page 63
"ITU 529-3 Propagation Model" on page 64
"Standard Propagation Model (SPM)" on page 65
"WLL Propagation Model" on page 75
"ITU-R P.526-5 Propagation Model" on page 76
"Erceg-Greenstein (SUI) Propagation Model" on page 78
"Sakagami Extended Propagation Model" on page 84
"Free Space Loss" on page 86
"Diffraction" on page 86
"Shadow Fading Model" on page 90
"Path Loss Matrices" on page 104
"File Formats" on page 108

56

Chapter 2: Radio Propagation
AT332_TRR_E0
2 Radio Propagation
Path loss calculations are carried out between a transmitter and a receiver using propagation models and other calculations
related to radio wave propagation such as diffraction and shadow fading. Propagation models are mathematical
representations of the average loss in signal strength over distance. Diffraction loss and shadow fading margins are added to
this average loss in order to get more precise path loss values.
Path loss matrices are calculated for each transmitter and their results used in other calculations (coverage predictions, Monte
Carlo simulations, point analysis, etc.). The method of calculation may differ depending on the analysis being performed:
Analysis type
Receiver
position
Calculation
Profile
extraction
Result
Coverage predictions
Centre of each bin

inside the calculation
area
Based on path loss

matrices
Radiala
One value for the bins

surface area
Point analysis (Profile)
Anywhere
Real-time
Systematic
Different values inside

a calculation bin
Point analysis (other)
Anywhere inside the

calculation areas
Based on path loss

matrices
Radiala
One value for the bins

surface area
Monte Carlo
simulations
Mobile coordinates
Based on path loss

matrices
Radiala
One value at the

mobile location
Subscriber lists
Subscriber coordinates
Real-time
Radiala
One value at the

subscriber location
a.
With the Standard Propagation Model, you can choose between radial or systematic.
This chapter describes the various propagation models available in Atoll, and other radio wave propagation phenomena such
as diffraction and shadow fading.
2.1 Path Loss Calculation Prerequisites

2.1.1 Ground Altitude Determination
Atoll determines reception and transmission site altitude from Digital Terrain Model map. The method used to evaluate site
altitude is based on a bilinear interpolation.
DTM files provide altitude value z (in metre) on evenly spaced points. Abscissa and ordinate axes are respectively oriented in
right and downwards directions. Space between points is defined by pixel size P (in metre). Pixel size must be the same in both
directions. The first point given in the file corresponds to the centre of the top-left pixel of the map (northwest point georeferenced by Atoll).
Figure 2.1: Digital Terrain Model

Four points (hence, four altitude values) are necessary to describe a bin; these points are bin vertices. Therefore, a DTM file
that contains N x N bins requires N2 points (altitude values).
57

Figure 2.2: Schematic view of a DTM file
In Atoll, DEM (Digital Elevation Model) is the same as Digital Terrain Model (DTM). In
literature, DEM and DTM do not always have the same meaning. By definition, DEM
refers to the altitude above sea level including ground and clutter, while DTM refers
to the ground altitude above sea level alone.
Let us suppose a site S located inside a bin. Atoll knows the altitudes of four bin vertices, S1, S1, S2 and S2, from the DTM
file (centre of each DTM pixel).
Figure 2.3: Ground Altitude Determination - 1

1. Atoll draws a vertical line through S. This line respectively intersects (S1,S1) and (S2, S2) lines at S1 and S2.

2. Atoll determines the S1 and S2 altitudes using a linear interpolation method.

3. Atoll performs a second linear interpolation to evaluate the S altitude.
58

AT332_TRR_E0
2.1.2 Clutter Determination

Some propagation models need clutter class and clutter height as information at receiver or along a transmitter-receiver
profile.
2.1.2.1 Clutter Classes

Atoll uses clutter classes file to determine the clutter class. The clutter classes map is a grid representing the ground with each
bin assigned a clutter class code corresponding to its clutter type. It is possible to specify an average height for each clutter
class in Atoll.
Clutter class files provide a clutter code per bin. Bin size is defined by pixel size P (in metre). Pixel size must be the same in
both directions. Abscissa and ordinate axes are respectively oriented in right and downwards directions. The first point given
in the file corresponds to the centre of the top-left pixel of the map (northwest point geo-referenced by Atoll.
Figure 2.7: Clutter Classes

Atoll supports a maximum of 255 clutter classes (8 bits/pixel). A clutter classes file file that contains N x N bins requires N2
code values.
2.1.2.2 Clutter Heights

To evaluate the clutter height, Atoll uses clutter heights file if available in the .atl document; clutter height of a site is the
height of the nearest point in the file.
Example: Let us suppose a site S. In the clutter heights file, Atoll reads clutter heights of four points around the site, S1, S1,
S2 and S2. Here, the nearest point to S is S2; therefore Atoll takes the S2 clutter height as clutter height of S.
Figure 2.8: Clutter Height

If you do not have any clutter height file, Atoll takes clutter height information in clutter classes file. In this case, clutter height
is an average height related to a clutter class.
59

2.1.3 Geographic Profile Extraction

Geographic profile extraction is needed in order to calculate diffraction losses. Profiles can be based on DTM only or on DTM
and clutter both, depending on the selected propagation model.
Method 1: Radial Extraction
Atoll draws radials from the site (where transmitter is located) to each calculation bin located along the transmitter
calculation area border. In other words, Atoll determines a geographic profile between site and each bin centre.
Figure 2.9: Radial calculation method

Transmitter location
Radials (Atoll extracts a geographic profile for each radial)
Centres of bins located on the calculation border
Receiver location
Figure 2.10: Site-bin centre profile

Depending on the calculation being carried out, the receiver may be located at the centre of a calculation bin (coverage
predictions) or anywhere within a calculation bin. Atoll uses the profile nearest to the receiver for calculations (the receiver
is assumed to be located on the profile).
60

AT332_TRR_E0
Method 2: Systematic Extraction

Atoll extracts a precise geographic profile between the site and the receiver.
Figure 2.11: Radial calculation method

Transmitter location
Geographic profile
Receiver location
2.1.4 Resolution of the Extracted Profiles

Geographic profile resolution depends on resolution of geographic data used by the propagation model (DTM and/or clutter).
The selected profile resolution does not depend on the geographic layer order.
If the propagation model uses both DTM and clutter heights along the profile, the profile resolution will be the highest
of the two.
Example 1 (Using the Standard Propagation Model)
A DTM map with a 40 m resolution and a clutter heights map with a 20 m resolution are available. The profile
resolution will be 20 m. It means that Atoll will extract geographic information, ground altitude and clutter height,
every 20 m.
To get ground altitude every 20 m, Atoll uses the bilinear interpolation method described in "Ground Altitude
Determination" on page 57. Clutter heights are read from the clutter heights map. Atoll takes the clutter height of the
nearest point every 20 m.
Example 2 (Using the Standard Propagation Model)
A DTM map with a 40 m resolution and a clutter classes map with a 20 m resolution are available. No clutter height
file has been imported in the document. The profile resolution will be 20 m. It means that Atoll will extract geographic
information, ground altitude and clutter height, every 20 m.
Determination" on page 57. Atoll uses the clutter classes map to determine clutter height. Every 20 m, it determines
clutter class and takes associated average height.
If the propagation model uses only DTM along the profile, the profile resolution will be the highest resolution among
the DTM files.
Example (Using the Cost-Hata Propagation Model)
61

DTM maps with 40 m and 25 m resolutions and a clutter map with a 20 m resolution are available. The profile
resolution will be 25 m. It means that Atoll will extract geographic information, only the ground altitude, every 25 m.
The profile resolution does not depend on the geographic layer order in the Geo tab of the Explorer window.
However, the geographic layer order has influence on the usage of the data. For example, when DTM 1 is on the top
of DTM 2, Atoll will use DTM 1 for extracting the profile where DTM a is available and it will use DTM 2 elsewhere.
Determination" on page 57.
Geo Tab of the Explorer Window
> DTM
> DTM 1 (25m)
> DTM 2 (40m)
> Clutter
> Clutter (20m)
2.2 List of Default Propagation Models

Propagation models available in Atoll are listed in the table below along with their main characteristics.
Propagation
model
ITU 370-7
(Vienna 93)
ITU 1546
ITU 526-5
WLL
Frequency band
100-400 MHz
30-3000 MHz
30-10000 MHz
30-10000 MHz
Physical
phenomena
Free space loss

Corrected standard loss
Free space loss + corrections
Free space loss

Diffraction loss
Free space loss

Diffraction loss
Diffraction
calculation
method
Deygout (3 obstacles)
Deygout corrected
(3 obstacles)
Profile based on
DTM
DTM
Clutter
Profile extraction
mode
Radial
Radial
Cell size
Macro cell
Macro cell
Macro cell
Receiver location
Rooftop
Rooftop
Street
Street
Rooftop
Receiver
Fixed
Mobile
Fixed
Fixed
Use
d > 10 km
Low frequencies
Broadcast
1 < d < 1000 km

Land and maritime mobile,
broadcast
Fixed receivers
WLL
Fixed receivers
WLL, Microwave links,
WiMAX
Propagation
model
Standard Propagation
Model
Erceg-Greenstein (SUI)
ITU 529-3
COST-Hata
Okumura-Hata
Frequency band
150-3500 MHz
1900-6000 MHz
300-1500 MHz
150-2000 MHz
Physical
phenomena
L(d, HTxeff, HRxeff, Diff loss,

clutter)
L(d, f, HTx, HRx)

(per environment)
Diffraction loss
L(d, f, HRx)
(per environment)
Diffraction loss
L(d, f, HRx)
(per environment)
Diffraction loss
Diffraction
calculation
method
Epstein-Peterson
(3 obstacles)
Deygout corrected
(3 obstacles)
Millington (1 obstacle)
Deygout
(1 obstacle)
Deygout
(1 obstacle)
Deygout
(1 obstacle)
62

AT332_TRR_E0
Propagation
model
Standard Propagation
Model
Erceg-Greenstein (SUI)
ITU 529-3
COST-Hata
Okumura-Hata
Profile based on
DTM
Clutter
DTM
DTM
DTM
Profile extraction
mode
Radial
Systematic
Radial
Radial
Radial
Cell size
Macro cell
Mini cell
Macro cell
Mini cell
Macro cell
Mini cell
Macro cell
Mini cell
Receiver location
Street
Rooftop
Street
Street
Street
Receiver
Mobile and Fixed
Fixed
Mobile
Mobile
Use
1 < d < 20 km
GSM, UMTS, CDMA2000,
WiMAX, LTE
Urban and suburban areas

100 m < d < 8 km
Fixed WiMAX
1 < d < 100 km

GSM, CDMA2000, LTE
GSM, UMTS, CDMA2000, LTE
2.3 Okumura-Hata and Cost-Hata Propagation Models

2.3.1 Hata Path Loss Formula
Hata formula empirically describes the path loss as a function of frequency, receiver-transmitter distance and antenna heights
for an urban environment. This formula is valid for flat, urban environments and 1.5 metre mobile antenna height.
Path loss (Lu) is calculated (in dB) as follows:
Lu = A 1 + A 2 log f + A 3 log h Tx + B 1 + B 2 log h Tx + B 3 h Tx log d
f is the frequency (MHz).
hTx is the transmitter antenna height above ground (m) (Hb notation is also used in Atoll).
d is the distance between the transmitter and the receiver (km).
The parameters A1, A2, A3, B1, B2, and B3 can be user-defined. Default values are proposed in the table below:
Parameters
Okumura-Hata
f 1500 MHz
Cost-Hata
f > 1500 MHz
A1
69.55
49.30
A2
26.16
33.90
A3
-13.82
-13.82
B1
44.90
44.90
B2
-6.55
-6.55
B3
2.3.2 Corrections to the Hata Path Loss Formula

As described above, the Hata formula is valid for urban environment and a receiver antenna height of 1.5m. For other
environments and mobile antenna heights, corrective formulas must be applied.
For urban areas: L model1 = Lu a h Rx
f 2
For suburban areas: L model1 = Lu a h Rx 2 log ------ 5,4
28
For quasi-open rural areas: L model1 = Lu a h Rx 4,78 log f + 18,33 log f 35,94
For open rural areas: L model1 = Lu a h Rx 4,78 log f + 18,33 log f 40,94
a(hRx) is a correction for a receiver antenna height different from 1.5m.
For rural/small cities: a h Rx = 1,1 log f 0,7 h Rx 1,56 log f 0,8
For large cities: a h Rx = 3,2 log 11,75h Rx 4,97
63

When receiver antenna height equals 1.5m, a(hRx) is close to 0 dB regardless of frequency.
2.3.3 Calculations in Atoll

Hata models take into account topo map (DTM) between transmitter and receiver and morpho map (clutter) at the receiver.
1st step: For each calculation bin, Atoll determines the clutter bin on which the receiver is located. This clutter bin corresponds
to a clutter class. Then, it uses the Hata formula assigned to this clutter class to evaluate L model1 .
2nd step: This step depends on whether the Add diffraction loss option is checked.
If the Add diffraction loss option is unchecked, Atoll stops calculations.

L model = L model1
If the Add diffraction loss option is selected, Atoll proceeds as follows:

a. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation mode.
b. It determines the largest obstacle along the profile in accordance with the Deygout method and evaluates losses
due to diffraction L model2 .
L model = L model1 + L model2
2.4 ITU 529-3 Propagation Model

2.4.1 ITU 529-3 Path Loss Formula
The ITU 529.3 model is a Hata-based model. For this reason, its formula empirically describes the path loss as a function of
frequency, receiver-transmitter distance and antenna heights for a urban environment. This formula is valid for flat, urban
environments and 1.5 metre mobile antenna height.
The standard ITU 529-3 formula, for a receiver located on a urban environment, is given by:
E = 69,82 6,16 log f + 13,82 log h Tx 44,9 6,55 log h Tx log d
where:
E is the field strength for 1 kW ERP
f is the frequency (MHz).
h Tx is the transmitter antenna height above ground (m) (Hb notation is also used in Atoll)
h Rx is the receiver antenna height above ground (m)
d is the distance between the transmitter and the receiver (km)
b is the distance correction
The domain of validity of such is formula is:
Frequency range: 300-1500 MHz

Base Station height: 30-200 m
Mobile height: 1-10 m
Distance range: 1-100 km
Since Atoll needs the path loss (Lu) formula, a conversion has to be made. One can find the following conversion formula:
Lu = 139,37 + 20 log f E
which gives the following path loss formula for the ITU 529-3 model:
Lu = 69,55 + 26,16 log f 13,82 log h Tx + 44,9 6,55 log h Tx log d
2.4.2 Corrections to the ITU 529-3 Path Loss Formula

Environment Correction
As described above, the Hata formula is valid for urban environment. For other environments and mobile antenna heights,
corrective formulas must be applied.
64

AT332_TRR_E0
L model1 = Lu a h Rx for large city and urban environments

f 2
L model1 = Lu a h Rx 2 log ------ 5,4 for suburban area
28
2
L model1 = Lu a h Rx 4,78 log f + 18,33 log f 40,94 for rural area

Area Size Correction
In the formulas above, a h Rx is the environment correction and is defined according to the area size.
For rural/small cities: a h Rx = 1,1 log f 0,7 h Rx 1,56 log f 0,8
For large cities: a h Rx = 3,2 log 11,75h Rx 4,97
Distance Correction
The distance correction refers to the term b above.
d<20 km: b = 1
h Tx
4
3
d 0,8
d>20 km: b = 1 + 0,14 + 1,87 10 f + 1,07 10 h' Tx log ------ with h' Tx = ---------------------------------------- 20
6 2
1 + 7 10 h Tx

Hata-based models take into account topo map (DTM) between transmitter and receiver and morpho map (clutter) at the
receiver.
1st step: For each calculation bin, Atoll determines the clutter bin on which the receiver is located. This clutter bin corresponds
to a clutter class. Then, it uses the ITU 529-3 formula assigned to this clutter class to evaluate L model1 .
2nd step: This step depends on whether the Add diffraction loss option is checked.
If the Add diffraction loss option is unchecked, Atoll stops calculations.

L model = L model1

a. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation mode.
due to diffraction L model2 .
L model = L model1 + L model2
2.5 Standard Propagation Model (SPM)

2.5.1 SPM Path Loss Formula
SPM is based on the following formula:
L model = K 1 + K 2 log d + K 3 log H Txeff + K 4 DiffractionLoss + K 5 log d log H Txeff +
K 6 H Rxeff + K 7 log H Rxeff + K clutter f clutter
with,
K1: constant offset (dB).
K2: multiplying factor for log(d).
d: distance between the receiver and the transmitter (m).
K3: multiplying factor for log(HTxeff).
HTxeff: effective height of the transmitter antenna (m).
K4: multiplying factor for diffraction calculation. K4 has to be a positive number.
65

Diffraction loss: loss due to diffraction over an obstructed path (dB).

K5: multiplying factor for log d log H Txeff
K6: multiplying factor for H Rxeff
K7: multiplying factor for log H Rxeff .
H Rxeff : effective mobile antenna height (m).
Kclutter: multiplying factor for f(clutter).
f(clutter): average of weighted losses due to clutter.

2.5.2.1 Visibility and Distance Between Transmitter and Receiver
For each calculation bin, Atoll determines:
The distance between the transmitter and the receiver.
If the distance Tx-Rx is less than the maximum user-defined distance (break distance), the receiver is considered to be near
the transmitter. Atoll will use the set of values marked Near transmitter.
If the distance Tx-Rx is greater than the maximum distance, receiver is considered far from transmitter. Atoll will use the set
of values Far from transmitter.
Whether the receiver is in the transmitter line of sight or not.
If the receiver is in the transmitter line of sight, Atoll will take into account the set of values (K1,K2)LOS. The LOS is defined by
no obstruction along the direct ray between the transmitter and the receiver.
If the receiver is not in the transmitter line of sight, Atoll will use the set of values (K1,K2)NLOS.
2.5.2.2 Effective Transmitter Antenna Height

Effective transmitter antenna height (HTxeff) may be calculated with six different methods.
Height Above Ground
The transmitter antenna height is above the ground (HTx in m).
HTxeff = HTx
Height Above Average Profile
The transmitter antenna height is determined relative to an average ground height calculated along the profile between a
transmitter and a receiver. The profile length depends on distance min and distance max values and is limited by the
transmitter and receiver locations. Distance min and Distance max are minimum and maximum distances from the transmitter
respectively.
H Txeff = H Tx + H 0Tx H 0
where,
H 0Tx is the ground height (ground elevation) above sea level at transmitter (m).
H 0 is the average ground height above sea level along the profile (m).
If the profile is not located between the transmitter and the receiver, HTxeff equals HTx
only.
Slope at Receiver Between 0 and Minimum Distance

The transmitter antenna height is calculated using the ground slope at receiver.
H Txeff = H Tx + H 0Tx H 0Rx + K d
where,
66

AT332_TRR_E0
H 0Rx is the ground height (ground elevation) above sea level at receiver (m).
K is the ground slope calculated over a user-defined distance (Distance min). In this case, Distance min is a distance from
receiver.
If H Txeff 20m then, Atoll uses 20m in calculations.
If H Txeff 200m then, Atoll takes 200m.
Spot Ht
If H 0Tx H 0Rx then, H Txeff = H Tx + H 0Tx H 0Rx
If H 0Tx H 0Rx then, H Txeff = H Tx
Absolute Spot Ht
H Txeff = H Tx + H 0Tx H 0Rx
Distance min and distance max are set to 3000 and 15000 m according to ITU
recommendations (low frequency broadcast f < 500 Mhz) and to 0 and 15000 m
according Okumura recommendations (high frequency mobile telephony).
These values are only used in the two last methods and have different meanings according to the method.
Enhanced Slope at Receiver
Atoll offers a new method called Enhanced slope at receiver to evaluate the effective transmitter antenna height.
Figure 2.12: Enhanced Slope at Receiver

Let x-axis and y-axis respectively represent positions and heights. We assume that x-axis is oriented from the transmitter
(origin) towards the receiver.
This calculation is achieved in several steps:
1. Atoll determines line of sight between transmitter and receiver.
The LOS line equation is:
H 0Tx + H Tx H 0Rx + H Rx
- Res i
Los i = H 0Tx + H Tx ----------------------------------------------------------------------d
where,
H Rx is the receiver antenna height above the ground (m).
i is the point index.
Res is the profile resolution (distance between two points).
2. Atoll extracts the transmitter-receiver terrain profile.
67

3. Hills and mountains are already taken into account in diffraction calculations. Therefore, in order for them not to
unfavourably influence the regression line calculation, Atoll filters the terrain profile.
Atoll calculates two filtered terrain profiles; one established from the transmitter and another from the receiver. It
determines filtered height of every profile point. Profile points are evenly spaced on the basis of profile resolution. To
determine filtered terrain height at a point, Atoll evaluates ground slope between two points and compares it with a
threshold set to 0.05; where three cases are possible.
Some notations defined hereafter are used in next part.
H filt is the filtered height.
H orig is the original height. Original terrain height is determined from extracted ground profile.
Filter starting from transmitter

Let us assume that H filt Tx Tx = H orig Tx
For each point, we have three different cases:
i.
H orig i H orig i 1
- 0,05 ,
If H orig i H orig i 1 and --------------------------------------------------Res
Then, H filt Tx i = H filt Tx i 1 + H orig i H orig i 1
H orig i H orig i 1
- 0,05
ii. If H orig i H orig i 1 and --------------------------------------------------Res
Then, H filt Tx i = H filt Tx i 1
iii. If H orig i H orig i 1
Then, H filt Tx i = H filt Tx i 1
If H filt i H orig i additionally
Then, H filt Tx i = H orig i
Filter starting from receiver
Let us assume that H filt Rx = H orig Rx

For each point, we have three different cases:
i.
H orig i H orig i + 1
- 0,05 ,
If H orig i H orig i + 1 and --------------------------------------------------Res
Then, H filt Rx i = H filt Rx i + 1 + H orig i H orig i + 1
H orig i H orig i + 1
- 0,05
ii. If H orig i H orig i + 1 and --------------------------------------------------Res
Then, H filt Rx i = H filt Rx i + 1
iii. 3rd case: If H orig i H orig i + 1
Then, H filt Rx i = H filt Rx i + 1
If H filt i H orig i additionally
Then, H filt Rx i = H orig i
Then, for every point of profile, Atoll compares the two filtered heights and chooses the higher one.
H filt i = max H filt Tx i H filt Rx i
4. Atoll determines the influence area, R. It corresponds to the distance from receiver at which the original terrain profile
plus 30 metres intersects the LOS line for the first time (when beginning from transmitter).
The influence area must satisfy additional conditions:
68
R 3000m

AT332_TRR_E0
R 0,01 d
R must contain at least three bins.
When several influence areas are possible, Atoll chooses the highest one.
If d < 3000m, R = d.
5. Atoll performs a linear regression on the filtered profile within R in order to determine a regression line.
The regression line equation is:
y = ax + b
d i dm Hfilt i Hm
i
and b = H m ad m
a = --------------------------------------------------------------------2
d i dm
where,
1
H m = --n
Hfilt i
i
i is the point index. Only points within R are taken into account.
dm = d R
--2
d(i) is the distance between i and the transmitter (m).
Then, Atoll extends the regression line to the transmitter location. Therefore, its equation is:
regr i = a i Res + b
6. Then, Atoll calculates effective transmitter antenna height, H Txeff (m).
H 0Tx + H Tx b
H Txeff = --------------------------------2
1+a
If HTxeff is less than 20m, Atoll recalculates it with a new influence area, which begins at transmitter.
In case H Txeff 1000m , 1000m will be used in calculations.
If H Txeff is still less than 20m, an additional correction is taken into account (7th step).
7. If H Txeff is still less than 20m (even negative), Atoll evaluates path loss using H Txeff = 20m and applies a correction
factor.
Therefore, if H Txeff 20m ,
L model = L model H Txeff = 20m d f + K lowant
20 1 H Txeff 20
d
where, K lowant = -------- 0,3 H Txeff 20 ------------------------------------------------------------------------5
d
d
9,63 + ----------10
- 6,93 + ------------
1000
1000
2.5.2.3 Effective Receiver Antenna Height

H Rxeff = H Rx + H 0Rx H 0Tx
where,
H Rx is the receiver antenna height above the ground (m).
H 0Rx is the ground height (ground elevation) above sea level at the receiver (m).
H 0Tx is the ground height (ground elevation) above sea level at the transmitter (m).
69

The calculation of effective antenna heights ( H Rxeff and H Txeff ) is based on extracted
DTM profiles. They are not properly performed if you have not imported heights (DTM
file) beforehand.
2.5.2.4 Correction for Hilly Regions in Case of LOS

An optional corrective term enables Atoll to correct path loss for hilly regions when the transmitter and the receiver are in
Line-of-sight.
Therefore, if the receiver is in the transmitter line of sight and the Hilly terrain correction option is active, we have:
L model = K 1 LOS + K 2 LOS log d + K 3 log H Txeff + K 5 log H Txeff log d + K 6 H Rx + K clutter f clutter + K hill LOS
When the transmitter and the receiver are not in line of sight, the path loss formula is:
L model = K 1 NLOS + K 2 NLOS log d + K 3 log H Txeff + K 4 Diffraction + K 5 log H Txeff log d + K 6 H Rx + K clutter f clutter
K hill LOS is determined in three steps. Influence area, R, and regression line are supposed available.
1st step: For every profile point within influence area, Atoll calculates height deviation between the original terrain profile and
regression line. Then, it sorts points according to the deviation and draws two lines (parallel to the regression line), one which
is exceeded by 10% of the profile points and the other one by 90%.
2nd step: Atoll evaluates the terrain roughness, h; it is the distance between the two lines.
3rd step: Atoll calculates K hill LOS .
We have K hill LOS = K h + K hf
If 0 h 20m , K h = 0
2
Else K h = 7,73 log h 15,29 log h + 6,746

If 0 h 10m , K hf = 2 0,1924 H 0Rx + H Rx regr i Rx
H 0Rx + H Rx regr i Rx
2
Else K hf = 2 1,616 log h + 14,75 log h 11,21 ----------------------------------------------------h
iRx is the point index at receiver.
2.5.2.5 Diffraction
Four methods are available to calculate diffraction loss over the transmitter-receiver profile.
Along the transmitter-receiver profile, you may consider:
Either ground altitude and clutter height (Consider heights in diffraction option),
In this case, Atoll uses clutter height information from clutter heights file if available in the .atl document. Otherwise,
it considers average clutter height specified for each clutter class in the clutter classes file description.
Or only ground altitude.
2.5.2.6 Losses due to Clutter

n
Atoll calculates f(clutter) over a maximum distance from receiver: f clutter =
Li wi
i=1
where,
L: loss due to clutter defined in the Clutter tab by the user (in dB).
w: weight determined through the weighting function.
n: number of points taken into account over the profile. Points are evenly spaced depending on the profile resolution.
Four weighting functions are available:
70
Uniform weighting function: w i = --1n

AT332_TRR_E0
di
Triangular weighting function: w i = -----------n
dj
j
d i = D d' i , where di is the distance between the receiver and the ith point and D is the maximum distance defined.
d
log ----i + 1
D
Logarithmic weighting function: w i = -----------------------------------n

d
log ----j + 1
D
1
di
---D
e 1
Exponential weighting function: w i = -----------------------n
dj
---D
j=1
The chart below shows the weight variation with the distance for each weighting function.
Figure 2.13: Losses due to Clutter
2.5.2.7 Recommendations
Beware that the clutter influence may be taken into account in two terms, Diffraction loss and f(clutter) at the same time. To
avoid this, we advise:
1. Not to consider clutter heights to evaluate diffraction loss over the transmitter-receiver profile if you specify losses
per clutter class.
This approach is recommended if the clutter height information is statistical (clutter roughly defined, no altitude).
Or
2. Not to define any loss per clutter class if you take clutter heights into account in the diffraction loss.
In this case, f(clutter)=0. Losses due to clutter are only taken into account in the computed Diffraction loss term.
This approach is recommended if the clutter height information is either semi-deterministic (clutter roughly defined,
altitude defined with an average height per clutter class) or deterministic (clutter sharply defined, altitude defined
with an average height per clutter class or - even better - via a clutter height file).
In case of semi-deterministic clutter information, specify receiver clearance (m) per clutter class. Both ground altitude and
clutter height are considered along the whole transmitter-receiver profile except over a specific distance around the receiver
(clearance), where Atoll proceeds as if there was only the DTM map. The clearance information is used to model streets.
71

Figure 2.14: Tx-Rx profile

In the above figure, the ground altitude and clutter height (in this case, average height specified for each clutter class in the
clutter classes map description) are taken into account along the profile.
Clearance definition is not necessary in case of deterministic clutter height information. Clutter height information is accurate
enough to be used directly without additional information such as clearance. Two cases can be considered:
1. If the receiver is in the street (clutter height lower than receiver height), Atoll calculates the path loss by considering
potentially some diffraction loss at reception.
2. If the receiver is supposed to be inside a building (clutter height higher than receiver height), Atoll does not consider
any difraction (and clearance) from the building but takes into account the indoor loss as an additional penetration
loss.
To consider indoor losses in building only when using a deterministic clutter map
(clutter height map), the 'Indoor Coverage' box must not be checked in predictions
unless this loss will be counted twice inside buildings (on the entire reception clutter
class and not only inside the building).
Even with no clearance, the clutter height (extracted either from clutter class or
clutter height folders) is never considered at the last profile point.
2.5.3 Automatic Calibration

The goal of this tool is to calibrate parameters and methods of the SPM formula in a simple and reproducible way. Calibration
is based on imported CW measurement data. It is the process of limiting the difference between predicted and measured
values. For a complete description of the calibration procedure (including the very important prerequisite filtering work on
the CW measurement points), please refer to the User Manual and the SPM Calibration Guide.
The following SPM formula parameters can be estimated:
K1, K2, K3, K4, K5, K6 and K7

Losses per clutter class (Kclutter must be user-defined)
Effective antenna height method
Diffraction method
Automatic model calibration provides a mathematical solution. The relevance of this mathematical solution with a physical
and realistic solution must be determined before committing these results.
You must keep in mind that the model calibration and its result (standard deviation and root mean square) strongly depend
on the CW measurement samples you use. A calibrated model must restore the behaviour of CW measurements depending
on their configuration on a large scale, and not just totally coincide with a few number of CW measurements. The calibrated
model has to give correct results for every new CW measurement point in the same geographical zone, without having been
calibrated on these new CW measurements.
2.5.3.1 General Algorithm

Propagation model calibration is a special case of the more general Least-Square problems, i.e. given a real m x n matrix A,
and a real m-vector b, find a real n-vector x0 that minimises the Euclidean length of Ax - b.
Here,
m is the number of measurement points,
n is the number of parameters to calibrate,
72

AT332_TRR_E0
A is the values of parameter associated variables (log(d), log(heff), etc.) at each measurement point, and
b is the vector of measurement values.
The vector x0 is the set of parameters found at the end of the calibration.
The theoretical mathematical solution of this problem was found by Gauss (around 1830). Further enhancements to the
original method were proposed in the 60's in order to solve the numerical instability problem.
In 1974, Lawson & Hanson [2] proposed a theoretical solution of the least-square problem with general linear inequality
constraints on the vector x0. Atoll implementation is based on this method, which is explained in detail in [1].
References:
[1] Bjrck A. Numerical Methods for Least Square Problems, SIAM, 1996.
[2] Lawson C.L., Hanson R.J. Solving Least Squares Problems, SIAM, 1974.
2.5.3.2 Sample Values for SPM Path Loss Formula Parameters

The following tables list some sample orders of magnitudes for the different parameters composing the Standard Propagation
Model formula.
Minimum
Typical
Maximum
K1
Variable
Variable
Variable
K2
20
44.9
70
K3
-20
5.83
20
K4
0.5
0.8
K5
-10
-6.55
K6
-1
K7
-10
It is recommended to set K6 to 0, and use K7 instead of K6. K6 is a multiplicative coefficient to a value in dB, which means that
slight variations in K6 have considerable impact on the path loss.
K1 depends on the frequency and the technology. Here are some sample values:
Project type
Frequency (MHz)
K1
GSM 900
935
12.5
GSM 1800
1805
22
GSM 1900
1930
23
UMTS
2045a
23.8
1xRTT
1900
23
2300
25.6
2500
26.8
2700
27.9
3300
30.9
3500
31.7
WiMAX
a.
2045 MHz = (2140 + 1950)/2. It is the average of the downlink

and uplink centre frequencies of the band.
The above K1 values for WiMAX are extrapolated estimates for different frequency ranges. It is highly recommended to
calibrate the SPM using measurement data collected on the field for WiMAX networks before using the SPM for predictions.
All K paramaters can be defined by the automatic calibration wizard. Since Kclutter is a constant, its value is strongly dependant
on the values given to the losses per clutter classes. From experience, typical losses (in dB) per clutter class are:
Dense urban
From 4 to 5
Woodland
From 2 to 3
Urban
73

Suburban
From -5 to -3
Industrial
From -5 to -3
Open in urban
From -6 to -4
Open
From -12 to -10
Water
From -14 to -12
These values have to be entered only when considering statistical clutter class maps only.
The Standard Propagation Model is derived from the Hata formulae, valid for urban
environments. The above values are normalized for urban clutter types (0 dB for urban
clutter class). Positive values correspond to more dense clutter classes and negative
values to less dense clutter classes.
2.5.4 Unmasked Path Loss Calculation

You can use the SPM to calculate unmasked path losses. Unmasked path losses are calculated by not taking into account the
transmitter antenna patterns, i.e., the attenuation due to the transmitter antenna pattern is not included. Such path losses
are useful when using path loss matrices calculated by Atoll with automatic optimisation tools.
The instance of the SPM available by default, under the Propagation Models folder in the Modules tab, has the following
characteristics:
Signature:
Type:
{D5701837-B081-11D4-931D-00C04FA05664}
Atoll.StdPropagModel.1
You can access these parameters in the Propagation Models table by double-clicking the Propagation Models folder in the
Modules tab.
To make the SPM calculate path losses excluding the antenna pattern attenuation, you have to change the type of the SPM to:
Type:
Atoll.StdPropagModelUnmasked.1
However, changing the type only does not invalidate the already calculated path loss matrices, because the signature of the
propagation model is still the same. If you want Atoll to recognize that the SPM has changed, and to invalidate the path loss
matrices calculated with this model, you have to change the signature of the model as well. The default signature for the SPM
that calculates unmasked path loss matrices is:
Signature:
{EEE060E5-255C-4C1F-B36C-A80D3D972583}
The above signature is a default signature. Atoll automatically creates different signatures for different instances of the same
propagation model. Therefore, it is possible to create different instances of the SPM, with different parameter settings, and
create unmasked versions of these instances.
You can change the signature and type of the original instance of the SPM, but it is recommended to make a copy of the SPM
in order not to lose the original SPM parameters. So, you will be able to keep different versions of the SPM, those that
calculate path losses with antenna pattern attenuation, and others that calculate path losses without it.
The usual process flow of an ACP working on an Atoll document through the API would be to:
1. Backup the storage directory of path loss matrices.
2. Set a different storage directory for calculating and storing unmasked path loss matrices.
3. Select the SPM used, backup its signature, and change its signature and type as shown above.
4. Perform optimisation using the path loss matrices calculated by the unmasked version of the SPM.
5. Restore the type and the signature of the SPM.
6. Reset the path loss storage directory to the original one.
74

AT332_TRR_E0
It is not possible to calibrate the unmasked version of the SPM using measurement
data.
Using the SPM, you can also calculate the angles of incidence by creating a new
instance of the SPM with the following characteristics:
Type:
Signature:
Atoll.StdPropagModelIncidence.1
{659F0B9E-2810-4e59-9F0D-DA9E78E1E64B}
The "masked" version of the algorithm has not been changed. It still takes into
account Atoll.ini options. However, the "unmasked" version does not take Atoll.ini
options into account.
Its highly recommended to use one method (Atoll.ini options) or the other one (new
identifier & signature) but not to combine both.
2.6 WLL Propagation Model

2.6.1 WLL Path Loss Formula
L model = L FS + F Diff L Diff
Where L FS is the free space loss calculated using the formula entered in the model properties, L Diff is the diffraction loss
calculated using the 3-obstacle Deygout method, and F Diff is the diffraction multiplying factor defined in the model
properties.

Free Space Loss
For free space loss calculation, see "Free Space Loss" on page 86.
Diffraction
Atoll calculates diffraction loss along the transmitter-receiver profile built from DTM and clutter maps. Therefore, losses due
to clutter are taken into account in diffraction losses. Atoll takes clutter height information from the clutter heights file if
available in the .atl document. Otherwise, it considers average clutter height specified for each clutter class in the clutter
classes file description.
The Deygout construction (considering 3 obstacles) is used. This method is described under "Diffraction" on page 86. The final
diffraction losses are determined by multiplying the diffraction losses calculated using the Deygout method by the Diffraction
multiplying factor defined in the model properties.
Receiver Clearance
Define receiver clearance (m) per clutter class when clutter height information is either statistical or semideterministic. Both ground altitude and clutter height are considered along the whole profile except over a specific
distance around the receiver (clearance), where Atoll proceeds as if there was only the DTM map (see SPM part). Atoll
uses the clearance information to model streets.
If the clutter is deterministic, do not define any receiver clearance (m) per clutter class. In this case, clutter height
information is accurate enough to be used directly without additional information such as clearance (Atoll can locate
streets).
Receiver Height
Entering receiver height per clutter class enables Atoll to consider the fact that receivers are fixed and located on the
roofs.
Visibility
If the option Line of sight only is not selected, Atoll computes Lmodel on each calculation bin using the formula defined
above. When selecting the option Line of sight only, Atoll checks for each calculation bin if the Diffraction loss (as
defined in the Diffraction loss: Deygout part) calculated along profile equals 0.
In this case, receiver is considered in line of sight and Atoll computes Lmodel on each calculation bin using the
formula defined above.
Otherwise, Atoll considers that Lmodel tends to infinity.
75

2.7 ITU-R P.526-5 Propagation Model

L model = L FS + L Diff
Where L FS is the free space loss calculated using the formula entered in the model properties and L Diff is the diffraction loss
calculated using the 3-obstacle Deygout method.

Free Space Loss
Diffraction
Atoll calculates diffraction loss along the transmitter-receiver profile is built from the DTM map. The Deygout construction
(considering 3 obstacles), with or without correction, is used. These methods are described under "Diffraction" on page 86.

If d<1 km, L model = L FS
If d>1000 km, L model = 1000
If 1<d<1000 km, L model = max L FS CorrectedS tan dardLoss
d is the distance between the transmitter and the receiver (km).

Free Space Loss
Corrected Standard Loss
This formula is given for a 60 dBm (1kW) transmitter power.
CorrectedS tan dardLoss = 60 C n A H
Rxeff
A cl 108,75 + 31,54 20 log f
where,
Cn is the field strength received in dBV/m,
AH
Rxeff
is a correction factor for effective receiver antenna height (dB),
Acl is the correction for terrain clearance angle (dB),

f is the frequency in MHz.
Cn Calculation
The Cn value is determined from charts Cn=f(d, HTxeff).
In the following part, let us assume that Cn=En(d,HTxeff) (where En(d,HTxeff) is the field received in dBV/m) is read from
charts for a distance, d (in km), and an effective transmitter antenna height, HTxeff (in m).
First of all, Atoll evaluates the effective transmitter antenna height, H Txeff , as follows:
If 0 d 3km , H Txeff = H 0Tx + H Tx H 0Rx
76

AT332_TRR_E0
If 3 d 15km , H Txeff = H 0Tx + H Tx H 0 3 ;d

If 15 d , H Txeff = H 0Tx + H Tx H 0 3 ;15
where,
H Tx is the transmitter antenna height above the ground (m).
H 0Tx is the ground height (ground elevation) above sea level at the transmitter (m).
H 0 3 ;d is the average ground height (m) above sea level for the profile between a point 3 km from transmitter and
the receiver (located at d km from transmitter).
H 0 3 ;15 is the average ground height (m) above sea level for the profile between a point 3 km and another 15 km
from transmitter.
Then, depending on d and HTxeff, Atoll determines Cn using bilinear interpolation as follows.
If 37.5 HTxeff 1200, Cn= En(d,HTxeff)
Otherwise, Atoll considers d horizon = 4,1 H Txeff (d is stated in km)
Therefore,
If HTxeff < 37.5
If d d horizon , we have C n = E n d + 25 d horizon 37,5
Else Cn=En(d, 37.5) En(dhorizon, 37.5) + En(25, 37.5)
If HTxeff > 1200
If d d horizon , we have C n = E n d + 142 d horizon 1200
Else Cn=En(d, 1200) En(dhorizon, 1200) + En(142, 1200)
AHRxeff Calculation
AH
Rxeff
H Rx
c
= --- 20 log ------ 10
6
where,
HRx is the user-defined receiver height,
c is the height gain factor.
c values are provided in the recommendation 370-7; for example, c=4 in a rural case.
Acl Calculation
If f
300 MHz, A cl = 8,1 6,9 + 20 log 0,1 + 1 + 0,1

2
Otherwise, A cl = 14,9 6,9 + 20 log 0,1 + 1 + 0,1

f
With = 4000 --------300
where,
is the clearance angle (in radians) determined according to the recommendation 370-7 (figure 19),
f is the frequency stated in MHz.
77

2.9 Erceg-Greenstein (SUI) Propagation Model

Erceg-Greenstein propagation model is a statistical path loss model derived from experimental data collected at 1.9 GHz in 95
macrocells. The model is for suburban areas, and it distinguishes between different terrain categories called the Stanford
University Interim Terrain Models. This propagation model is well suited for distances and base station antenna heights that
are not well-covered by other models. The path loss model applies to base antenna heights from 10 to 80 m, base-to-terminal
distances from 0.1 to 8 km, and three distinct terrain categories.
The basic path loss equation of the Erceg-Greenstein propagation model is:
d
PL = A + 10 a H BS Log 10 -----
d 0
4d 0
Where A = 20 Log 10 ------------ . This is a fixed quantity which depends upon the frequency of operation. d is the distance
between the base station antenna and the receiver terminal and d0 is a fixed reference distance (100 m). a(HBS) is the
correction factor for base station antenna heights, HBS:
ca H BS = a b H BS + ------H BS
Where 10 m HBS 80 m , and a, b, and c are correction coefficients which depend on the SUI terrain type.
The Erceg-Greenstein propagation model is further developed through the correction factors introduced by the Stanford
University Interim model. The standards proposed by the IEEE working group 802.16 include channel models developed by
Stanford University. The basic path loss equation with correction factors is presented below:
d
PL = A + 10 a H BS Log 10 ----- + a f a H R
d 0
f
Where a(f) is the correction factor for the operating frequency, a f = 6 Log 10 ------------ , with f being the operating
2000
HR
frequency in MHz. a(HR) is the correction factor for the receiver antenna height, a H R = X Log 10 ------ , where d depends
2
on the terrain type.
a(HR) = 0 for HR = 2 m.
References:
[1] V. Erceg et. al, An empirically based path loss model for wireless channels in
suburban environments, IEEE J. Select Areas Commun., vol. 17, no. 7, July 1999,
pp. 1205-1211.
[2] Abhayawardhana, V.S.; Wassell, I.J.; Crosby, D.; Sellars, M.P.; Brown, M.G.;
"Comparison of empirical propagation path loss models for fixed wireless access
systems," Vehicular Technology Conference, 2005. IEEE 61st Volume 1, 30 May-1
June 2005 Page(s):73 - 77 Vol. 1
2.9.1 SUI Terrain Types

The SUI models are divided into three types of terrains2, namely A, B and C.
Type A is associated with maximum path loss and is appropriate for hilly terrain with moderate to heavy tree
densities.
Type B is characterised with either mostly flat terrains with moderate to heavy tree densities or hilly terrains with
light tree densities.
Type C is associated with minimum path loss and applies to flat terrain with light tree densities.
The constants used for a, b, and c are given in the table below.
2.
78
Model Parameter
Terrain A
Terrain B
Terrain C
4.6
4.0
3.6
b (m-1)
0.0075
0.0065
0.005
The word terrain is used in the original definition of the model rather than environment. Hence it is used
interchangeably with environment in this description.

AT332_TRR_E0
Model Parameter
Terrain A
Terrain B
Terrain C
c (m)
12.6
17.1
20
10.8
10.8
20
2.9.2 Erceg-Greenstein (SUI) Path Loss Formula

The Erceg-Greenstein (SUI) propagation model formula can be simplified from the following equation:
4d
d
PL = 20 Log 10 -----------0- + 10 a H BS Log 10 ----- + a f a H R
d 0

(1)
to the equation below:

PL = 7,366 + 26 Log 10 f + 10 a H BS 1 + Log 10 d a H R
(2)
Where,
f is the operating frequency in MHz

d is the distance from the transmitter to the received in m in equation (1) and in km in equation (2)
HBS is the transmitter height in m
HR is the receiver height in m
The above equation is divided into two parts in Atoll:

PL = Lu a H R
Where,
Lu = 7,366 + 26 Log 10 f + 10 a H BS 1 + Log 10 d
The above path loss formulas are valid for d > d0, i.e. d > 100 m. For d < 100 m, the path loss has been restricted to the free
space path loss with correction factors for operating frequency and receiver height:
4d
4d
PL = 20 Log 10 ------------------ + a f a H R instead of PL = 20 Log 10 ------------------
Where a(f) and a(Hr) have the same definition as given above. Simplifying the above equation, we get,
PL = 12,634 + 26 Log 10 f + 20 Log 10 d a H R , or Lu = 12,634 + 26 Log 10 f + 20 Log 10 d
The above equation is not user-modifiable in Atoll except for the coefficient of Log 10 f , i.e. 26. Atoll uses the same
coefficient as the one you enter for Log 10 f in Atoll for the case d > d0.
You can get the same resulting equation by setting a(hBS) = 2.

The Erceg-Greenstein (SUI) propagation model takes DTM into account between the transmitter and the receiver, and it can
also take clutter into account at the receiver location.
1st step: For each pixel in the calculation radius, Atoll determines the clutter bin on which the receiver is located. This clutter
bin corresponds to a clutter class. Atoll uses the Erceg-Greenstein (SUI) path loss formula assigned to this clutter class to
evaluate path loss.
2nd step: This step depends on whether the Add diffraction loss option is selected or not.
If the Add diffraction loss option is not selected, 1st step gives the final path loss result.
a. It extracts a geographic profile between the transmitter and the receiver using the radial calculation method.
due to diffraction L Diffraction . For more information on the Deygout method, see "3 Knife-edge Deygout Method"
on page 87.
The final path loss is the sum of the path loss determined in 1st step and L Diffraction .
79

Shadow fading is computed in Atoll independent of the propagation model. For more information on the shadow fading
calculation, see "Shadow Fading Model" on page 90.

This propagation model is based on the P.1546-2 recommendations of the ITU-R. These recommendations extend the P.3707 recommendations, and are suited for operating frequencies from 30 to 3000 MHz. The path loss is calculated by this
propagation model with the help of graphs available in the recommendations. The graphs provided in the recommendations
represent field (or signal) strength, given in db V m , as a function of distance for:
Nominal frequencies, f n : 100, 600, and 1000 MHz

The graphs provided for 100 MHz are applicable to frequencies from 30 to 300 MHz, those for 600 MHz are applicable
to frequencies from 300 to 1000 MHz, and the graphs for 1000 MHz are applicable to frequencies from 1000 to
3000 MHz. The method for interpolation is described in the recommendations (Annex 5, 6).
Transmitter antenna heights, h 1 : 10, 20, 37.5, 75, 150, 300, 600, and 1200 m
For any values of h 1 from 10 to 3000 m, an interpolation or extrapolation from the appropriate two curves is used, as
described in the recommendations (Annex 5, 4.1). For h 1 below 10 m, the extrapolation to be applied is given in
Annex 5, 4.2. It is possible for the value of h 1 to be negative, in which case the method is given in Annex 5, 4.3.
Time variability, t : 1, 10, and 50 %

The propagation curves represent the field strength values exceeded for 1, 10 and 50 % of time.
Receiver antenna height, h 2 : 10 m

For land paths, the graphs represent field strength values for a receiver antenna height above ground, equal to the
representative height of the clutter around the receiver. The minimum value of the representative height of clutter is
10 m. For sea paths, the graphs represent field strength values for a receiver antenna height of 10 m.
For other values of receiver antenna height, a correction is applied according to the environment of the receiver. The
method for calculating this correction is given in Annex 5, 9.
These recommendations are not valid for transmitter-receiver distances less than 1 km or greater than 1000 km. Therefore in
Atoll, the path loss between a transmitter and a receiver over less than 1 km is the same as the path loss over 1 km. Similarly,
the path loss between a transmitter and a receiver over more than 1000 km is the same as the path loss over 1000 km.
Moreover, these recommendations are not valid for transmitter antenna heights less than the average clutter height
surrounding the transmitter.
The cold sea graphs are used for calculations over warm and cold sea both.
The mixture of land and sea paths is not supported by Atoll.

The input to the propagation model are the transmission frequency, transmitter and receiver heights, the distance between
the transmitter and the receiver, the precentage of time the field strength values are exceeded, the type of environment (i.e.,
land or sea), and the clutter at the receiver location.
In the following calculations, f is the transmission frequency, d is the transmitter-receiver distance, and t is the percentage
of time for which the path loss has to be calculated.
The following calculations are performed in Atoll to calculate the path loss using this propagation model.
2.10.1.1 Step 1: Determination of Graphs to be Used

First of all, the upper and lower nominal frequencies are determined for any given transmission frequency. The upper and
lower nominal frequencies are the nominal frequencies (100, 600, and 2000 MHz) between which the transmission frequency
is located, i.e., f n1 f f n2 .
Once f n1 and f n1 are known, along with the information about the percentage of time t and the type of path (land or sea),
the sets of graphs which will be used for the calculation are also known.
80

AT332_TRR_E0
2.10.1.2 Step 2: Calculation of Maximum Field Strength

A field strength must not exceed a maximum value, E Max , which is given by:
E Max = E FS = 106,9 20 Log d for land paths, and
E Max = E FS + E SE = 106,9 20 Log d + 2,38 1 exp d 8,94 Log 50 t for sea paths.
Where E FS is the free space field strength for 1 kW ERP, E SE is an enhancement for sea graphs.
2.10.1.3 Step 3: Determination of Transmitter Antenna Height

The transmitter antenna height to be used in the calculation depends on the type and length of the path.
Land paths
h 1 = h eff
Sea paths
h 1 = Max 1 h a
Here, all antenna heights (i.e., h 1 , h eff , and h a ) are in expressed in m. h a is the antenna height above ground and h eff is the
effective height of the transmitter antenna, which is its height over the average level of the ground between distances of
0,2 d and d km from the transmitter in the direction of the receiver.
2.10.1.4 Step 4: Interpolation/Extrapolation of Field Strength

The interpolations are performed in series in the same order as described below. The first interpolation/extrapolation is
performed over the field strength values, E , from the graphs for transmitter antenna height to determine E h1 . The second
interpolation/extrapolation is performed over the interpolated/extrapolated values of E h1 to determine E d . And, the thrid
and final interpolation/extrapolation is performed over the interpolated/extrapolated values of E d to determine E f .
Step 4.1: Interpolation/Extrapolation of Field Strength for Transmitter Antenna Height
If the value of h 1 coincides with one of the eight heights for which the field strength graphs are provided, namely 10, 20, 37.5,
75, 150, 300, 600, and 1200 m, the required field strength is obtained directly from the corresponding graph. Otherwise:
If 10 m h 1 3000 m
The field strength is interpolated or extrapolated from field strengths obtained from two curves using the following
equation:
Log h 1 h Low
E h1 = E Low + E Up E Low ------------------------------------Log h Up h Low
Where h Low = 600 m if h 1 1200 m , otherwise h Low is the nearest nominal effective height below h 1 ,
h Up = 1200 m if h 1 1200 m , otherwise h Up is the nearest nominal effective height above h 1 , E Low is the field
strength value for h Low at the required distance, and E Up is the field strength value for h Up at the required distance.
If 0 m h 1 10 m
For land path if the transmitter-receiver distance is less than the smooth-Earth horizon distance
d H h 1 = 4,1 h 1 , i.e., if d 4,1 h 1 ,
E h1 = E 10 d H 10 + E 10 d E 10 d H h 1 , or
E h1 = E 10 12,9 km + E 10 d E 10 d H h 1 because d H 10 = 12,9 km
For land path if the transmitter-receiver distance is greater than or equal to the smooth-Earth horizon distance
d H h 1 = 4,1 h 1 , i.e., if d 4,1 h 1 ,
E h1 = E 10 d H 10 + d d H h 1 , or E h1 = E 10 12,9 km + d d H h 1 because d H 10 = 12,9 km
Where E x y is the field strength value read for the transmitter-receiver distance of y from the graph available
for the transmitter antenna height of x.
81

If in the above equation, d H 10 + d d H h 1 1000 km even though d 1000 km , the field strength is
determined from linear extrapolation for Log (distance) of the graph given by:
Log d D Low
E h1 = E Low + E Up E Low -------------------------------------Log D Up D Low
Where D Low is penultimate tabulation distance (km), D Up is the final tabulation distance (km), E Low is the field
strength value for D Low , and E Up is the field strength value for D Up .
For sea path, h 1 should not be less than 1 m. This calculation requires the distance at which the path has 0.6 of
the first Fresnel zone just unobstructed by the sea surface. This distance is given by:
D h1 = D 0,6 f h 1 h 2 = 10 m (km)
Df Dh
Where D 0,6 = Max 0,001 ----------------- (km) with D f = 0,0000389 f h 1 h 2 (frequency-dependent term), and
Df + Dh
D h = 4,1 h 1 + h 2 (asymptotic term defined by the horizon distance).
If d D h1 the 0.6 Fresnel clearance distance for the sea path where the transmitter antenna height is 20 m is also
calculated as:
D 20 = D 0,6 f h 1 = 20 m h 2 = 10 m (km)
Once D h1 and D 20 are known, the field strength for the required distance is given by:
E h1
E Max
Log d D h1
= E D + E D E D ---------------------------------h1
20
h1
Log D 20 D h1
E' 1 F S + E'' F S
for d D h1
for D h1 d D 20
for d D 20
Where E Max is the maximum field strength at the required distance as calculated in "Step 2: Calculation of
Maximum Field Strength" on page 81, E D
ED
20
h1
is E Max for d = D h1 ,
Log h1 10
Log h1 10
= E 10 D 20 + E 20 D 20 E 10 D 20 ------------------------------- , E' = E 10 d + E 20 d E 10 d ------------------------------- , and
Log 20 10
Log 20 10
E'' is the field strength calculated as described for land paths. E 10 y and E 20 y are field strengths interpolated
for distance y and h 1 = 10 m and 20 m , respectively, and F S = d D 20 d .
If h 1 0 m
A correction is applied to the field strength, E h1 , calculated in the above description in order to take into account the
diffraction and tropospheric scattering. This correction is the maximum of the diffraction correction,, and
tropospheric scattering correction, .
C h1 = Max C h1d C h1t
Where
C h1d = 6,03 J
with
J = 6,9 + 20 Log 0,1 + 1 + 0,1
and
= K eff2 ,
h1
- , and K is 1.35 for 100 MHz, 3.31 for 600 MHz, 6.00 for 2000 MHz.
eff2 = arc tan ----------9000
e
180 d- a = 6370 km
- with e = -------------------C h1t = 30 Log ---------------------,
(radius of the Earth), and k = 4 3 is the
e + eff2
ak
effective Earth radius factor for mean refractivity conditions.
Step 4.2: Interpolation/Extrapolation of Field Strength for Transmitter-Receiver Distance
In the field strength graphs in the recommendations, the field strength is plotted against distance from 1 km to 1000 km. The
distance values for which field strengths are tabulated are 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25,
30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 225, 250, 275, 300,
325, 350, 375, 400, 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675, 700, 725, 750, 775, 800, 825, 850, 875, 900, 925,
950, 975, 1000. If the transmitter-receiver distance is a value from this list, then interpolation of field strength is not required
and the field strength can be directly read from the graphs.
82

AT332_TRR_E0
If the transmitter-receiver distance does not coincide with the list of distances for which the field strengths are accurately
available from the graphs, the field strength are linearly interpolated or extrapolated for the logarithm of the distance using
the following equation:
Log d d Low
E d = E Low + E Up E Low -------------------------------------Log d Up d Low
Where d Low is the lower value of the nearest tabulated distance to d , d Up is the higher value of the nearest tabulated
distance to d , E Low is the field strength value for d Low , and E Up is the field strength value for d Up .
Step 4.3: Interpolation/Extrapolation of Field Strength for Transmission Frequency
The field strength at the transmission frequency is interpolated from the graphs available for the upper and lower nominal
frequencies as follows:
Log f f Low
E f = E Low + E Up E Low ----------------------------------Log f Up f Low
Where f Low is the lower nominal frequency (100 MHz if f < 600 MHz, 600 MHz otherwise), f Up is the higher nominal
frequency (600 MHz if f < 600 MHz, 2000 MHz otherwise), E Low is the field strength value for f Low , and E Up is the field
strength value for f Up .
In the case of transmission frequencies below 100 MHz or above 2000 MHz, the field strength values are extrapolated from
the two nearer nominal frequency values. The above equation is used for all land paths and sea paths.
2.10.1.5 Step 5: Calculation of Correction Factors

Step 5.1: Correction for Receiver Antenna Height
The receiver antenna height correction depends on the type of path and clutter in which the receiver is located. The field
strength values given by the graphs for land paths are for a reference receiver antenna at a height, R (m), representative of
the height of the clutter surrounding the receiver, subject to a minimum height value of 10 m. Examples of reference heights
are 20 m for an urban area, 30 m for a dense urban area, and 10 m for a suburban area. For sea paths the notional value of R
is 10 m.
For land paths, the elevation angle of the arriving ray is taken into account by calculating a modified representative clutter
1000 d R 15 h 1
height R' , given by R' = Max 1 ----------------------------------------------------------.
1000 d 15
Note that for h 1 6,5 d + R , R' R .
The different correction factors are calculated as follows:
For land path in urban and suburban zones
C Receiver
6,03 J
for h 2 R'
=
h
3,2 + 6,2 Log f Log ----2- for h 2 R'
R'
R' h
2
With J = 6,9 + 20 Log 0,1 + 1 + 0,1 and = 0,0108 f R' h 2 arc tan ---------------2 .
27
10
If R' 10 m , C Receiver is reduced by 3,2 + 6,2 Log f Log ------ .
R'
For land path other zones

h
C Receiver = 3,2 + 6,2 Log f Log -----2-
10
For sea path

d 10 and d h2 are determined as distances at which at which the path has 0.6 of the first Fresnel zone just unobstructed
by the sea surface with h 2 = 10 m
and variable h 2 , respectively. These distances are given by
83

D f D h
d 10 = D 0,6 f h 1 h 2 = 10 m and d h2 = D 0,6 f h 1 h 2 (km), respectively. Here D 0,6 = Max 0,001 ----------------as
D f + D h
explained earlier.
h
If h 2 10 m , C Receiver = 3,2 + 6,2 Log f Log -----2-
10
h
If h 2 10 m and d d 10 , C Receiver = 3,2 + 6,2 Log f Log -----2-
10
If h 2 10 m and d d 10 and d d h2 , C Receiver = 0
Log d d h2
h
If h 2 10 m and d d 10 and d d h2 , C Receiver = 3,2 + 6,2 Log f Log -----2- -------------------------------- 10 Log d 10 d h2
Step 5.2: Correction for Short Urban/Suburban Paths

This correction is only applied when the path loss is to be calculated over land paths, over a transmitter-receiver distance less
than 15 km, in urban and suburban zones. This correction takes into account the presence of buildings in these zones. The
buildings are assumed to be of uniform height.
The correction represents a reduction in the field strength due to building clutter. It is added to the field strength and is given
by:
C Building = 3,3 Log f 1 0,85 Log d 1 0,46 Log 1 + h a R
Where h a is the antenna height above the ground, and R is the clutter height of the clutter class where the receiver is located.
This correction is only applied when d 15 km and h 1 R 150 m .
Step 5.3: Correction for Receiver Clearance Angle
This correction is only applied when the path loss is to be calculated over land paths, and over a transmitter-receiver distance
less than 16 km. This correction gives more precise field strength prediction over small reception areas. The correction is
added to the field strength and is given by:
C Clearance = J ' J
2
Where J = 6,9 + 20 Log 0,1 + 1 + 0,1 , ' = 0,036 f , and = 0,065 Clearance f
Clearance is the clearance angle in degrees determined from:
: The elevation angle of the line from the receiver which just clears all terrain obstructions in the direction of the
transmitter over a distance of up to 16 km but not going beyond the transmitter.
h 1S h 2S
- .
Ref : The reference angle, Ref = arc tan ------------------- 1000 d
Where h 1S and h 2S are the heights of the transmitter and the receiver above sea level, respectively.
2.10.1.6 Step 6: Calculation of Path Loss

First, the final field strength is calculated from the interpolated/extrapolated field strength, E f , by applying the corrections
calculated earlier. The calculated field strength is given by:
E Calc = E f + C Receiver + C Building + C Clearance
The resulting field strength is given by E = Min E Calc E Max , from which the path loss (basic transmission loss, L B ) is
calculated as follows:
L B = 139 E + 20 Log f
2.11 Sakagami Extended Propagation Model

The Sakagami extended propagation model is based on the simplification of the extended Sakagami-Kuboi propagation
model. The Sakagami extended propagation model is valid for frequencies above 3 GHz.
84

AT332_TRR_E0
The Sakagami-Kuboi propagation model requires detailed information about the environment, such as widths of the streets
where the receiver is located, the angles formed by the street axes and the directions of the incident waves, heights of the
buildings close to the receiver, etc. The path loss formula for the Sakagami-Kuboi propagation model is [1]:
H 2
L Model = 100 7,1 Log W + 0,023 + 1,4 Log h s + 6,1 Log H 1 24,37 3,7 -------- Log h b +
h b0
43,2 3,1 Log h b Log d + 20 Log f + e
13 Log f 3,23
Where,
W is the width (in meters) of the streets where the receiver is located
is the angle (in degrees) formed by the street axes and the direction of the incident wave
hs is the height (in meters) of the buildings close to the receiver
H1 is the average height (in meters) of the buildings close to the receiver
hb is the height (in meters) of the transmitter antenna with respect to the observer
hb0 is the height (in meters) of the transmitter antenna with respect to the ground level
H is the average height (in meters) of the buildings close to the base station
d is the separation (in kilometres) between the transmitter and the receiver
f is the frequency (in MHz)
The Sakagami-Kuboi propagation model is valid for:

5m
<W<
50 m
< <
90
5m
< hs <
80 m
5m
< H1 <
50 m
20 m
< hb <
100 m
0.5 km
<d<
10 km
450 MHz
<f<
2200 MHz
h b0 H
Studies [2] have shown that the Sakagami-Kuboi propagation model can be extended to frequencies higher than 3 GHz, which
also allows a simplification in terms of the input required by the model.
The path loss formula for the extended Sakagami-Kuboi propagation model is:
L Model = 54 + 40 Log d 30 Log h b + 21 Log f + a
Where a is a corrective factor with three components:
hm
H
W
a = a H 0 + a W + a h m = 11 Log -----0- 7,1 Log ------ 5 Log ----- 20
20
1,5
W is the width (in meters) of the streets where the receiver is located
H0 (= hs = H1) is the height (in meters) of the buildings close to the receiver
hb (= hb0) is the height (in meters) of the transmitter antenna with respect to the ground
hm is the height (in meters) of the receiver antenna
H is the average height (in meters) of the buildings close to the base station
d is the separation (in metres) between the transmitter and the receiver
f is the frequency (in GHz)
The extended Sakagami-Kuboi propagation model is valid for:

5m
<W<
50 m
10 m
< H0 <
30 m
10 m
< hb <
100 m
0.1 km
<d<
3 km
0.8 GHz
<f<
8 GHz
1.5 m
< hm <
5m
85

Studies also show that above 3 GHz, the path loss predicted by the extended model is almost independant of the input
parameters such as street widths and angles. Therefore, the extended Sakagami-Kuboi propagation model can be simplified
to the extended Sakagami propagation model:
L Model = 54 + 40 Log d 30 Log h b + 21 Log f 5 Log h m
The extended Sakagami propagation model is valid for:
10 m
< hb <
0.1 km
<d<
3 km
3 GHz
<f<
8 GHz
1.5 m
< hm <
5m
100 m
The path loss calculation formula of the Sakagami extended propagation model resembles the formula of the Standard
Propagation Model. In Atoll, this model is in fact a copy of the Standard Propagation Model with the following values assigned
to the K coefficients:
K1
65.4
(calculated for 3.5 GHz)
K2
40
K3
-30
K4
K5
K6
K7
-5
For more information on the Standard Propagation Model, see "Standard Propagation Model (SPM)" on page 65.
References:
[1] Manuel F. Catedra, Jesus Perez-Arriaga, "Cell Planning for Wireless

Communications," Artech House Publishers, 1999.
[2] Koshiro Kitao, Shinichi Ichitsubo, "Path Loss Prediction Formula for Urban and
Suburban Areas for 4G Systems," IEEE, 2006.
2.12 Free Space Loss

The calculation of free space loss is based on ITU 525 recommendations.
L FS = 32,4 + 20 log f + 20 log d
where,
f is the frequency in MHz,
d is the Tx-Rx distance in km,
Free space loss is stated in dB.
2.13 Diffraction
The calculation of diffraction is based on ITU 526-5 recommendations. General method for one or more obstacles (knife-edge
diffraction) is used to evaluate diffraction losses (Diffraction loss in dB). Four construction modes are implemented in Atoll.
All of them are based on this same physical principle presented hereafter, but differ in the way they consider one or several
obstacles. Calculations take the earth curvature into account through the effective Earth radius concept (K factor=1.333).
2.13.1 Knife-edge Diffraction

The procedure checks whether a knife-edge obstructs the first Fresnel zone constructed between the transmitter and the
receiver. The diffraction loss, J(), depends on the obstruction parameter (), which corresponds to the ratio of the
obstruction height (h) and the radius of the Fresnel zone (R).
86

AT332_TRR_E0
Figure 2.15: Knife-Edge Diffraction

R =
c0 n d1 d2
-------------------------------f d1 + d2
where,
n is the Fresnel zone index,
c0 is the speed of light (2.99792 x108 ms-1),
f is the frequency in Hz
d1 is the distance from the transmitter to obstacle in m,
d2 is the distance from obstacle to receiver in m.
We have: = h
--r
where,
R
r = ------2
h is the obstruction height (height from the obstacle top to the Tx-Rx axis).
Hence,
2
For 1 knife-edge method, if 0,7 , J = 6,9 + 20 log 0,1 + 1 + 0,1

Else, J = 0
In case of multiple-knife edge method, the minimum required to estimate diffraction
loss is -0.78.
2.13.2 3 Knife-edge Deygout Method

The Deygout construction, limited to a maximum of three edges, is applied to the entire profile from transmitter to receiver.
This method is used to evaluate path loss incurred by multiple knife-edges. Deygout method is based on a hierarchical knifeedge sorting used to distinguish the main edges, which induce the largest losses, and secondary edges, which have a lesser
effect. The edge hierarchy depends on the obstruction parameter () value.
87

1 Obstacle
Figure 2.16: Deygout Construction 1 Obstacle

A straight line between transmitter and receiver is drawn and the height of the obstacle above the Tx-Rx axis, hi, is calculated.
The obstruction position, di, is also recorded. i are evaluated from these data. The point with the highest value is termed
the principal edge, p, and the corresponding loss is J(p).
Therefore, we have
DiffractionLoss = J P
3 Obstacles
Then, the main edge (point p) is considered as a secondary transmitter or receiver. Therefore, the profile is divided in two
parts: one half profile, between the transmitter and the knife-edge section, another half, constituted by the knife-edgereceiver section.
Figure 2.17: Deygout Construction 3 Obstacles

The same procedure is repeated on each half profile to determine the edge with the higher . The two obstacles found, (points
t and r), are called secondary edges. Losses induced by the secondary edges, J(t) and J(r), are then calculated.
Once the edge hierarchy is determined, the total loss is evaluated by adding all the intermediary losses obtained.
Therefore, if P 0
we have DiffractionLoss = J P + J t + J r
Otherwise, If P 0,7 , DiffractionLoss = J P
88

AT332_TRR_E0
In case of ITU 526-5 and WLL propagation models, Diffraction loss term is determined as
follows:
If P 0,78 , we have DiffractionLoss = J P + J t + J r t
Otherwise DiffractionLoss = 0
J P
- 1
Here, t = min ----------
6
2.13.3 Epstein-Peterson Method

The Epstein-Peterson construction is limited to a maximum of three edges. First, Deygout construction is applied to determine
the three main edges over the whole profile as described above. Then, the main edge height, hp, is recalculated according to
the Epstein-Peterson construction. hp is the height above a straight line connecting t and r points. The main edge position dp
is recorded and p and J(p) are evaluated from these data.
Figure 2.18: Epstein-Peterson Construction

Therefore, we have
DiffractionLoss = J P + J t + J r
2.13.4 Deygout Method with Correction

The Deygout method with correction (ITU 526-5) is based on the Deygout construction (3 obstacles) plus an empirical
correction, C.
Therefore, If P 0 ,
we have DiffractionLoss = J P + J t + J r + C
Otherwise DiffractionLoss = J P + C
In case of ITU 526-5 propagation model, Diffraction loss term is determined as follows:
If P 0,78 , we have DiffractionLoss = J P + t J t + J r + C
Otherwise DiffractionLoss = 0
J P
Here, t = min ------------ 1 and C = 8,0 + 0,04d with d = distance stated in km between
6
the transmitter and the receiver.
2.13.5 Millington Method

The Millington construction, limited to a single edge, is applied over the entire profile. Two horizon lines are drawn at the
transmitter and at the receiver. A straight line between the transmitter and the receiver is defined and the height of the
89

intersection point between the two horizon lines above the Tx-Rx axis, hh, is calculated. The position dh is recorded and then,
from these values, h and J(h) are evaluated using the same previous formulas.
Therefore, we have
DiffractionLoss = J h
Figure 2.19: Millington Construction
2.14 Shadow Fading Model

Propagation models predict the mean path loss as a function of transmission and reception parameters such as frequency,
antenna heights, and distance, etc. Therefore, the predicted path loss between a transmitter and a receiver is constant, in a
given environment and for a given distance. However, in reality different types of clutter may exist in the transmitter-receiver
path. Therefore, the path losses for the same distance could be different along paths that pass throught different types of
environments. The location of the receiver in different types of clutter causes variations with respect to the mean path loss
values given by the path loss models. Some paths undergo more loss while others are less obstructed and may have higher
received signal strength. The variation of path loss with respect to the mean path loss values predicted by the propagation
models, depending on the type of environment is called shadow fading (shadowing) or slow fading. "Slow" fading implies that
the variations in the path loss due to shadow fading occur comparatively slower than the fast fading effect (Rayleigh fading),
which is due to the mobile receiving multipath copies of a signal.
Different types of clutter (buildings, hills, etc.) make large shadows that cause variations in the path loss over long distances.
As a mobile passes under a shadow, the path loss to the mobile keeps varying from point to point. Shadow fading varies as
the mobile moves, while fast fading can vary even if the mobile remains at the same location or moves over very small
distances. It is crucial to account for the shadow fading in order to predict the reliability of coverage provided by any mobile
cellular system.
The shadowing effect is modelled by a log-normal (Gaussian) distribution, as shown in Figure 2.20 on page 90, whose standard
deviation depends on the type of clutter.
Figure 2.20: Log-normal Probability Density Function
90

AT332_TRR_E0
Different clutter types have different shadowing effects. Therefore, each clutter type in Atoll can have a different standard
deviation representing its shadowing characteristics. For different standard deviations, the shape of the Gaussian distribution
curve remains similar, as shown in Figure 2.20 on page 90.
The accuracy of this model depends upon:
The suitability of the range of standard deviation used for each clutter class,
The definition (bin size) of the digital map,
How up-to-date the digital map is,
The number of clutter classes,
The accuracy of assignment of clutter classes.
Shadowing is applied to the predicted path loss differently depending on the technology, and whether it is applied to
predictions or simulations. The following sections explain how shadowing margins are calculated and applied to different
technology documents.
Shadowing margins are calculated for a given cell edge coverage probability. The cell edge coverage probability is the
probability of coverage at a pixel located at the cell edge, and corresponds to the reliability of coverage that you are planning
to achieve at the cell edge. For example, a cell edge coverage probability of 75 % means that the users located at the cell edge
will receive adequate signal level during 75 % of the time. Therefore, a coverage prediction with a cell edge coverage
probability of x % means that the signal level predicted on each pixel is reliable x % of the time, and the overall predicted
coverage area is reliable at least x % of the time.
References:
[1] Saunders S. Antennas and propagation for Wireless Communication Systems

pp. 180-198
[2] Holma H., Toskala A. WCDMA for UMTS
[3] Jhong S., Leonard M. CDMA systems engineering handbook pp. 309-315, 10511053
[4] Remy J.G., Cueugnet J., Siben C. Systmes de radiocommunications avec les
mobiles pp. 309-310
[5] Laiho J., Wacker A., Novosad T. Radio network planning and optimisation for
UMTS pp. 80-81
GSM GPRS EGPRS Documents

The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 96, and applied
to signal level or C/I as explained below.
Signal Level-Based Predictions

Signal level-based predictions include coverage predictions (Coverage by Transmitter, Coverage by Signal Level, and
Overlapping Zones) and calculations in point analysis tabs (Profile and Reception) that require calculation of the
received signal level only, and do not depend on interference.
In these calculations (signal level calculations), a shadowing margin ( M Shadowing model ) is applied to the received
signal level calculated for each pixel. The shadowing margin is calculated for a given cell edge coverage probability,
and depends on the model standard deviation ( model in dB) associated to the clutter class where the receiver is
located.
Interference-Based Predictions
Interference-based predictions include coverage predictions (Coverage by C/I Level, Interfered Zones, Coverage by
GPRS/EDGE Coding Scheme, RLC Throughout/Timeslot, Application Throughput/Timeslot, Circuit Quality Indicator
Analysis) and calculations in point analysis windows Interference tab that require calculation of the received signal
level and interference received from other base stations.
In these calculations, ( C I calculations), the shadowing margin ( M Shadowing C I ) is applied to the ratio of the carrier
power (C) and the interfering signal levels (I) received from the interfering base stations. This shadowing margin is
calculated for a given cell edge coverage probability and depends on the C/I standard deviation ( C I in dB) associated
to the clutter class where the receiver is located.
UMTS HSPA and CDMA2000 1xRTT 1xEV-DO Documents

The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 96 and
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97, and applied to signal level, Ec/I0, or Eb/Nt as
explained below.
91

located.
Interference+noise-Based Predictions
Interference+noise-based predictions include coverage predictions (Pilot Quality Analysis, Downlink Total Noise,
Service Area Analyses, Handoff Status, etc.) and point analysis (AS Analysis tab) that require calculation of the received
signal level and interference and noise received from other base stations.
In these calculations, the shadowing margins ( M Shadowing Ec Io , M Shadowing Eb Nt
DL
, or M Shadowing Eb Nt
UL
are applied to Ec/I0 or Eb/Nt. These shadowing margins are calculated for a given cell edge coverage probability and
depend on the Ec/I0 or Eb/Nt standard deviations ( Ec Io , Eb Nt
DL
, or Eb Nt
UL
, in dB) associated to the clutter
class where the receiver is located.
Macro-Diversity Gains
UL
DL
Atoll calculates the uplink and downlink macro-diversity gains ( G macro diversity and G macro diversity ) depending on
the receiver handover status. These gains are respectively taken into account to evaluate the uplink Eb/Nt in case of
soft handover and the downlink Ec/Io from best server. For detailed description of the calculation of macro-diversity
gains, please refer to "Macro-Diversity Gains Calculation" on page 98.
Monte-Carlo Simulations
Random values for shadowing margins are calculated for each transmitter-receiver link and applied to the predicted
signal level. A shadowing margin for each transmitter-receiver link in each simulation is obtained by taking a random
value from the probability density distribution for the appropriate clutter class. The probability distribution is a lognormal distribution as explained above.
TD-SCDMA Documents
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97, and applied to signal level or interference+noise
predictions as explained below.

Signal level-based predictions include coverage predictions (Best Server and RSCP P-CCPCH Coverages, P-CCPCG
Pollution Analysis, Baton Handover Coverage, DwPCH and UpPCH Coverages, Cell to Cell Interference, and Scrambling
Code Interference) and calculations in point analysis tabs (Profile and Reception) that require calculation of the
located.
Interference+noise-based predictions include coverage predictions (P-CCPCH Eb/Nt and C/I Coverages, Service Area
Analsyses for downlink and uplink Eb/Nt and C/I, etc.) that require calculation of the received signal level and
interference received from other base stations.
In these calculations, the shadowing margins ( M Shadowing Eb Nt
M Shadowing Eb Nt
UL
P CCPCH
M Shadowing Eb Nt
DL
, or
) are applied to Eb/Nt. These shadowing margins are calculated for a given cell edge coverage
probability and depend on the Eb/Nt standard deviations ( Eb Nt
P CCPCH
, Eb Nt
DL
, or Eb Nt
UL
, in dB)
associated to the clutter class where the receiver is located.
92

AT332_TRR_E0
WiMAX, Wi-Fi, and LPWA Documents

"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97 , and applied to signal level or C/(I+N) as explained
below.

located.
Interference-based predictions include coverage predictions (Coverage by C/(I+N) Level, Coverage by Bearer,
Coverage by Throughput, etc.) that require calculation of the received signal level and interference.
In these calculations, (C/(I+N) calculations), in addition to the shadowing margin ( M Shadowing model ) applied to the
received signal level calculated for each pixel, the ratio M Shadowing model M Shadowing C I is applied to the
interfering signal levels (I). M Shadowing C I is calculated for a given cell edge coverage probability and depends on the
C/I standard deviation ( C I in dB) associated to the clutter class where the receiver is located.
The reason why the ratio M Shadowing model M Shadowing C I is used can be understood from the following
derivation (linear, not it dB):
Inputs
C P : The predicted received carrier power without any shadowing margin.
I P : The predicted received interference power without any shadowing margin.
m C : Shadowing margin based on the model standard deviation ( 10
m C I : Shadowing margin based on the C/I standard deviation ( 10
N : Thermal noise
M
Shadowing model
---------------------------------------------------------10
M Shadowing C I
-------------------------------------------------10
Calculations
The effective received carrier power is given by:
C = mC CP
The effective C/I is given by:
C
C
--- = m C I ----PIP
I
The above equations lead to:
mC
mC CP
C - = ----------------------I
- = ---------I = ---------------------mC I P
CP
CP
m C I ----m C I ----IP
IP
mC
Where ----------- corresponds to M Shadowing model M Shadowing C I in dB.
mC I
Therefore, the effective C/(I+N) is given by:
mC CP
C - = -------------------------------------------------m
I + N
C
---------
-I +N
mC I P
93

LTE Documents
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97 , and applied to signal level or C/(I+N) as explained
below.

In these calculations (signal level calculations), a shadowing margin ( M Shadowing model ) is applied to the signal level
calculated for each pixel. The shadowing margin is calculated for a given cell edge coverage probability, and depends
on the model standard deviation ( model in dB) associated to the clutter class where the receiver is located.
Interference-based predictions include coverage predictions (Coverage by C/(I+N) Level, Coverage by Bearer,
Coverage by Throughput, etc.) that require calculation of the received signal level and received interference.
In these calculations, (C/(I+N) calculations), in addition to the shadowing margin ( M Shadowing model ) applied to the
signal level calculated for each pixel, the ratio M Shadowing model M Shadowing C I is applied to the interfering signal
levels (I). M Shadowing C I is calculated for a given cell edge coverage probability and depends on the C/I standard
deviation ( C I in dB) associated to the clutter class where the receiver is located.
The reason why the ratio M Shadowing model M Shadowing C I is used can be understood from the following
derivation (linear, not it dB):
Inputs
C P : The predicted received carrier power without any shadowing margin.
I P : The predicted received interference power without any shadowing margin.
m C : Shadowing margin based on the model standard deviation ( 10
m C I : Shadowing margin based on the C/I standard deviation ( 10
N : Thermal noise
M Shadowing model
---------------------------------------------------------10
M Shadowing C I
-------------------------------------------------10
Calculations
The effective received carrier power is given by:
C = mC CP
The effective C/I is given by:
C
C
--- = m C I ----PIP
I
The above equations lead to:
mC
mC CP
C
I = ---------------------- = ---------------------- = ----------- I P
mC I
C
CP
m C I ----Pm C I ----IP
IP
mC
Where ----------- corresponds to M Shadowing model M Shadowing C I in dB.
mC I
Therefore, the effective C/(I+N) is given by:
94

AT332_TRR_E0
mC CP
C --------------= -----------------------------------m
I + N
C
---------- I P + N
mC I
Downlink Macro-Diversity CoMP Gain

MD
Atoll calculates the downlink macro-diversity CoMP gain ( G CoMP ) depending on the terminals CoMP support as well
as the CoMP mode used by the CoMP set to which the studied serving cell belongs (downlink dynamic point selection
or coherent joint transmission). This gain is taken into account to evaluate the downlink signal level from best server.
With CoMP macro-diversity gain, the received carrier power and C/(I+N) become:
C = m C g MD C P
m C g MD C P
C - = -------------------------------------------------mC
I + N
---------
-I +N
mC I P
Here g MD is the linear value of the downlink macro-diversity CoMP gain. For detailed description of the calculation of
macro-diversity gains, please refer to "Downlink Macro-Diversity Gain Evaluation" on page 101.
2.14.1 Shadowing Margin Calculation

The following sections describe the calculation method used for determining different shadowin margins.
The following shadowing margins are calculated using the method described below:
Network Type
GSM GPRS EGPRS
UMTS HSPA
Standard Deviation
MShadowing
Applied to
model
M Shadowing model
C I
M Shadowing C I
C/I
model
M Shadowing model
Ec Io
M Shadowing Ec Io
Ec/I0
Eb Nt
Eb Nt
CDMA2000
DL
UL
Eb/Nt (DL)
UL
Eb/Nt (UL)
M Shadowing model
Ec Io
M Shadowing Ec Io
Ec/I0
Eb Nt
DL
UL
model
Eb Nt
P CCPCH
Eb Nt
Eb Nt
WiMAX
M Shadowing Eb Nt
DL
model
Eb Nt
TD-SCDMA
M Shadowing Eb Nt
DL
UL
M Shadowing Eb Nt
M Shadowing Eb Nt
DL
Eb/Nt (DL)
UL
Eb/Nt (UL)
M Shadowing model
M Shadowing Eb Nt
P CCPCH
M Shadowing Eb Nt
M Shadowing Eb Nt
C
Eb/Nt P-CCPCH
DL
Eb/Nt (DL)
UL
Eb/Nt (UL)
model
M Shadowing model
C and C/(I+N)
C I
M Shadowing C I
C/(I+N)
95

Network Type
Standard Deviation
MShadowing
Applied to
model
M Shadowing model
C and C/(I+N)
C I
M Shadowing C I
C/(I+N)
LTE
2.14.1.1 Shadowing Margin Calculation in Predictions

Shadowing margins, MShadowing, are calculated from standard deviation values defined for the clutter class where the pixel
(probe mobile) is located, and required cell edge coverage probability, and applied to the path loss, Lpath.
Shadowing Error PDF (1 Signal)
The measured path loss in dB can be expressed as a Gaussian random variable:
L = L path + dB G 0 1
where,
Lpath is the predicted path loss,

dB is the user-defined standard deviation of the error,
G(0,1) is a zero-mean unit-variance Gaussian random variable.
Therefore, the probability density function (pdf) for the random (shadowing) part of path loss is:
1
p L x = -------------------- e
dB 2
2
x
-------------2
2
dB
The probability that the shadowing error exceeds z dB is

2
x ------------2
2 dB
PL x z =
pL x dx
z
1
= -------------------- e
dB 2
dx
Normalising x by dividing it bydB:
1
P L x z = ----------
2
2
x----2
z
dx = Q --------
dB
z-------- dB
where Q is the complementary cumulative function.

To ensure a given cell edge coverage probability, R L , for the predicted value, a shadowing margin, M Shadowing , is added to
the link budget.
Confidence in the prediction can be expressed as:
C d = P' Tx L P rec L P' Tx P rec G 0 1 dB M Shadowing
where,
P rec is the signal level predicted at the receiver. P rec = P' Tx L path M Shadowing
P' Tx = EIRP + G antRx L Rx
EIRP is the effective isotropic radiated power of the transmitter.

L Rx are receiver losses.
G antRx is the receiver antenna gain.
The shadowing margin is calculated such that:

M Shadowing
P C d P rec = R L M Shadowing = 1 P L x M Shadowing 0 = 1 Q -------------------------
dB
A lookup table is used for mapping the values of Q vs. a set of cell edge coverage probabilities.
96

AT332_TRR_E0
M Shadowing
Figure 2.21: Normalised Margin M arg in = ------------------------- dB
In interference-based predictions, where signal to noise ratio is calculated, the shadowing margin is only applied to the signal
from the interfered transmitter (C). We consider that the interference value is not altered by the shadowing margin. Random
variations also exist in the interfering signals, but taking only the average interference gives accurate results. [3] explains how
a certain level of interference is maintained by congestion control in CDMA-based networks.
2.14.1.2 Shadowing Margin Calculation in Monte-Carlo Simulations

Shadowing margins, MShadowing, are calculated from standard deviation values defined for the clutter class where the pixel
(probe mobile) is located, and required cell edge coverage probability, and added to the path loss, Lpath.
Random values are generated during Monte-Carlo simulation. Each user is assigned a service, a mobility type, an activity
status, a geographic position and a random shadowing value.
For each link, path loss (L) can be broken down to L = L path + .
Here, is a zero mean gaussian random variable G 0 dB representing variation due to shadowing. It can be expressed as
the sum of two uncorrelated zero mean gaussian random variables, L and P . L models the error related to the receivers
location (surrounding environment), and remains the same for all links between the receiver and the base stations from which
it is receiving signals. P models the error related to the path between the transmitter and the receiver.
Therefore, in case of two links, we have:
1
1 = L + P for link 1
2
2 = L + P for link 2
i
Standard deviations of L L and P P can be calculated from i , the model standard deviation model , and the
correlation coefficient between 1 and 2 .
Assuming all P have the same standard deviations, we have:
2
model = L + P
2
L
= --------------2
model
Therefore,
2
P = model 1
L = model
is set to 0.5 in Atoll, which gives:
model
model
- and P = -------------- L = --------------2
2
97


Receiver
Therefore, to model shadowing error common to all the signals received at a receiver ( E Shadowing model ), values are
randomly generated for each receiver. These values have a zero-mean gaussian distribution with a standard deviation of
model
--------------- , where model is the model standard deviation associated with the receivers clutter class.
2
Next, Atoll generates another random value for each transmitter-receiver pair. This values represents the shadowing error
Path
not related to the location of the receiver ( E Shadowing model ). These values also have a zero-mean gaussian distribution with
model
- .
a standard deviation --------------
2
So, we have:
Receiver
Path
E Shadowing model = E Shadowing model + E Shadowing model

Random shadowing error has its mean value at zero. Hence, this shadowing modelling method has no impact on the simulated
network load. On the other hand, as shadowing errors on the transmitter-receiver links are uncorrelated, the method
influences the calculated macro-diversity gain in case the mobile is in soft handover.
2.14.2 Macro-Diversity Gains Calculation

The following sections explain how uplink and downlink macro-diversity gains are calculated in UMTS HSPA and CDMA2000
1xRTT 1xEV-DO documents for predictions and AS Analysis tab of the point analysis tool.
The calculation and use of macro-diversity gains can be disabled through the Atoll.ini file.
For more information, see the Administrator Manual.
2.14.2.1 Uplink Macro-Diversity Gain Evaluation

In UMTS HSPA and CDMA2000 1xRTT 1xEV-DO, mobiles may be in soft handoff (mobile connected to cells located on different
sites). In this case, we can consider the shadowing error pdf described below.
2.14.2.1.1
Shadowing Error PDF (n Signals)

For each link, path loss (L) can be broken down as:
L = L path +
is a zero mean gaussian random variable G 0 dB representing variation due to shadowing. It can be expressed as the
sum of two uncorrelated zero mean gaussian random variables, L and P . L models error related to the receiver local
environment; it is the same whichever the link. P models error related to the path between transmitter and receiver.
1
1 = L + P for the link 1

2

Knowing i , the uplink Eb/Nt standard deviation Eb Nt and the correlation coefficient between 1 and 2 , we can
UL
calculate standard deviations of L L and P P (assuming all P have the same standard deviations).
We have:
2
Eb Nt
UL
= L + P
2
L
= ----------------------2
Eb Nt
UL
Therefore,
2
P = Eb Nt
98
UL

AT332_TRR_E0
2
L = Eb Nt
UL
2 Signals Without Recombination

In technologies supporting soft handoff (UMTS and CDMA2000), cell is interference limited. As for one link, to ensure a
required cell edge coverage probability R L for the prediction, we add to each link budget a shadowing margin,
2signals
M Shadowing Eb Nt
UL
Prediction reliability in order to have Eb/Nt higher or equal to Eb/Nt from the best server can be expressed as:
Cd
1
1
-------1 = P' Tx1 L 1 N 1 CI pred 1 P' Tx1 L path N 1 CI pred
1
N1
or
Cd
1
1
-------2 = P' Tx2 L 2 N 2 CI pred 2 P' Tx2 L path N 2 CI pred
2
N2
where
i
CI pred is the quality level (signal to noise ratio) predicted at the receiver for link i.
Ni is the noise level for link i.
We note:
2signals
M Shadowing Eb Nt
UL
= P' Txi L path N i CI pred

i
and
2
1 = CI pred CI pred
2
1 is the minimum needed margin on each link.

Therefore, the probability of having a quality at least equal to the best predicted one is:
noMRC
RL
noMRC
RL
Cd
Cd
2signals
1
1
M Shadowing Eb Nt = 1 P L1 L2 -------1 CI pred -------2 CI pred
UL
N
N
1
2
2signals
M Shadowing Eb Nt = 1 P
UL
1 2
2signals
2signals
1 M Shadowing Eb Nt 2 M Shadowing Eb Nt
UL
UL
We can express it using L , P and P

P
1
1 2
2signals
UL
= P L P
L

1 2 1
2signals
M Shadowing Eb Nt 2 M Shadowing Eb Nt
1
1 2 P
P P
2signals
M Shadowing Eb Nt
2signals
noMRC
RL
2signals
2signals
UL
2signals
UL
2
1 L = L
L P M Shadowing Eb Nt
M Shadowing Eb Nt 2 M Shadowing Eb Nt
= P L P P M Shadowing Eb Nt
L
UL
UL
UL
UL
1 L
2
1 L = L
2signals
L P P M Shadowing Eb Nt
P
UL
1 L
2signals
M Shadowing Eb Nt
UL
1
2signals
2
2signals
2
= 1 P L P P M Shadowing Eb Nt L P P M Shadowing Eb Nt 1 L d L
L
P
UL
P
UL
99

2signals
P P M Shadowing Eb Nt
P
1
---------------=
P 2 2signals
M
Shadowing Eb Nt
UL
e
UL
2
x
---------2
2
P
2signals
M Shadowing Eb Nt UL L
dx = Q ----------------------------------------------------------------
P
Then, we have:
noMRC
RL
2signals
M Shadowing Eb Nt
UL
2signals
M Shadowing Eb Nt UL
2signals
2
L
M Shadowing Eb Nt UL 1 L
= 1 P L Q ---------------------------------------------------------------- Q ---------------------------------------------------------------------------- d L
L
P
P
If we introduce user defined standard deviation Eb Nt and correlation coefficient , and consider that P is a
UL
Gaussian pdf:
noMRC
RL
2signals
M Shadowing Eb Nt
UL
1
= 1 ---------- e
2
2
xL
--------2
2
M 2signals
M 2signals
x L Eb Nt
x L Eb Nt
1
Shadowing Eb Nt
Shadowing Eb Nt
UL
UL
UL
UL
Q ----------------------------------------------------------------------------------------------- Q ------------------------------------------------------------------------------------------------------------ dx L
Eb Nt
1
Eb Nt
1

UL
UL
n Signals Without Recombination

We can generalize the previous expression to n signals (n is the number of available signals - Atoll may consider up to 3
signals):
noMRC
RL
nsignals
M Shadowing Eb Nt
UL
1
= 1 ---------- e
2
2
xL
--------2
M nsignals
M nsignals
Shadowing Eb Nt UL x L Eb Nt UL
Shadowing Eb Nt UL x L Eb Nt UL 1
Q ----------------------------------------------------------------------------------------------- Q ------------------------------------------------------------------------------------------------------------ dx L
Eb Nt UL
Eb Nt UL
The case where softer handoff occurs (two signals from co-site cells) is equivalent to the one signal case. The Softer/soft case
is equivalent to the two signals case. For the path associated with the softer recombination, we will use combined SNR to
calculate the availability of the link.
Correlation Coefficient Determination
There is currently no agreed model for predicting correlation coefficient between 1 and 2 . Two key variables influence
correlation:
The angle between the two signals. If this angle is small, correlation is high.
The relative values of the two signal lengths. If angle is 0 and lengths are the same, correlation is zero. Correlation is
different from zero when path lengths differ.
A simple model has been found [1]:

= -----T D1
------- when T
D2
T is a function of the mean size of obstacles near the receiver and is also linked to the receiver environment.
In a normal handover status, assuming a hexagonal design for sites, is close to (+/- /3) and D1/D2 is close to 1.
In [1,5], = 0,5 when = 0,3 and T = ------ .

10
In Atoll, is set to 0.5.
100

AT332_TRR_E0
2.14.2.1.2
Uplink Macro-Diversity Gain

UL
Atoll determines the uplink macro-diversity gain ( G macro diversity ) from the shadowing margins calculated in case of one
signal and n signals.
Therefore, we have:
UL
nsignals
G macro diversity = M Shadowing Eb Nt
UL
M Shadowing Eb Nt
UL
Where n is the number of cell-mobile signals.
2.14.2.2 Downlink Macro-Diversity Gain Evaluation

In UMTS HSPA and CDMA2000 1xRTT 1xEV-DO, mobiles in soft handoff are able to switch from one cell to another if the best
pilot suddenly fades. In LTE, mobiles using CoMP and served by cells using downlink dynamic point selection or coherent joint
transmission CoMP are either able to switch from one serving cell to another dynamically or are simultaneously served by
more than one cell.
To model this function, we have to consider the probability of fading over the shadowing margin, both for the best signal and
for all the other available signals, in the shadowing margin calculation.
Let us consider the shadowing error pdf described below.
2.14.2.2.1
Shadowing Error PDF (n Signals)

For each link, path loss (L) can be broken down as:
L = Lpath +
is a zero mean gaussian random variable G 0 dB representing variation due to shadowing. It can be expressed as the
sum of two uncorrelated zero mean gaussian random variables, L and P . L models the error related to the receiver local
environment, which is the same for all links. P models the error related to the path between the transmitter and the
receiver.
1

2

Knowing i , the Ec/Io standard deviation Ec I o and the correlation coefficient between 1 and 2 , we can calculate
standard deviations of L L and P P (assuming all P have the same standard deviations).
We have:
2
Ec I o = L + P
2
L
= ------------2
Ec I o
Therefore,
2
P = Ec I o 1
L = Ec I o
2 Available Signals
In technologies supporting soft handoff (UMTS and CDMA2000) cells are interference limited. As for one link, to ensure a
2signals
required cell edge coverage probability R L for the prediction, we add a shadowing margin, M Shadowing Ec Io , to each link
budget.
Ec
Ec
Prediction reliability to have ----- -----
for the best server can be expressed as:
Io
Io pred
Ec
Ec 1
Ec 1
-------1- = P pilot L 1 Io -----
1 P pilot L m Io -----
Io pred
Io pred
1
1
1
Io
101

Or
Ec 2
Ec 1
Ec 1
-------- = P pilot L 2 Io -----
2 P pilot L m Io -----
2
2
2
Io pred
Io pred
Io
We note:
1
Ec
2signals
M Shadowing Ec Io = P pilot L m Io -----
i
i
Io pred
Ec 1
Ec 2
2
1 = -----
-----
Io pred Io pred
2
1 is the minimum needed margin on each link.

Therefore, probability of having a quality at least equal to the best predicted one is:
noMRC
RL
noMRC
RL
Ec
Ec
Ec 1
Ec 1
2signals
M Shadowing Ec Io = 1 P L1 L2 -------1- -----
-------2- -----
Io Io pred Io Io pred
2signals
2signals
2signals
M Shadowing Ec Io = 1 P 1 2 1 M Shadowing Ec Io 2 M Shadowing Ec Io 1

1
We can express it by using L , P and P

2signals
2signals
P 1 2 1 M Shadowing Ec Io 2 M Shadowing Ec Io 1 L = L
= P L P
L
1
1 2 P

P P
2signals
2signals
M Shadowing Ec Io L P M Shadowing Ec Io 1 L
2signals
2signals
P 1 2 1 M Shadowing Ec Io 2 M Shadowing Ec Io 1 L = L
1
2signals
2signals
= P L P P M Shadowing Ec Io L P P M Shadowing Ec Io 1 L
L
noMRC
RL
2signals
M Shadowing Ec Io
= 1
PL L PP P MShadowing Ec Io L PP P MShadowing Ec Io 1 L dL
1
2signals
2signals
1
i
2signals
P P M Shadowing Ec Io L = ----------------P
P 2
SHO L
2
x --------2
2 P
2signals
M Shadowing Ec Io L
dx = Q ------------------------------------------------------
P
Then, we have:
noMRC
2signals
RL
M Shadowing Ec Io
= 1
2signals
2signals
M Shadowing Ec Io L
M Shadowing Ec Io 1 L
- Q ----------------------------------------------------------------- d L
P L Q ----------------------------------------------------L
P
P
If we introduce a user defined Ec/Io standard deviation and a correlation coefficient and consider that P is a
L
Gaussian pdf:
noMRC
RL
2signals
M Shadowing Ec Io
1
= 1 ---------2
2
x
L
--------2
2signals
2signals
M Shadowing Ec Io x L Ec I o
M Shadowing Ec Io 1 x L Ec I o
- Q --------------------------------------------------------------------------------------- dx L
Q -------------------------------------------------------------------------- Ec I o 1
Ec I o 1
n Available Signals
We can generalize the previous expression for n signals (n is the number of available signals - Atoll may consider up to 3
signals):
102

AT332_TRR_E0
noMRC
RL
nsignals
M Shadowing Ec Io
1
= 1 ---------2
2
x
L
--------2
nsignals
M Shadowing Ec Io x L Ec I o
Q ---------------------------------------------------------------------------- x
Ec I o 1
i=2
nsignals
M Shadowing Ec Io 1 x L Ec I o
Q ---------------------------------------------------------------------------------------- dx L
Ec I o 1
1 =1 dB
2
1 =5 dB
2
1 =10 dB
Figure 2.22: Margin - Probability (Case of 2 Signals)
2 signals
3
1 =5 dB
3
1 =10 dB
Figure 2.23: Margin - Probability (Case of 3 Signals with sigma = 8dB, delta1 = 1dB)
2 signals
3
1 =5 dB
3
1 =10 dB
Figure 2.24: Margin - Probability (Case of 3 Signals with sigma = 8dB, delta1 = 2dB)
Correlation Coefficient Determination
For further information about determination of the correlation coefficient, please see "Correlation Coefficient Determination"
on page 103.
2.14.2.2.2
Downlink Macro-Diversity Gain

DL
Atoll determines the downlink macro-diversity gain ( G macro diversity ) from the shadowing margins calculated in case of one
signal and n signals.
Therefore, we have:
103

DL
nsignals
G macro diversity = M Shadowing Ec Io M Shadowing Ec Io

Where n is the number of available signals.
2.15 Path Loss Matrices

Atoll is able to calculate two path loss matrices per transmitter, a first matrix over a smaller radius computed with a high
resolution and a propagation model (main matrix), and a second matrix over a larger radius computed with a low resolution
and another propagation model (extended matrix).
To be considered for calculations, a transmitter must fulfil the following conditions:
It must be active,
It must satisfy filter criteria defined in the Transmitters folder, and
It must have a calculation area.
In the rest of the document, a transmitter fulfilling the conditions detailed above will be called TBC transmitter.
The path loss matrix size of a TBC transmitter depends on its calculation area. Atoll determines a path loss value ( L path ) on
each calculation bin (calculation bin is defined by the resolution) of the calculation area of the TBC transmitter. You may have
one or two path loss matrices per TBC transmitter.
2.15.1 Calculation Area Determination

Transmitter calculation area is made of a rectangle or a square depending on transmitter calculation radius and the
computation zone.
Calculation radius enables Atoll to define a square around the transmitter. One side of the square equals twice the entered
calculation radius.
Since the computation zone can be made of one or several polygons, transmitter calculation area corresponds to the
intersection area between its calculation square and the rectangle containing the computation zone area(s).
Figure 2.25: Example 1: Single Calculation Area
Figure 2.26: Example 2: Multiple Calculation Areas

Computation zone
Rectangle containing the computation zone(s)
104

AT332_TRR_E0
Calculation area defined (square)

Transmitter
Actual calculation area on which Atoll calculates path losses
2.15.2 Validity of Path Loss Matrices

Most geographic data modifications and some radio data changes make path loss matrices invalid. This table lists these
modifications and also changes that have an impact only on coverage predictions.
Modification
Matrix validity
Impact on
Calculate
Force calculation
Frequency
Invalid
Path loss
matrices
Sufficient
Not necessary
Antenna coordinates
(site coordinate: X and Y, Dx and Dy)
Invalid
Path loss
matrices
Sufficient
Not necessary
Antennaa height
Invalid
Path loss
matrices
Sufficient
Not necessary
Antennaa pattern
Invalid
Path loss
matrices
Sufficient
Not necessary
Downtilta
Invalid
Path loss
matrices
Sufficient
Not necessary
Azimutha
Invalid
Path loss
matrices
Sufficient
Not necessary
% Power (secondary antennas)
Invalid
Path loss
matrices
Sufficient
Not necessary
Site position/altitude
Invalid
Path loss
matrices
Sufficient
Not necessary
Grid resolution (main or extended)
Invalid
Path loss
matrices
Sufficient
Not necessary
Propagation model (main or extended)
Invalid
Path loss
matrices
Sufficient
Not necessary
Propagation model parameters
Invalid
Path loss
matrices
Sufficient
Not necessary
Calculation areas
(Calculation areas gets smaller)
Valid
Coverage
predictions
Sufficient
Not necessary
Calculation areas
(Calculation areas gets larger)
Invalid
Path loss
matrices
Sufficient
Not necessary
Receiver height
Invalid
Path loss
matrices
Sufficient
Not necessary
Receiver losses
Valid
Coverage
predictions
Sufficient
Not necessary
Receiver gain
Valid
Coverage
predictions
Sufficient
Not necessary
Receiver antenna
Valid
Coverage
predictions
Sufficient
Not necessary
Geographic layer order
Invalid
Path loss
matrices
Insufficientb
Necessary
Geographic file resolution
Invalid
Path loss
matrices
Insufficientb
Necessary
New DTM map
Invalid
Path loss
matrices
Insufficientb
Necessary
Clutter class edition
Invalid
Path loss
matrices
Insufficientb
Necessary
Coverage prediction resolution
Valid
Coverage
predictions
Sufficient
Not necessary
105

Modification
Matrix validity
Impact on
Calculate
Force calculation
Cell edge coverage probability
Valid
Coverage
predictions
Sufficient
Not necessary
Coverage prediction conditions
Valid
Coverage
predictions
Sufficient
Not necessary
Coverage prediction display options
Valid
Coverage
predictions
Sufficient
Not necessary
a.
b.
Modification of any parameter related to main or other antennas makes matrix invalid.
Except if this action has an impact on the site positions/altitudes.
2.15.3 Path Loss Tuning

Atoll can tune path loss matrices obtained from propagation results by the use of real measurements (CW Measurements or
Test Mobile Data). For each measured transmitter, Atoll tries to merge measurements and predictions on the same points
and to smooth the surrounding points of the path loss matrices for homogeneity reasons. A transmitter path loss matrix can
be tuned several times by the use of several measurement paths. All these tuning paths are stored in a catalogue. This
catalogue is stored under a .tuning folder containing a .dbf file and one .pts file per tuned transmitter. Since a tuning file can
contain several measurement paths, all these measurements are added to the tuning file.
For more information on the tuning files, see the Administrator Manual.
2.15.3.1 Transmitter Path Loss Tuning

The same algorithm is used for CW Measurement and Test Mobile Data. It is also the same for main and extended matrices.
Path Losses tuning will be done using two steps.
1. Total matrix correction
A mean error is calculated between each measured value and the corresponding bin in the pathloss matrix. Mean
error is calculated for each pathloss matrix (main and extended) of each transmitter. This mean error is then applied
to all the matrix bins. This tuning is done to smooth the local corrections (step 2) of measured values and not the tuned
bins.
2. Local correction for each measured value
For each measured value, an ellipse is used to define the pathloss area which has to be tuned. The main axis of the
ellipse is oriented to the transmitter.The ellipse is user-defined by two parameters:
The radius of the axis parallel to the Profile (A)

The radius of the axis perpendicular to the Profile (B)
Lets take M a measurement value and P i the path loss value at point i, before any tuning.
M is limited by the minimum measurement threshold defined in the interface.
The squared elliptic distance between i and M is given by:

2
Xi XM
Yi YM
D i = ----------------------- + ----------------------2
2
A
B
Where:
X i and X M are the X-coordinates of i and M respectively
Y i and Y M are the Y-coordinates of i and M respectively
The mean error for the first tuning is given by:
1
E = ---
n
ei
i
Where e i is the error between measurement and prediction at point i
106

AT332_TRR_E0
E is limited by the maximum total correction defined in the interface.
Then, the path loss value is tuned using E:

Pi
new
= Pi
old
+E
Finally, a second tuning ( R i ) is applied where:

Ri = 1 Di M g Pi
new
so R i = 1 D i M g P i
old
+ E
Where g is (measurement gain - losses).

R i is limited by the maximum local correction defined in the interface.
So, the final tuned path loss is:

Pi
tuned
= Pi
new
+ R i so P i
tuned
= Pi
old
+ E + Ri
When several ellipses overlap a pathloss bin, the final tuned path loss is given by:
1 d j P j
tuned
j
= ------------------------------------------------
d j
n
Pi
tuned
Where n is the number of overlapping ellipses
2.15.3.2 Repeater Path Loss Tuning

In the case of repeaters, Atoll provides only a composite measured value per pixel which is a combination of the contribution
of both a transmitter and one or several repeaters. In order to tune the path loss matrices of donor transmitters and repaters,
its is mandatory to split the contribution of each element in the measured value as starting point.
Lets take M the measured value.
M = Md + Mr
where :
M d represents the contribution of the donor transmitter in the measured value.
M r represents the contribution of the repeater in the measured value.
All the values are used in Watts.
If C d and C r represent respectively the filtered signal level from the donor transmitter and the repeater on a pixel, one can
define the contribution of each element as follows:
Cd
Cr
M d = M ---------------and M r = M ---------------- .
Cd + Cr
Cd + Cr
Following the path loss tuning process described in "Transmitter Path Loss Tuning" on page 106, the donor transmitter (resp.
the repeater) is then tuned using M d (resp. M r ) values.
107

2.16 File Formats

2.16.1 Path Loss Matrix File Format
When path loss matrices are stored externally, i.e., outside the ATL file, the path loss matrices folder contains a pathloss.dbf
file containing the calculation parameters of the transmitters and one LOS (path loss results) file per calculated transmitter.
The path loss matrices folder also contains a LowRes folder with another pathloss.dbf file and one LOS (path loss results) file
per transmitter that has an extended path loss matrix.
The formats of the pathloss.dbf and LOS files are described here.
2.16.1.1 Pathloss.dbf File Format

The pathloss.dbf file has a standard DBF (dBase III) format. The file can be opened in Microsoft Access, but it should not be
modified without consulting the Forsk customer support.
For general information, the format of DBF files in any Xbase language is as follows:
Notations used in the following tables: FS = FlagShip; D3 = dBaseIII+; Fb = FoxBase; D4 =
dBaseIV; Fp = FoxPro; D5 = dBaseV; CL = Clipper
DBF Structure
Byte
Description
Remarks
0...n
DBF header (see next part for size, byte 8)
n+1
1st record of fixed length (see next parts); 2nd record (see next part for size, byte10) ; last
record
last
optional: 0x1a (eof byte)
If .dbf is not empty
DBF Header
The DBF header size is variable and depends on the field count.
Byte
Size
Contents
Description
Applies to
00
0x03
plain .dbf
FS, D3, D4, D5, Fb, Fp, CL
0x04
plain .dbf
D4, D5 (FS)
0x05
plain .dbf
D5, Fp (FS)
with .dbv memo var size
FS
with .dbv and .dbt memo
FS
0x83
with .dbt memo
FS, D3, D4, D5, Fb, Fp, CL
0x8B
with .dbt memo in D4 format
D4, D5
0x8E
with SQL table
D4, D5
0xF5
with .fmp memo
Fp
01
YYMMDD
Last update digits
All
04
ulong
Number of records in file
All
08
ushort
Header size in bytes
All
10
ushort
Record size in bytes
All
12
0,0
Reserved
All
14
0x01
Begin transaction
D4, D5
0x00
End Transaction
D4, D5
0x00
ignored
FS, D3, Fb, Fp, CL
0x01
Encrypted
D4, D5
0x00
normal visible
All
0 (1)
multi-user environment use
D4,D5
15
16
108
0x43
0xB3
1
12

AT332_TRR_E0
Byte
Size
Contents
Description
Applies to
28
0x01
production index exists
Fp, D4, D5
0x00
index upon demand
All
language driver ID
D4, D5
0x01
codepage437 DOS USA
Fp
0x02
codepage850 DOS Multi ling
Fp
0x03
codepage1251 Windows ANSI
Fp
0xC8
codepage1250 Windows EE
Fp
0x00
ignored
FS, D3, Fb, Fp, CL
29
30
32
n*32
+1
0,0
0x0D
reserved
All
Field Descriptor (see next paragraph)
all
Header Record Terminator
all
Field descriptor array in the DBF header (32 bytes for each field):
Byte
Size
Contents
Description
Applies to
11
ASCI
field name, 0x00 termin
all
11
ASCI
field type (see next paragraph)
all
12
n,n,n,n
Fld address in memory
D3
n,n,0,0
offset from record begin
Fp
0,0,0,0
ignored
FS, D4, D5, Fb, CL
16
byte
Field length, bin (see next paragraph)
all \ FS,CL: for C field type
17
byte
decimal count, bin
all / both used for fld lng
18
0,0
reserved
all
20
21
23
byte
Work area ID
D4, D5
0x00
unused
FS, D3, Fb, Fp, CL
n,n
multi-user dBase
D3, D4, D5
0,0
ignored
FS, Fb, Fp, CL
0x01
Set Fields
D3, D4, D5
0x00
ignored
FS, Fb, Fp, CL
24
0...0
reserved
all
31
0x01
Field is in .mdx index
D4, D5
0x00
ignored
FS, D3, Fb, Fp, CL
Field type and size in the DBF header, field descriptor (1 byte):
Size
Type
Description/Storage
Applies to
C 1...n
Char
ASCII (OEM code page chars)

rest= space, not \0 term.
all
n = 1...64kb (using deci count)
FS
n = 1...32kb (using deci count)
Fp, CL
n = 1...254
all
Date
8 ASCII digits (0...9) in the YYYYMMDD format
all
F 1...n
Numeric
ASCII digits (-.0123456789)

variable pos. of float.point
n = 1...20
FS, D4, D5, Fp
N 1...n
Numeric
ASCII digits (-.0123456789)

fix posit/no float.point
all
n = 1...20
FS, Fp, CL
n = 1...18
D3, D4, D5, Fb
ASCII chars (YyNnTtFf space)
FS, D3, Fb, Fp, CL
D8
L1
Logical
109

Size
Type
Description/Storage
Applies to
ASCII chars (YyNnTtFf?)
D4, D5 (FS)
Memo
10 digits repres. the start block posit. in .dbt file, or 10

spaces if no entry in memo
all
V 10
Variable
Variable, bin/asc data in .dbv

4bytes bin= start pos in memo
4bytes bin= block size
1byte = subtype
1byte = reserved (0x1a)
10 spaces if no entry in .dbv
FS
P 10
Picture
binary data in .ftp

structure like M
Fp
B 10
Binary
binary data in .dbt

structure like M
D5
G 10
General
OLE objects
structure like M
D5, Fp
22
short int
binary int max +/- 32767
FS
44
long int
binary int max +/- 2147483647
FS
88
double
binary signed double IEEE
FS
M 10
Each DBF record (fixed length):

Byte
Size
Description
Applies to
deleted flag "*" or not deleted " "
All
1n
x-times contents of fields, fixed length, unterminated.

For n, see (2) byte 1011
All
2.16.1.2 Pathloss.dbf File Contents

The DBF file provides information that is needed to check validity of each path loss matrix.
110
Field
Type
Description
TX_NAME
Text
Name of the transmitter
FILE_NAME
Text
Name (and optionally, path) of .los file
MODEL_NAME
Text
Name of propagation model used to calculate path loss
MODEL_SIG
Text
Signature (identity number) of model used in calculations. You can check it in the
propagation model properties (General tab).
The Model_SIG is used for the purpose of validity. A unique Model_SIG is assigned to
each propagation model. When model parameters are modified, the associated model
ID changes. This enables Atoll to detect path loss matrix invalidity. In the same way, two
identical propagation models in different projects do not have the same model IDa.
ULXMAP
Float
X-coordinate of the top-left corner of the path loss matrix upper-left pixel
ULYMAP
Float
Y-coordinate of the top-left corner of the path loss matrix upper-left pixel
RESOLUTION
Float
Resolution of path loss matrix in metre
NROWS
Float
Number of rows in path loss matrix
NCOLS
Float
Number of columns in path loss matrix
FREQUENCY
Float
Frequency band
TILT
Float
Transmitter antenna mechanical tilt
AZIMUTH
Float
Transmitter antenna azimuth
TX_HEIGHT
Float
Transmitter height in metre
TX_POSX
Float
X-coordinate of the transmitter
TX_POSY
Float
Y-coordinate of the transmitter
ALTITUDE
Float
Ground height above sea level at the transmitter in metre
RX_HEIGHT
Float
Receiver height in metre
ANTENNA_SI
Float
Logical number referring to antenna pattern. Antennas with the same pattern will have
the same number.

AT332_TRR_E0
Field
Type
Description
MAX_LOS
Float
Maximum path loss stated in 1/16 dB. This information is used, when no calculation
radius is set, to check the matrix validity.
CAREA_XMIN
Float
Lowest x-coordinate of centre pixel located on the calculation radiusb
CAREA_XMAX
Float
Highest x-coordinate of centre pixel located on the calculation radius
CAREA_YMIN
Float
Lowest y-coordinate of centre pixel located on the calculation radius
CAREA_YMAX
Float
Highest y-coordinate of centre pixel located on the calculation radius
WAREA_XMIN
Float
Lowest x-coordinate of centre pixel located in the computation zonec
WAREA_XMAX
Float
Highest x-coordinate of centre pixel located in the computation zone
WAREA_YMIN
Float
Lowest y-coordinate of centre pixel located in the computation zone
WAREA_YMAX
Float
Highest y-coordinate of centre pixel located in the computation zone
LOCKED
Boolean
Locking status
0: path loss matrix is not locked
1: path loss matrix is locked.
Boolean
Atoll indicates if losses due to the antenna pattern are taken into account in the path
loss matrix.
0: antenna losses not taken into account
1: antenna losses included
INC_ANT
a.
b.
c.
In order to benefit from the calculation sharing feature, users must retrieve the propagation models from the same
central database. This can be done using the Open from database command for a new document or the Refresh
command for an existing one. Otherwise, Atoll generates different model_ID (even if same parameters are applied on
the same kind of model) and calculation sharing become unavailable due to inconsistency.
These coordinates enable Atoll to determine the area of calculation for each transmitter.
These coordinates enable Atoll to determine the rectangle including the computation zone.
2.16.1.3 LOS File Format

The LOS (path loss results) files are binary files with a standard row-column structure. Data are stored starting from the
southwest to the northeast corner of the area. The file contains 16-bit signed integer values in the range [-32768; +32767]
with a 1/16 dB precision. "No data" values are represented by +32767.
2.16.2 Path Loss Tuning File Format

Atoll can tune path losses calculated by propagation models using CW measurements or drive test Data. Path losses are tuned
by merging measurement data with propagation results on pixels corresponding to the measurement points and the pixels in
the vicinity. Path losses surrounding the measurement points are smoothed for homogeneity.
Measuremment paths that are used for path loss tuning are stored as a catalogue in a folder containing a pathloss.dbf file and
one PTS (path loss tuning) file per transmitter. A tuning file can contain several measurement paths.
For more information on the path loss tuning algorithm, see the Technical Reference Guide.
2.16.2.1 Pathloss.dbf File Format

See "Pathloss.dbf File Format" on page 108.
2.16.2.2 Pathloss.dbf File Contents

The DBF file provides information about the measured transmitters involved in the tuning.
Field
Type
Description
TX_NAME
Text
FILE_NAME
Text
Name (and optionally, path) of .pts file
AREA_XMIN
Float
Not used
AREA_XMAX
Float
Not used
AREA_YMIN
Float
Not used
AREA_YMAX
Float
Not used
111

2.16.2.3 PTS File Format

The PTS (path loss tuning) files contain a header and the list of measurement points.
Header:
4 bytes: version
4 bytes: flag (can be used to manage flags like active flag)
50 bytes: GUID
4 bytes: number of points
255 bytes: original measurement name (with prefix "Num" for drive test data and "CW" for CW measurements)
256 bytes: comments
4 bytes: X_RADIUS
4 bytes: Y_RADIUS
4 bytes: gain = measurement gain - losses
4 bytes: global error
4 bytes: rx height
4 bytes: frequency
8 bytes: tx Position
List of measurement points:
4 bytes: X
4 bytes: Y
4 bytes: measurement value
4 bytes: incidence angle.
2.16.3 Interference Matrix File Formats

Interference matrices are used by GSM, LTE, and WiMAX AFPs (automatic frequency planning tools). Interference matrices
can be imported and exported using the following formats:
GSM: CLC, IM0, IM1, IM2

LTE and WiMAX: IM2, TXT, CSV
Interference matrix files must contain interference probability values between 0 and 1, and not in precentage (between 0 and
100%). When interference matrix files are imported, Atoll does not check their validity and imports interference probability
values for loaded transmitters only.
In the following format descriptions and samples, lines starting with the "#" are considered as comments.
In GSM interference matrices:
The interferer TRX type is not specified and is always considered to be BCCH.
Subcells have different powers defined as offsets with respect to the BCCH. For
subcells other than the BCCH, if the power offset of a subcell is X dB, then its
interference histogram will be shifted by X dB with respect to the BCCH interference
histogram.
If no power offset is defined on the interfered TRX type, it is possible to set "All".
For each interfered subcell-interferer subcell pair, Atoll saves probabilities for
several C/I values (6 to 24 values), including five fixed ones: 9, 1, 8, 14, and 22 dB.
Between two fixed C/I value, there can be up to three additional values (this number
depends on the probability variation between the fixed values). The C/I values have
0.5 dB accuracy and probability values are calculated and stored with an accuracy of
0.002 for probabilities between 1 and 0.05, and with an accuracy of 0.0001 for
probabilities lower than 0.05.
2.16.3.1 CLC Format (One Value per Line)

The CLC format uses wo ASCII text files: a CLC file and a DCT file. Interference matrices are imported by selecting the CLC file
to import. Atoll looks for the associated DCT file in the same directory and uses it to decode transmitter identifiers. If no DCT
file is available, Atoll assumes that the transmitter identifiers are the transmitter names, and the columns 1 and 2 of the CLC
file must contain the names of the interfered and interfering transmitters instead of their identification numbers.
2.16.3.2 CLC File Format

The CLC file consists of two parts. The first part is a header used for format identification. It must start with and contain the
following lines:
112

AT332_TRR_E0
# Calculation Results Data File.

# Version 1.1, Tab separated format. Commented lines start with #.
The second part details interference histogram of each interfered subcell-interfering subcell pair. The lines after the header
are considered as comments if they start with "#". If not, they must have the following format:
<Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline>
The 5 tab-separated columns are defined in the table below:
Column
Name
Description
Column1
Interfered transmitter
Identification number of the interfered transmitter. If the column is empty,

its value is identical to the one of the line above.
Column2
Interfering transmitter
Identification number of the interferer transmitter. If the column is null, its

value is identical to the one of the line above.
Column3
Interfered TRX type
Interfered subcell. If the column is null, its value is identical to the one of the
line above. In order to save storage, all subcells with no power offset are not
duplicated (e.g. BCCH, TCH).
Column4
C/I threshold
C/I value. This column cannot be null.
Column5
Probability C/I > Threshold
Probability to have C/I the value specified in column 4 (C/I threshold). This
field must not be empty.
The columns 1, 2, and 3 must be defined only in the first line of each histogram.
Sample
# Version 1.1,
# Remark:
Tab separated format. Commented lines start with #.
C/I results do not incorporate power offset values.
# Fields are:
##------------#------------#------------#-----------#------------------#
#| Interfered | Interfering| Interfered | C/I
#| Transmitter| Transmitter| Trx type
| Probability
| Threshold | C/I >= Threshold |
##------------#------------#------------#-----------#------------------#
#
# Warning, The parameter settings of this header can be wrong if
# the "export" is performed following an "import". They
# are correct when the "export" follows a "calculate".
#
# Service Zone Type is "Best signal level of the highest priority HCS layer".
# Margin is 5.
# Cell edge coverage probability 75%.
# Traffic spreading was Uniform
##---------------------------------------------------------------------#
1
TCH_INNER
8
9
1
0.944
10
0.904
11
0.892
14
0.844
15
0.832
16
0.812
17
0.752
22
0.316
25
0.292
113

BCCH,TCHa
a.
0.944
10
.904
13
0.872
14
0.84
17
0.772
If the TCH and BCCH histograms are the same, they are not repeated. A single record indicates that the histograms
belong to TCH and BCCH both.
2.16.3.3 DCT File Format

The .dct file is divided into two parts. The first part is a header used for format identification. It must start with and contain
the following lines:
# Calculation Results Dictionary File.
The second part provides information about transmitters taken into account in AFP. The lines after the header are considered
as comments if they start with "#". If not, they must have the following format:
<Column1><tab><Column2><newline>
Column
Name
Type
Description
Column1
Transmitter name
Text
Column2
Transmitter Identifier
Integer
Identification number of the transmitter
Column3
BCCH during calculation
Integer
BCCH used in calculations
Column4
BSIC during calculation
Integer
BSIC used in calculations
Column5
% of vic coverage
Float
Percentage of overlap of the victim service area
Column6
% of int coverage
Float
Percentage of overlap of the interferer service area
The last four columns describe the interference matrix scope. One transmitter per line is described separated with a tab
character.
Sample
# Calculation Results Dictionary File.
# Version 2.1,
Tab separated format. Commented lines start with #.
# Fields are:
##-----------#-----------#-----------#-----------#---------#---------#
#|Transmitter|Transmitter|BCCH during|BSIC during|% of vic'|% of int'|
#|Name
|Identifier |calculation|calculation|coverage |coverage |
##-----------#-----------#-----------#-----------#---------#---------#
#
#
# Service Zone Type is "Best signal level per HCS layer".
# Margin is 5.
# Cell edge coverage probability is 75%.
# Traffic spreading was Uniform (percentage of interfered area)
##---------------------------#
114
Site0_0
-1
-1
100
100
Site0_1
-1
-1
100
100

AT332_TRR_E0
Site0_2
-1
-1
100
100
Site1_0
-1
-1
100
100
Site1_1
-1
-1
100
100
Site1_2
-1
-1
100
100
Site2_0
-1
-1
100
100
Site2_1
-1
-1
100
100
2.16.3.4 IM0 Format (One Histogram per Line)

This file contains one histogram per line for each interfered/interfering subcell pair. The histogram is a list of C/I values with
associated probabilities.
The .im0 file consists of two parts. The first part is a header used for format identification. It must start with and contain the
following lines:
The second part details interference histogram of each interfered subcell-interferer subcell pair. The lines after the header
<Column1><tab><Column2><tab><Column3><tab><Column4><newline>
Column
Name
Description
Column1
Name of the interfered transmitter.
Column2
Name of the interferer transmitter.
Column3
Interfered TRX type
Interfered subcell. In order to save storage, all subcells with no power offset
are not duplicated (e.g. BCCH, TCH).
Column4
C/I probability
C/I value and the probability associated to this value separated by a space
character. This entry cannot be null.
Sample
# Remark:
# Fields are:
#-----------------------------------------------------------------------#Transmitter
Interferer
TRX type
{C/I Probability} values
#-----------------------------------------------------------------------#
#
# Margin is 5.
##---------------------------------------------------------------------#
#
Site0_2
Site0_1
BCCH,TCH-10 1 -9 0.996 -6 0.976 -4 0.964 -1 0.936

0 0.932 1 0.924 4 0.896 7 0.864 8 0.848
9 0.832 10 0.824 11 0.804 14 0.712 17 0.66
115

Site0_2
Site0_3
BCCH,TCH-10 1 -9 0.996 -6 0.976 -4 0.972 -1 0.948

0 0.94 1 0.928 4 0.896 7 0.856 8 0.84
11 0.772 13 0.688 14 0.636 15 0.608 18 0.556
Site0_3
Site0_1
BCCH,TCH-10 1 -9 0.996 -6 0.98 -3 0.948 0 0.932

1 0.924 4 0.892 7 0.852 8 0.832 9 0.816
10 0.784 11 0.764 14 0.644 15 0.616 18 0.564
Site0_3
Site0_2
BCCH,TCH-9 1 -6 0.972 -3 0.964 -2 0.96 0 0.94

1 0.932 4 0.904 7 0.876 8 0.86 9 0.844
11 0.804 13 0.744 14 0.716 15 0.692 18 0.644
2.16.3.5 IM1 Format (One Value per Line, TX Name Repeated)

This file contains one C/I threshold and probability pair value per line for each interfered/interfering subcell pair. The
histogram is a list of C/I values with associated probabilities.
The .im1 file consists of two parts. The first part is a header used for format identification. It must start with and contain the
following lines:
The second part details interference histogram of each interfered subcell-interferer subcell pair. The lines after the header
<Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline>
Column
Name
Description
Column1
Name of the interfered transmitter.
Column2
Name of the interferer transmitter.
Column3
Interfered TRX type
Interfered subcell. In order to save storage, all subcells with no power offset
are not duplicated (e.g. BCCH, TCH).
Column4
C/I threshold
C/I value. This column cannot be null.
Column5
Probability C/I > Threshold
Probability to have C/I the value specified in column 4 (C/I threshold). This
field must not be empty.
Sample
# Remark:
# Fields are:
#-----------------------------------------------------------------------#Transmitter
Interferer
TRX type
C/I
Probability
#-----------------------------------------------------------------------#
#
# Margin is 5.
##---------------------------------------------------------------------#
116

AT332_TRR_E0
Site0_2
Site0_1
BCCH,TCH
-10
Site0_2
Site0_1
BCCH,TCH
-9
0.996
Site0_2
Site0_1
BCCH,TCH
-6
0.976
Site0_2
Site0_1
BCCH,TCH
-4
0.964
Site0_2
Site0_1
BCCH,TCH
-1
0.936
Site0_2
Site0_1
BCCH,TCH
0.932
Site0_2
Site0_1
BCCH,TCH
0.924
Site0_2
Site0_1
BCCH,TCH
0.896
Site0_2
Site0_1
BCCH,TCH
0.864
Site0_2
Site0_1
BCCH,TCH
0.848
Site0_2
Site0_1
BCCH,TCH
0.832
Site0_2
Site0_1
BCCH,TCH
10
0.824
...
2.16.3.6 IM2 Format (Co- and Adjacent-channel Probabilities)

IM2 files contain co-channel and adjacent-channel interference probabilities for each interfered transmitter interfering
transmitter pair. In GSM, there is only one set of values for all the subcells of the interfered transmitter. Each line must have
the following format:
<Column1><SEP><Column2><SEP><Column3><SEP><Column4><newline>
Where the separator (<SEP>) can either be a tab or a semicolon.
The four columns are defined in the table below:
Column
Name
Description
Column1
Name of the interfered transmitter
Column2
Name of the interferer transmitter
Column3
Co-channel probability
Co-channel interference probability
Column4
Adjacent-channel probability
Adjacent channel interference probability
Sample
# Remark:
# Fields are:
#-----------------------------------------------------------------------#Transmitter
Interferer
Co-channel
Adjacent channel
#-----------------------------------------------------------------------#
#
# Margin is 5.
##---------------------------------------------------------------------#
Site0_2
Site0_1
0.226667
0.024
Site0_2
Site0_3
0.27
0.024
Site0_3
Site0_1
0.276
0.02
117

Site0_3
Site0_2
0.226
0.028
The columns in the sample above are separated with a tab. These columns can also be separated with a semilcolon:
Site0_2;Site0_1;0.226667;0.024
Site0_2;Site0_3;0.27;0.024
Site0_3;Site0_1;0.276;0.02
Site0_3;Site0_2;0.226;0.028
2.16.4 "Per Transmitter" Prediction File Format

When a coverage prediction is calculated by value intervals it is stored externally, i.e., outside the ATL file. A corresponding
<doc_name>\{<GUID>} folder is actually created where the ATL document is located, as soon as the latter is saved.
The calculation of the coverage prediction is either global or "per transmitter".
When the calculation is global, the results are stored in two files for the entire prediction: one HDR file and one BIL
file (both identified by the prediction name).
When a calculation is "per transmitter", one HDR file and one BIL file are created for each transmitter in the prediction
(both identified by the transmitters name). In some "per transmitter" predictions, an additional DBF file is created for
the entire prediction (identified by the prediction name). The DBF file contains information on each transmitter and a
pointer to each transmitters specific HDR and BIL files.
In LTE, when a Cell Identifier Collision Zones (DL) prediction is calculated by value intervals with the display type set
"No. of interferers per cell", the HDR file and the BIL file are created for each cell in the prediction (both identified by
the cells name).
The format and the content of the DBF file is described here.
In both cases, an XML file describing the prediction is also created in the corresponding
<doc_name>\{<GUID>} folder.
2.16.4.1 <per_transmitter_prediction>.dbf File Format

The format of <per_transmitter_prediction>.dbf files is identical to the format described in "Pathloss.dbf File Format" on
page 108.
2.16.4.2 <per_transmitter_prediction>.dbf File Contents

The <per_transmitter_prediction>.dbf files generated in specific {<GUID>} folders provide information that is needed to
check the validity of each "per transmitter" prediction> calculated by value intervals.
Field
Type
Description
TX_NAME
Text
FILE_NAME
Text
Name of the transmitters BIL result file
RESOLUTION
Float
Resolution of the calculation, same as xdim and ydim in the HDR file
AREA_XMIN
Float
Same as ulxmap in the HDR file
AREA_XMAX
Float
Same as ulxmap + xdim * ncols in the HDR file
AREA_YMIN
Float
Same as ulymap in the HDR file
AREA_YMAX
Float
Same as ulymap + ydim * nrows in the HDR file
NBITS
Float
Same as nbits in the HDR file
NBANDS
Float
Same as nbands in the HDR file
BYTE_ORDER
Float
Same as byteorder in the HDR file
BAND_ROW_BYTES
Float
Same as bandrowbytes in the HDR file
TOTAL_ROW_BYTES
Float
Same as totalrowbytes in the HDR file
SKIP_BYTES
Float
Same as skipbytes in the HDR file
DATA_TYPE
Text
Same as datatype in the HDR file
NO_DATA_VALUE
118
Same as nodatavalue in the HDR file

AT332_TRR_E0
2.16.5 Coverage Prediction Export and Reports

2.16.5.1 Filtering Coverage Predictions at Export
Raster and vector coverage predictions can be filtered at export in order to exclude holes and islands. Predictions are filtered
by setting the colour of a pixel to the dominant colour of the bounding box, i.e., surrounding pixels, using a dispersion factor:
2
exp D X 2 .
Here, D is the distance from the pixel to be coloured to each pixel within the bounding box and X is the value at that pixel.
In other words, the pixel will be coloured by the most representative value within this bounding box.
Figure 2.27: Bounding box for prediction filtering

The user-defined filtering percentage Y gives the size of the bounding box: Y 10 pixels in each direction. In other words, the
bounding box is increased by one pixel every 10 % (since Y is defined as a percentage).
2.16.5.2 Smoothing Coverage Predictions at Export

Vector coverage predictions can be smoothed at export in orer to simplify its contours. Predictions are smoothed by reducing
the number of points defining the contours of the polygons using a vertex reduction routine that successively reduces the
number of closely clustered vertices (vertex reduction within tolerance of prior vertex cluster, Douglas-Peucher polyline
simplification).
Two smoothing methods exist for defining the degree of coverage smoothing: smoothing by percentage and smoothing by
the maximum number of points.
Smoothing by Percentage
2
Z
The user-defined smoothing percentage Z gives the approximation tolerance: ------- R ------ , where R is the user-defined
2
20
export resolution. Tolerance is the interval within which Atoll tries to reduce the number of points.
Figure 2.28: Smoothing Tolerance Definition

For example, for three successive points, A1, A2, and A3 as shown in Figure 2.29 on page 120, A2 will be deleted if within this
tolerance (and A1 and A3 will be directly linked) and A2 will be conserved if outside this tolerance.
119

A2 inside the tolerance interval
A2 outside the tolerance interval
Figure 2.29: Smoothing by Percentage

Smoothing by Number of Points
The second method consists in defining a maximum number of points to be deleted. This number of points helps the algorithm
to determine the optimised tolerance (see "Smoothing by Percentage" on page 119) such that, with this obtained tolerance,
the number of points to be deleted will be lower than this value.
Lets consider the following example ( 1 ). Starting from the maximum possible tolerance, the number of points to be filtered
out are estimated (circled in red in the following example ( 2 )). If this number is greater than the maximum number of points
defined by the user, Atoll reduces the tolerance until reaching the requested maximum number of points or less ( 3 ). The
first the number of points respecting the constraint is obtained, smoothing is applied by deleting these points and linking the
remaining closest points ( 4 ).
1
Figure 2.30: Smoothing by Number of Points
2.16.5.3 Examples of Prediction Export Filtering and Smoothing

Figure 2.31 on page 121 shows the original signal level coverage prediction whose filtered and smoothed exported results are
presented in Figure 2.32 on page 121.
120

AT332_TRR_E0
Figure 2.31: Bounding box for prediction filtering
Filtering Percentage: 0 %
Smoothing Percentage: 0 %


Figure 2.32: Exported prediction with filtering and smoothing
2.16.5.4 Coverage Prediction Reports Over Focus/Computation Zones

Statistics are calculated in coverage prediction reports over the focus zone or the computation zone, if no focus zone exists,
or the covered area, if neither zone exists.
If the reference surface area for the statistics is based on a focus or computation zone, there may be minute inaccuracies in
the calculated statistics because of the difference in the surface area calculation methods:
The surface areas of the zones (polygons) are calculated by triangulation.

The surface area of a coverage predictions is calculated by counting the number of covered pixels and multiplying this
number with the area of one pixel, calculated from resolution of the coverage prediction.
At the border of the focus or computation zone, a pixel is considered inside the zone if its centre is inside. Otherwise, the pixel
is considered outside the zone. This estimation may give rise to inaccuracies.
121

122
Chapter 3
GSM GPRS EDGE
Networks
"Signal Level Calculation" on page 125
"Interference-based DL Calculations" on page 131
"GPRS/EDGE Calculations" on page 136
"Codec Mode Selection and CQI Calculations" on

page 145
"UL Coverage Predictions" on page 152
"Traffic Analysis" on page 159
"Network Dimensioning" on page 171
"Key Performance Indicators Calculation" on page 181
"Simulations" on page 184
"Automatic Neighbour Allocation" on page 191
"AFP Appendices" on page 195

124

Chapter 3: GSM GPRS EDGE Networks
AT332_TRR_E0
3 GSM GPRS EDGE Networks

This chapter describes all the calculations performed in Atoll GSM/GPRS/EDGE documents. The first four sections describe the
signal level, interference, GPRS/EDGE-specific, and CQI calculations, respectively. The following three sections explain the
traffic analysis, network dimensioning, and KPI calculation processes. The last section describes the neighbour allocation
process in GSM.
All the calculations are performed on TBC (to be calculated) transmitters. For the
definition of TBC transmitters please refer to "Path Loss Matrices" on page 104.
Logarithms used in this chapter (Log function) are base-10 unless stated otherwise.
3.1 Signal Level Calculation

3.1.1 DL Signal Level
Three parameters can be studied in point analysis (Profile tab) and in signal level-based coverage predictions:
Studied Parameter
Txi
DL Signal level ( P rec )
Formulas
Signal level received from a transmitter on a TRX type
Txi
P rec tt
Txi
= EIRP tt P tt L path M Shadowing model L Indoor + G ant
Txi
L Term
Txi
Txi
L path = L model + L ant
Path loss ( L path )

DL Total losses ( L total )
Term
Txi
Tx
Txi
L total = L path + M Shadowing model + L Indoor + L Tx DL + L Term G ant + G ant

Tx
Term
Here,
EIRP is the effective isotropic radiated power of the transmitter,

L model is the loss on the transmitter-terminal path (path loss) calculated by the propagation model,
L ant
Tx
is the transmitter antenna attenuation (from antenna patterns),
M Shadowing model is the shadowing margin. This parameter is taken into account when the option Shadowing taken
into account is selected,

L Indoor are the indoor losses, taken into account when the option Indoor coverage is selected,
L Term are the terminal losses,
L Tx DL are the total transmitter DL losses,
G ant
P is the power offset defined for the selected TRX type in the transmitter property dialog box,
tt is the TRX type (in the GSM GPRS EDGE.mdb document template, there are three possible TRX types, BCCH, TCH and
inner TCH).
Term
is the terminal antenna gain,
3.1.2 UL Signal Level

Two parameters can be studied in UL signal level-based coverage predictions:
Studied Parameter
Term
UL Signal level ( P rec )

Term
UL Total losses ( L total )
Formulas
Signal level received from a terminal at a transmitter
Term
P rec
Term
max
P Term
Txi
+ G ant L path M Shadowing model L Indoor + G ant

Tx
Term
L Tx UL L Term
Txi
L total = L path + M Shadowing model + L Indoor + L Tx UL + L Term G ant + G ant

Tx
Term
Here,
max
P Term is the maximum terminal allowed power,
L path is the path loss on the transmitter-terminal link, identical in DL and in UL (see above),
Txi
125

M Shadowing model is the shadowing margin. This parameter is taken into account when the option Shadowing taken
into account is selected,

L Term are the terminal losses,
L Tx UL are the total transmitter UL losses,
G ant
G ant
Tx
is the transmitter antenna gain,
Term
is the terminal antenna gain.
3.1.3 Point Analysis

3.1.3.1 Profile Tab
Txi
Txi
For a selected transmitter, it is possible to display the signal level received from a TRX type ( P rec tt ), the path loss, L path , or
Txi
the DL total losses, L total DL . Path loss and DL total losses are the same for all TRX types.
If the power reduction values defined for all the subcells are the same, the received signal level from the selected transmitter
will be the same for all TRX types.
3.1.3.2 Reception Tab

Analysis provided in the Reception tab is based on path loss matrices. Therefore, it is possible to display the signal levels
received from TBC transmitters for which path loss matrices have been calculated over their calculation areas.
Txi
Txi
For each transmitter, Atoll can display the signal level received from a TRX type ( P rec tt ), the path loss, L path , or the total
Txi
losses, L total DL . Path loss and total losses are the same for all TRX types.
If the power reduction values defined for all the subcells are the same, the received signal level from the selected transmitter
will be the same for all TRX types.
Reception level bars are displayed in the order of decreasing signal level. The number of displayed bars depends on the signal
level received from the best server. Bars are only displayed for transmitters whose signal level is within a 30 dB margin from
the best server signal level.
You can use a value other than 30 dB for the margin from the best server signal level, for
example a smaller value for improving the calculation speed. For more information on
defining a different value for this margin, see the Administrator Manual.
3.1.4 Signal Level-based DL Coverage Predictions

For each TBC transmitter, Txi, Atoll calculates the selected parameter on each pixel inside the Txi calculation area. In other
words, each pixel inside the Txi calculation area is considered a probe (non-interfering) receiver.
Coverage prediction parameters to be set are:
The coverage conditions in order to determine the service area of each TBC transmitter, and
The display settings to select the displayed parameter and its shading levels.
3.1.4.1 DL Service Area Determination

Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialog box to determine the areas
coverage will be displayed.
We can distinguish eight cases as below. Let us assume that:
3.1.4.1.1
Each transmitter, Txi, belongs to a Hierarchical Cell Structure (HCS) layer, k, with a defined priority and a defined
reception threshold.
No max range is set.
All Servers
The service area of Txi corresponds to the pixels :
Txi
MinimumThreshold P rec tt MaximumThreshold
126

AT332_TRR_E0
Txi
For pure signal level-based calculations (not C/I or C/(I+N)), P rec tt can be replaced
Txi
Txi
with L total DL or L path .
3.1.4.1.2
Best Signal Level and an Overlap Margin

Txi

Txi
Txi
Txi
Txi
Txj
And P rec tt Best P rec tt M
ji
M is the specified overlap margin (dB). The Best function considers the highest value from a list of values.
3.1.4.1.3
If M = 0 dB, Atoll considers pixels the received signal level from Txi is the highest.
If M = 2 dB, Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from
the highest.
If M = -2 dB, Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from
transmitters which are 2nd best servers.
Second Best Signal Level and an Overlap Margin

The service area of Txi corresponds to the pixels:
Txi

Txi
Txi
Txi
nd
Txi
Txj
And P rec tt 2 Best P rec tt M
ji
M is the specified overlap margin (dB). The 2nd Best function considers the second highest value from a list of values.
3.1.4.1.4
If M = 0 dB, Atoll considers pixels the received signal level from Txi is the second highest.
If M = 2 dB, Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin
from the second highest.
transmitters which are 3rd best servers.
Best Signal Level per HCS Layer and an Overlap Margin

For each HCS layer, k, the service area of Txi corresponds to the pixels :
Txi

Txi
Txi
Txi
Txi
Txj
And P rec BCCH Best P rec BCCH M
ji
127

3.1.4.1.5
the highest.
Second Best Signal Level per HCS Layer and an Overlap Margin
Txi

Txi
Txi
Txi
nd
Txi
Txj
And P rec BCCH 2 Best P rec BCCH M
ji
3.1.4.1.6
HCS Servers and an Overlap Margin

Txi

Txi
Txi
Txi
Txi
Txj
ji
Txi
And the received P rec tt exceeds the reception threshold defined per HCS layer.
3.1.4.1.7
the highest.
Highest Priority HCS Server and an Overlap Margin

Txi

Txi
Txi
Txi
128

AT332_TRR_E0
Txi
Txj
ji
And Txi belongs to the HCS layer with the highest priority. The highest priority is defined by the priority field (0: lowest).
Txi
In the case two layers have the same priority, the traffic is served by the transmitter for
which the difference between the received signal strength and the HCS threshold is the
highest. The way the competition is managed between layers with the same priority can
be modified. For more information, see the Administrator Manual.
3.1.4.1.8
the highest.
Best Idle Mode Reselection Criterion (C2)

Such type of coverage is useful :
To compare idle and dedicated mode best servers for voice traffic
Display the GPRS/EDGE best server map (based on GSM idle mode)
The path loss criterion C1 used for cell selection and reselection is defined by:
Txi
C1 = P rec BCCH MinimumThreshold BCCH

The path loss criterion (GSM03.22) is satisfied if C1 0 .
The reselection criterion C2 is used for cell reselection only and is defined by:
C2 = C1 + CELL_RESELECT_OFFSET
CELL_RESELECT_OFFSET is the Cell Reselect Offset defined for the transmitter.
Txi
MinimumThreshold P rec BCCH MaximumThreshold

Txi
Txi
Txi
And C2
Txi
Txj
BCCH = Best C2 BCCH
j
The Best function considers the highest value from a list of values.
On each pixel, the transmitter with the highest C2 value is kept. It corresponds to the best server in idle mode. C2 is defined
as an integer in the 3GPP specifications, therefore, the C2 values in the above calculations are rounded down to the nearest
integer.
3.1.4.2 Coverage Display

3.1.4.2.1
Coverage Resolution
The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data
and can be defined separately for each coverage prediction. Coverage predictions are generated using a bilinear interpolation
method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes, see "Path Loss Calculation
Prerequisites" on page 57 for more information).
3.1.4.2.2
Display Types
It is possible to display the coverage predictions with colours depending on any transmitter attribute or other criteria such as:
129

Signal Level (in dBm, dBV, dBV/m)

Atoll calculates signal level received from the transmitter on each pixel of each transmitter service area. A pixel of a service
area is coloured if the signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on signal
level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as transmitter service areas. Each layer shows the different signal levels available in the transmitter service area.
Best Signal Level (in dBm, dBV, dBV/m)
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. When other service
areas overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the signal level is greater
than or equal to the defined thresholds (the pixel colour depends on the signal level). Coverage consists of several
independent layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each
layer corresponds to an area the signal level from the best server exceeds a defined minimum threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if path loss is greater than or equal to the defined minimum thresholds (pixel colour depends on path loss). Coverage
consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as service
areas. Each layer shows the different path loss levels in the transmitter service area.
Total Losses (dB)
Atoll calculates DL total losses from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if total losses is greater than or equal to the defined minimum thresholds (pixel colour depends on total losses).
Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many layers
as service areas. Each layer shows the different total losses levels in the transmitter service area.
Best Server Path Loss (dB)
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. When other service
areas overlap the studied one, Atoll determines the best transmitter and evaluates path loss from the best transmitter. A pixel
of a service area is coloured if the path loss is greater than or equal to the defined thresholds (pixel colour depends on path
loss). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as defined thresholds. Each layer corresponds to an area the path loss from the best server exceeds a defined minimum
threshold.
Best Server Total Losses (dB)
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. service areas overlap
the studied one, Atoll determines the best transmitter and evaluates DL total losses from the best transmitter. A pixel of a
service area is coloured if the total losses is greater than or equal to the defined thresholds (pixel colour depends on total
losses). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as defined thresholds. Each layer corresponds to an area the total losses from the best server exceed a defined
minimum threshold.
Number of Servers
Atoll evaluates how many service areas cover a pixel in order to determine the number of servers. The pixel colour depends
on the number of servers. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area the number of servers is greater than or
equal to a defined minimum threshold.
Cell Edge Coverage Probability (%)
On each pixel of each transmitter service area, the coverage corresponds to the pixels the signal level from this transmitter
fulfils signal conditions defined in Conditions tab with different cell edge coverage probabilities. There is one coverage area
per transmitter in the explorer.
Best Cell Edge Coverage Probability (%)
On each pixel of each transmitter service area, the coverage corresponds to the pixels the best signal level received fulfils
signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability in the explorer.
Best C2 (dBm)
Atoll calculates C2 values received from transmitters on each pixel of each transmitter service area. When other service areas
overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the C2 value is greater than or
130

AT332_TRR_E0
equal to the defined thresholds (the pixel colour depends on the C2 value). Coverage consists of several independent layers
whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer corresponds to
an area the best C2 value exceeds a defined minimum threshold.
3.2 Interference-based DL Calculations

Interference-based calculations include all the calculations that involve the calculation of interference received from
interfering transmitters in addition to the signal level received from the server.
3.2.1 DL Carrier-to-Interference Ratio Calculation

MSA (Mobile Station Allocation) Definition
In order to understand the difference between each frequency hopping mode from the point of view of a mobile, it is interesting to consider the Mobile Station Allocation. MSA is characterised by the pair (Channel list, MAIO). In the following, we
will use this notion to characterise the interference and resources set of a mobile.
For non-hopping (NH) mode, the channel list is 1 channel. For base-band hopping (BBH) or synthesized frequency hopping
(SFH), the channel list corresponds to the mobile allocation list (MAL).
For BBH, channels of MAL belong to the same TRX type.
Examples:
Non-hopping (NH): An MSA is the channel assigned to a TRX used by a mobile.

TRX index
Channel list
MAIO
MSA
53
(53,-)
54
(54,-)
Baseband hopping (BBH): An MSA is the Mobile Allocation List (MAL) and the TRX index.
TRX index
Channel list
MAIO
MSA
53
([53,54,55],0)
54
([53,54,55],1)
55
([53,54,55],2)
Synthesised frequency hopping (SFH): An MSA is the Mobile Allocation List (MAL) and the Mobile Allocation Index
Offset (MAIO).
TRX index
Channel list
MAIO
MSA
53 54 55 56
([53,54,55,56],2)
53 54 55 56
([53,54,55,56],3)
Therefore, from the point of view of a mobile station, BBH and SFH work in the same way. An MSA will be attached to each
mobile considered during the simulation and the level of interference will be evaluated on this MSA.
Notations and Assumptions
In the following description:
v is a victim transmitter,
MSAS(v) is the set of MSAs (Mobile Station Allocations) associated to v,
The number of MSAS(v) depends on TRX types to be analysed. You may study a given TRX type tt (there will be as many
MSA(v) as TRXs allocated to the subcell (v,tt)) or all the TRX types (the number of MSA(v) will correspond to the
number of TRXs allocated to v).
C v m
- for each victim
Several MSAs, m, are related to a transmitter. Therefore, Atoll calculates the DL C/I ------------- Iv m
transmitter v with MSA m (m MSAS(v)).
131

C v m
C
- or
Atoll considers the most interfered MSA, therefore, the displayed C/I or C/(I+N) are --- = Min ------------- Iv
k Iv m
v
C -
C m -
--------------------------------=
Min
--------------------
, respectively. If the Detailed Results check box is selected, the C/I values for all
k I v m + N Term
I + N Term
tot
tot
v
MSAs are displayed.
i is any potential interfering transmitter (TBC transmitters whose calculation areas intersect the service area of v),
MSAS(i) is the set of MSAs related to potential interferers i,
INT(v) is the set of transmitters that interfere v,
C m is the carrier power level received from v on m,
I m corresponds to the interference received from interfering transmitters i on m,
M Shadowing used in the C/I calculation is based on the C/I standard deviation.
The C/I shadowing margin is applied on the carrier power level. The interference levels
are not changed.
Calculations
The carrier power level is the power received from the victim transmitter at the terminal.
v
C m = P rec m
Term
If the interference conditions are based on C/(I+N), Atoll takes the total noise N tot
into account. The total noise is the sum
of the thermal noise N thermal (-121 dBm by default or user-defined), the terminal noise figure NF
technology downlink noise rise
Term
N tot
= N thermal + NF
Term
v DL
NR inter techno log y
Term
, and the inter-
v DL
+ NR inter techno log y
Interference can be received from interfering transmitters i on co-channel and adjacent channels. Interference may also be
received from the transmitters of another technology.
v
DL
Therefore, I m = I co m + I adj m + I inter techno log y G PC G Div

i
G PC is the average power control gain defined for the interfering transmitter i and G Div is the diversity gain defined for the
considered subcell.
Each interference component is explained below.
Co- and Adjacent Channel Interference:
v
I co m is the interference received at v on m on co-channel, given by:
I co m =
i INT v n MSAS i
v
I adj m
v i
i
p m n P rec n T i n
co
is the interference received at v on m on adjacent channels, given by:
I adj m =
i INT v n MSAS i
P rec n
v i
- T i n
p m n ---------------F
adj
Here, P rec n is the carrier power level received from i on n.

Ti(n) is occupancy of the MSA n:
i
T i n = L traffic n f act n
i
L traffic n is the traffic load defined for the MSA n or i. It can be set to 100% in the coverage prediction properties.
132

AT332_TRR_E0
i
f act n is the activity factor defined for the MSA n of i. If the subcell (i,tt) supports DTX, the value specified in the
coverage prediction properties is used. Otherwise, the activity factor is 1.
BCCH TRXs are always on. Therefore, DTX and traffic loads do not impact the interference
i
from BCCH. In other words, f act n = 1 and L traffic n = 1 for the BCCH TRXs of the
interferers.
v i
p m n is the probability of having a co- or adjacent channel collision between MSAs n and m, depending on the used
frequency hopping mode.
Collision Probability for Non Hopping Mode:

v i
p m n = 1
Collision Probability for BBH and SFH Modes:

MSA m of v can be defined as the pair ([f1,f2,.fn], MAIO) and MSA n of i as the pair ([f1,f2,.fn], MAIO) ( f and
f are channels).
v
An occurence OCCUR f m f' n refers to the event when a channel f of m encounters a channel f of n during
hopping. A collision occurs when f and f are co- or adjacent channels:
v
Collision = OCCUR f m f' n such that f m f' n = 0 or 1

The probability of collision is the ratio of the number of collisions to the number of occurences:
n collision
v i
p m n = ---------------------n occurence
The probibility of collision depends on the correlation between m and n. There can be two cases:
i.
MSAs m and n are correlated

m and n must have identical HSN and synchronisation. The number of occurrences depends on the MAL
length, MAIO, and MAIO.
Example:
Schematic view of hopping sequences
MSA m of v
([34 37 39], MAIO=0)
34 37 39
MSA n of i
([38 36 34], MAIO=2)
38 36 34
Here, the number of occurrences is 3, the number of co-channel collisions is 1, and the number of adjacent
channel collisions is 1. Therefore,
v i
v i
p m n co = 1
--- and p m n adj = 1
--3
3
ii. MSAs m and n are not correlated

m and n do not have identical HSN and synchronisation. The probability of collision is the same for all the
channels.
Example:
Schematic view of hopping sequences
MSA m of v
([34 37 39], MAIO=0)
34 37 39
MSA n of i
([38 36 34], MAIO=2)
38 36 34
Here, the number of occurrences is 9, the number of co-channel collisions is 1, and the number of adjacent
channel collisions is 3. Therefore,
133

v i
v i
p m n co = 1
--- and p m n adj = 1
--9
3
Diversity gain:
v
G Div is the diversity gain defined for the victim subcell.

Two types of diversity modes can be defined. In Tx Diversity, the signal is transmitted as many times that there are
antennas. In, the signal is successively transmitted on the various antennas.
For Tx Diversity mode, the diversity gain is defined as:
v
Tx_Div
G Div = 3dB + Gclutter

Tx_Div
G clutter is the additional transmit diversity gain defined for the clutter class on which is located m.
For Antenna Hopping mode, the diversity gain is defined as:
v
Ant_Div
G Div = G clutter
Ant_Div
G clutter is the antenna hopping gain defined for the clutter class on which is located m.
Inter-technology Downlink Interference:
DL
I inter techno log y is the total inter-technology interference level on m due to transmitters in a linked Atoll document.
The interference from a transmitter Tx in a linked Atoll document is given as:
Tx
DL
I inter techno log y =
P Transmitted ic i
------------------------------------Tx
Tx
L
ICP
ni
ic i is the i
th
ic i f
total
Tx
frequency used by the transmitter Tx within its list of frequencies, P Transmitted ic i is the total
Tx
Tx
transmitted Tx power on ic i , L total are the total losses between the transmitter Tx and the terminal, and ICP ic f is
i
the inter-technology channel protection between the frequencies used by the transmitter Tx and the victim
transmitter v.
In case of frequency hopping, the ICP value is weighted according to the fractional
load.
In the ICP, the frequency gap is based on the defined base frequency for each
technology (e.g., 935 MHz in GSM 900)

Analysis provided in the Interference tab is based on path loss matrices. Therefore, it is possible to display the interference
levels received from TBC transmitters for which path loss matrices have been calculated over their calculation areas.
Atoll displays the following at the terminal:
The carrier power level received from the victim transmitter v on the most interfered MAS m,
Co-channel, adjacent channel, or both co- and adjacent channel interference received from interfering transmitters i
on MAS m (for further information about noise calculation, please refer to Signal to noise calculation: noise calculation
part),
Interferers are sorted in the order of descending carrier power levels.
Neither DTX nor traffic load of TRXs are taken into account to evaluate interference
i
levels. Therefore, we have T i n = L traffic n f act n = 1 .
134
The C/I shadowing margin is applied on the carrier power level. The interference
levels are not changed.

AT332_TRR_E0
3.2.3 Interference-based DL Coverage Predictions

Two interference-based DL coverage predictions are available:
Coverage by C/I Level (DL): Provides a global analysis of the network quality.
Atoll calculates the C/I on each pixel within the service area of studied transmitters, determines the pixels where the
calculated C/I exceeds the defined minimum threshold, and colours these pixels depending on C/I value.
Interfered Zones: Shows the areas a transmitter is interfered.

Atoll calculates the C/I on each pixel within the service area of studied transmitters, determines the pixels where the
calculated C/I is lower than the defined maximum threshold, and colours these pixels depending on colour of the
interfered transmitter.
The coverage conditions in order to determine the service area of each TBC transmitter,
The interference conditions to meet for a pixel to be covered, and
The thermal noise (N = -121 dBm, by default) is used in the calculations if the coverage prediction is based on C/(I+N). This
value can be modified by the user.
3.2.3.1 Service Area Determination

coverage will be displayed. Service areas are determined in the same manner as for signal level-based coverage predictions.
See "DL Service Area Determination" on page 126 for more information.
3.2.3.2 Coverage Area Determination

C
C
For each victim transmitter v, coverage area corresponds to pixels where DL --- or ----------- is between the lower and upper
I v
I+N v
thresholds defined in the coverage prediction properties.
The two options defining the thresholds are explained below.
3.2.3.2.1
Interference Condition Satisfied by At Least One TRX

In this case, the coverage area of a transmitter Txi corresponds to the pixels :
C
Minimum threshold ---
I v
TRX
C
Maximum threshold or Minimum threshold -----------
I+N v
Maximum threshold
TRX
, TRXj is any TRX belonging to Txi.
3.2.3.2.2
Interference Condition Satisfied by The Worst TRX

In this case, the coverage area of a transmitter Txi corresponds to the pixels :
C
Minimum threshold ---
Iv
TRX j
C
Maximum threshold or Minimum threshold -----------
I + N v
Maximum threshold
TRX j
, TRXj is the TRX (belonging to Txi) with the worst C/I or C/(I+N) at the pixel.

3.2.3.3.1
Coverage Resolution
3.2.3.3.2
Display Types
135

C/I Level
Each pixel of the transmitter coverage area is coloured if the calculated DL C/I (or C/(I+N)) level is greater than or equal to the
specified minimum thresholds (pixel colour depends on DL C/I (or C/(I+N)) level). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as transmitter coverage areas. Each layer
shows the different DL C/I levels available in the transmitter coverage area.
Max C/I Level
Atoll compares calculated DL C/I (or C/(I+N)) levels received from transmitters on each pixel of each transmitter coverage area
coverage areas overlap the studied one and chooses the highest value. A pixel of a coverage area is coloured if the DL C/I (or
C/(I+N)) level is greater than or equal to the specified thresholds (the pixel colour depends on the DL C/I (or C/(I+N)) level).
as defined thresholds. Each layer corresponds to an area the highest received DL C/I level exceeds a defined minimum
threshold.
Min C/I Level
Atoll compares DL C/I (or C/(I+N)) levels received from transmitters on each pixel of each transmitter coverage area the
coverage areas overlap the studied one and chooses the lowest value. A pixel of a coverage area is coloured if the DL C/I (or
C/(I+N)) level is greater than or equal to the specified thresholds (the pixel colour depends on the DL C/I (or C/(I+N)) level).
as defined thresholds. Each layer corresponds to an area the lowest received DL C/I level exceeds a defined minimum
threshold.
3.3 GPRS/EDGE Calculations

GPRS/EDGE calculations include coding scheme selection and throughput calculation. Coding schemes may be selected using
ideal link adaptation or without it. Once coding schemes have been selected, throughputs corresponding to these coding
schemes are readily determined from the look-up tables.
The following sections describe the two categories of calculations, i.e., with and without ideal link adaptations. Ideal link
adaptation implies that the selected coding scheme corresponds to the highest available throughput under the given radio
conditions.
GPRS/EDGE calculations may be based on signal levels (C) alone, on C/I, or on C/(I+N). For calculating the noise, either the
noise figure defined for the calculations or that of the selected terminal type is used.
Different GPRS/EDGE configurations may be defined for transmitter and terminals. In this case, Atoll only selects the coding
schemes that are common in the two, and gives priority to the thresholds defined in the terminal configuration. If no terminal
type is defined for the calculation, or if the terminal type does not have any GPRS/EDGE configuration assigned to it, Atoll only
uses the GPRS/EDGE configuration of the transmitter. If the transmitter does not have any GPRS/EDGE configuration assigned
to it, no coding scheme selection and throughput calculation is carried out.
In the following calculations, we assume that:
Txi
P rec TRX is the signal level received from the selected TRX type (tt) or on all the TRXs of Txi on each pixel of the Txi
coverage area,
Txi
P Backoff TRX is the Power Backoff defined for the subcell for 8PSK, 16QAM, or 32QAM modulations,
CS is the set of all available coding schemes,

Reception Threshold CS are the values of reception thresholds for the coding schemes available in the GPRS/EDGE
configuration,
-- I Threshold CS are the values of C/I thresholds for the coding schemes available in the GPRS/EDGE configuration,
C ---------
I + N Threshold CS are the values of C/(I+N) thresholds for the coding schemes available in the GPRS/EDGE
configuration,
The priorities of the coding scheme lists are as follows: DBS > DAS > MCS > CS.
When the calculations are based on C/I and C/(I+N):
136
Atoll calculates the carrier-to-interference ratio for all the GPRS/EDGE TBC transmitters but takes into account all the
TBC transmitters (GSM and GPRS/EDGE) to evaluate the interference.
The reception thresholds given for signal level C are internally converted to C/N thresholds ( N is the thermal noise
defined in the document database at -121 dBm by default) in order to be indexed by C/(I+N) values. C/I thresholds are
also indexed by the C/(I+N) value.

AT332_TRR_E0
The selection of coding schemes is mainly based on the radio conditions mentionned
above. Nevertheless, you can optionally define some specific coding scheme graphs
accoding to a specific hopping mode, mobility type, frequency band and MAL. As an
example, you can model the gain due to longer MALs in coding scheme selection.
For more information on interference (I) calculation, see "DL Carrier-to-Interference Ratio Calculation" on page 131.
3.3.1 Coding Scheme Selection and Throughput Calculation

Without Ideal Link Adaptation
3.3.1.1 Calculations Based on C
Coding Scheme Selection
Atoll selects a coding scheme, cs, from among the coding schemes available in the GPRS/EDGE configuration, such that:
For each TRX type, tt, cs = Lowest CS
Txi
Txi
P rec TRX P Backoff TRX Reception Threshold CS
The selected coding scheme, cs, is the coding scheme with the lowest coding scheme number from the lowest priority coding
scheme list.
Throughput Calculation
Once the coding scheme cs is selected, Atoll reads the corresponding throughput value for the received signal level from the
Throughput=f(C) graph associated with cs.
3.3.1.2 Calculations Based on C/I

Atoll selects two coding schemes from among the coding schemes available in the GPRS/EDGE configuration, such that:
For each TRX type, tt, cs C = Lowest CS
And, cs C I
= Lowest CS
Txi
Txi
Txi
Txi
P
TRX P
TRX
rec
Backoff
C
------------------------------------------------------------------------- --- Threshold

I
CS
I
csC is the coding scheme determined from the signal level, and csC/I is the coding scheme determined from the C/I level. Both
coding schemes are the coding schemes with the lowest coding scheme number from the lowest priority coding scheme list.
The selected coding scheme, cs, is the coding scheme with the lower coding scheme number among csC and csC/I:
cs = Min cs C cs C I .
Throughput Calculation Based on the Worst Case Between C and C/I
For the coding scheme csC determined above, a throughput value, TPC, corresponding to the signal level is determined from
the TP = f(C) graph.
For the coding scheme csC/I determined above, a throughput value, TPC/I, corresponding to the C/I is determined from the TP
= f(C/I) graph.
The resulting throughput TP is the lower of the two values, TPC and TPC/I: TP = Min TP C TP C I .
3.3.1.3 Calculations Based on C/(I+N)

137

For each TRX type, tt, cs C N = Lowest CS
And, cs C I + N
= Lowest CS
Txi
Txi
P
TRX P
TRX
rec
Backoff
C
------------------------------------------------------------------------- ----------- Threshold
I + N
CS
N
Txi
Txi
P rec TRX P Backoff TRX
C
------------------------------------------------------------------------- --------- I + N Threshold CS
I+N
csC/N is the coding scheme determined from the C/N, and csC/(I+N) is the coding scheme determined from the C/(I+N) level.
Both coding schemes are the coding schemes with the lowest coding scheme numbers from the lowest priority coding scheme
list.
The selected coding scheme, cs, is the coding scheme with the higher coding scheme number among csC/N and csC/(I+N):
cs = Max cs C N cs C I + N .
Throughput Calculation Based on Interpolation Between C/N and C/(I+N)
For the coding scheme csC/N determined above, the TP = f(C) graph is internally converted to TP = f(C/N) graph. A throughput
value, TPC/N, corresponding to the C/(I+N) is determined from the TP = f(C/N) graph.
For the coding scheme csC/(I+N) determined above, the TP = f(C/I) graph is internally converted to TP = f(C/(I+N)) graph. A
throughput value, TPC/(I+N), corresponding to the C/(I+N) is determined from the TP = f(C/(I+N)) graph.
The final throughput is computed by interpolating between the throughput values obtained from these two graphs. The
throughput interpolation method consists in interpolating TPC/N and TPC/(I+N) according to the respective weights of I and N
values.
The resulting throughput TP is given by: TP = TP C N + 1 TP C I + N
pN - , pN is the thermal noise power (value in Watts), and p(I+N) is the interferences + thermal noise power (value
= ------------------pI + N
in Watts).
3.3.2 Coding Scheme Selection and Throughput Calculation With

Ideal Link Adaptation
3.3.2.1 Calculations Based on C
Throughput Calculation
For the received signal level, and coding schemes whose reception thresholds are lower than the received signal level, Atoll
determines the highest throughput from the TP=f C graphs available in the GPRS/EDGE configuration.
Txi
Txi
TP C = Highest TP=f C = P rec TRX P Backoff TRX CS
Txi
Txi

The selected coding scheme, cs, is the one corresponding to the highest throughput calculated above.
If there are more than one coding schemes providing the highest throughput at the pixel, the selected coding scheme, cs, is
the one with the lowest coding scheme number from the lowest priority coding scheme list.
3.3.2.2 Calculations Based on C/I

Throughput Calculation Based on Worst Case Between C and C/I
For the received signal level, and coding schemes whose reception thresholds are lower than the received signal level, Atoll
determines the highest throughput from the TP=f C graphs available in the GPRS/EDGE configuration.
138

AT332_TRR_E0
Txi
Txi
TP C = Highest TP=f C = P rec TRX P Backoff TRX CS
Txi
Txi
TRX P
TRX Reception Threshold
rec
Backoff
CS
For the received C/I, and coding schemes whose C/I thresholds are lower than the received C/I, Atoll determines the highest
throughput from the TP=f C I graphs available in the GPRS/EDGE configuration.
TP C I
Txi
Txi
= Highest TP=f C I = ---------------------------------------------------------------- CS

I
Txi
Txi
------------------------------------------------------------------------- --- Threshold

I
CS
I
The resulting throughput TP is the lower of the two values, TPC and TPC/I.
TP = Min TP C TP C I
The selected coding scheme, cs, is the one corresponding to the lower of the two highest throughputs calculated above.
If there are more than one coding schemes providing the highest throughputs at the pixel, the selected coding scheme, cs, is
the one with the lowest coding scheme number from the lowest priority coding scheme list.

Throughput Calculation Based on Interpolation Between C/N and C/(I+N)
Atoll internally converts the TP = f(C) graphs into TP = f(C/N) graphs. For the received C/(I+N), and coding schemes whose C/
(I+N) thresholds are lower than the received C/(I+N), Atoll determines the highest throughput from the TP = f(C/N) graphs
available in the GPRS/EDGE configuration.
TP C N
Txi
Txi
C
= Highest TP=f --- = ---------------------------------------------------------------- CS

I+N
N
Txi
Txi
C
------------------------------------------------------------------------- ----------- Threshold
I + N
CS
I+N
Atoll internally converts the TP = f(C/I) graphs into TP = f(C/(I+N)) graphs. For the received C/(I+N), and coding schemes whose
C/(I+N) thresholds are lower than the received C/(I+N), Atoll determines the highest throughput from the TP = f(C/(I+N))
graphs available in the GPRS/EDGE configuration.
TP C I + N
Txi
Txi
C
= Highest TP=f ----------- = ---------------------------------------------------------------- CS

I+N
I+N
Txi
Txi
C
------------------------------------------------------------------------- ----------- Threshold
I + N
CS
I+N
values.
pN = ------------------, pN is the thermal noise power (value in Watts), and p(I+N) is the interferences + thermal noise power (value
pI + N
in Watts).
The selected coding scheme, cs, is the one corresponding to the higher of the two highest throughputs calculated above.
If there are more than one coding schemes providing the highest throughputs at the pixel, the selected coding scheme, cs, is
the one with the highest coding scheme number from the highest priority coding scheme list.
3.3.3 Application Throughput Calculation

Application throughput is calculated from the effective RLC throughput as follows:
SF- TP
TP Application = TP RLC MAC -------Offset
100
139

TP RLC MAC is the effective RLC throughput, and TP Offset and SF are the throughput offset (kbps) and the throughput scaling
factor (%) defined for the selected service.
3.3.4 BLER Calculation

Block error rate is calculated as follows:
TP
--------------- If TP TP MAX
BLER = TP MAX
0 If TP TP MAX
TP is the throughput per timeslot calculated for a pixel and TPMAX is the maximum throughput per timeslot read from the
GPRS/EDGE configuration used for the calculations.
3.3.5 GPRS/EDGE Coverage Predictions

Two GPRS/EDGE coverage predictions are available:
Coverage by GPRS/EDGE Coding Scheme: Shows the areas various coding schemes are available.
Packet Throughput and Quality Analysis: Shows the throughputs corresponding to the coding schemes available.
The interference conditions to meet for a pixel to be covered, and

3.3.5.1.1
Each transmitter, Txi, is GPRS/EDGE-capable.
All Servers
Txi
SubcellReceptionThreshold P rec tt
3.3.5.1.2

Txi
Txi
Txj
ji
140
the highest.

AT332_TRR_E0
3.3.5.1.3

Txi
SubcellReceptionThreshold Prec tt
nd
Txi
Txj
ji
3.3.5.1.4

Txi
Txi
Txj
ji
3.3.5.1.5
the highest.
Txi
nd
Txi
Txj
ji
3.3.5.1.6

Txi
Txi
Txj
ji
Txi
the highest.
141

3.3.5.1.7

Txi
Txi
Txj
ji
Txi
3.3.5.1.8
the highest.
Best Idle Mode Reselection Criterion (C2)

Such type of coverage is useful:
Txi

Txi
SubcellReceptionThreshold P rec BCCH

And C2
Txi
Txj
BCCH = Best C2 BCCH
j
integer.

3.3.5.2.1
Coverage Resolution
142

AT332_TRR_E0
3.3.5.2.2
Display Types
It is possible to display the coverage predictions with colours depending on criteria such as:
Coverage by GPRS/EDGE Coding Scheme: Coding Schemes
Only the pixels with a coding scheme assigned are coloured. The pixel colour depends on the assigned coding scheme.
Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many layers
as transmitter coverage areas. Each layer shows the coding schemes available in the transmitter coverage area.
Coverage by GPRS/EDGE Coding Scheme: Max Coding Schemes
On each pixel, Atoll chooses the highest coding scheme available from the TRXs of different transmitters covering that pixel.
Only the pixels with a coding scheme assigned are coloured. The pixel colour depends on the assigned coding scheme.
as possible coding schemes. Each layer shows the areas a given coding scheme can be used.
Packet Throughput and Quality Analysis: Effective RLC Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated effective RLC throughput per timeslot from any transmitter covering
that pixel exceeds the defined minimum threshold. The pixel colour depends on the effective RLC throughput per timeslot.
as transmitter coverage areas and throughput display thresholds. Each layer shows the effective RLC throughput that a
transmitter can provide on one timeslot.
Packet Throughput and Quality Analysis: Max Effective RLC Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated highest effective RLC throughput per timeslot from any transmitter
covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the highest effective RLC throughput
per timeslot. Coverage consists of several independent layers whose visibility in the map window can be managed. There are
as many layers as throughput display thresholds. Each layer shows the best effective RLC throughput that any transmitter can
provide on one timeslot.
Packet Throughput and Quality Analysis: Average Effective RLC Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated average effective RLC throughput per timeslot from all the
transmitters covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the average effective
RLC throughput per timeslot. Coverage consists of several independent layers whose visibility in the map window can be
managed. There are as many layers as throughput display thresholds. Each layer shows the average effective RLC throughput
that all the transmitters can provide on one timeslot.
Packet Throughput and Quality Analysis: Application Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated application throughput per timeslot from any transmitter covering
that pixel exceeds the defined minimum threshold. The pixel colour depends on the application throughput per timeslot.
as transmitter coverage areas and throughput display thresholds. Each layer shows the application throughput that a
transmitter can provide on one timeslot.
Packet Throughput and Quality Analysis: Best Application Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated highest application throughput per timeslot from any transmitter
covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the highest application throughput
as many layers as throughput display thresholds. Each layer shows the best application throughput that any transmitter can
provide on one timeslot.
Packet Throughput and Quality Analysis: Average Application Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated average application throughput per timeslot from all the transmitters
covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the average application throughput
as many layers as throughput display thresholds. Each layer shows the average application throughput that all the
transmitters can provide on one timeslot.
Packet Throughput and Quality Analysis: Effective RLC Throughput (kbps)
A pixel of the coverage area is coloured if the calculated effective RLC throughput from any transmitter covering that pixel
exceeds the defined minimum threshold. The pixel colour depends on the effective RLC for all the timeslots supported by the
143

selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL timeslots is the
minimum between the number of DL timeslots defined in the selected terminal and service. Coverage consists of several
independent layers whose visibility in the map window can be managed. There are as many layers as transmitter coverage
areas and throughput display thresholds. Each layer shows the application throughput that a transmitter can provide on all
available timeslots in the terminal.
Packet Throughput and Quality Analysis: Max Effective RLC Throughput (kbps)
A pixel of the coverage area is coloured if the calculated highest effective RLC throughput from any transmitter covering that
pixel exceeds the defined minimum threshold. The pixel colour depends on the highest effective RLC throughput for all the
timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number
of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. Coverage
consists of several independent layers whose visibility in the map window can be managed. There are as many layers as
throughput display thresholds. Each layer shows the highest effective RLC throughput that any transmitter can provide on all
Packet Throughput and Quality Analysis: Average Effective RLC Throughput (kbps)
A pixel of the coverage area is coloured if the calculated average effective RLC throughput from all the transmitters covering
that pixel exceeds the defined minimum threshold. The pixel colour depends on the average effective RLC throughput for all
the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The
number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service.
as throughput display thresholds. Each layer shows the average effective RLC throughput that all the transmitters can provide
on all available timeslots in the terminal.
Packet Throughput and Quality Analysis: Application Throughput (kbps)
A pixel of the coverage area is coloured if the calculated application throughput from any transmitter covering that pixel
exceeds the defined minimum threshold. The pixel colour depends on the application throughput for all the timeslots
supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL
timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. Coverage consists
of several independent layers whose visibility in the map window can be managed. There are as many layers as transmitter
coverage areas and throughput display thresholds. Each layer shows the application throughput that a transmitter can provide
on all available timeslots in the terminal.
Packet Throughput and Quality Analysis: Max Application Throughput (kbps)
A pixel of the coverage area is coloured if the calculated highest application throughput from any transmitter covering that
pixel exceeds the defined minimum threshold. The pixel colour depends on the highest application throughput for all the
throughput display thresholds. Each layer shows the highest application throughput that any transmitter can provide on all
Packet Throughput and Quality Analysis: Average Application Throughput (kbps)
A pixel of the coverage area is coloured if the calculated average application throughput from all the transmitters covering
that pixel exceeds the defined minimum threshold. The pixel colour depends on the average application throughput for all the
throughput display thresholds. Each layer shows the average application throughput that all the transmitters can provide on
all available timeslots in the terminal.
Packet Throughput and Quality Analysis: Throughput per User (kbps)
A pixel of the coverage area is coloured if the calculated throughput per user from any transmitter covering that pixel exceeds
the defined minimum threshold. The pixel colour depends on the throughput per user for all the timeslots supported by the
selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL timeslots is the
minimum between the number of DL timeslots defined in the selected terminal and service. The throughput per user is
calculated by applying the throughput reduction factor, determined using the selected dimensioning model, to the application
throughput. Coverage consists of several independent layers whose visibility in the map window can be managed. There are
as many layers as transmitter coverage areas and throughput display thresholds. Each layer shows the throughput per user
that a transmitter can provide on all available timeslots in the terminal.
144

AT332_TRR_E0
Packet Throughput and Quality Analysis: Max Throughput per User (kbps)
A pixel of the coverage area is coloured if the calculated highest throughput per user from any transmitter covering that pixel
exceeds the defined minimum threshold. The pixel colour depends on the highest throughput per user for all the timeslots
supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL
timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. The throughput
per user is calculated by applying the throughput reduction factor, determined using the selected dimensioning model, to the
application throughput. Coverage consists of several independent layers whose visibility in the map window can be managed.
There are as many layers as throughput display thresholds. Each layer shows the highest throughput per user that any
transmitter can provide on all available timeslots in the terminal.
Packet Throughput and Quality Analysis: Average Throughput per User (kbps)
A pixel of the coverage area is coloured if the calculated average throughput per user from all the transmitters covering that
pixel exceeds the defined minimum threshold. The pixel colour depends on the average throughput per user for all the
of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. The
throughput per user is calculated by applying the throughput reduction factor, determined using the selected dimensioning
model, to the application throughput. Coverage consists of several independent layers whose visibility in the map window can
be managed. There are as many layers as throughput display thresholds. Each layer shows the average throughput per user
that all the transmitters can provide on all available timeslots in the terminal.
Packet Throughput and Quality Analysis: BLER (%)
A pixel of the coverage area is coloured if the calculated BLER from any transmitter exceeds the defined minimum threshold.
The pixel colour depends on the BLER. Coverage consists of several independent layers whose visibility in the map window
can be managed. There are as many layers as transmitter coverage areas and BLER display thresholds. Each layer shows the
BLERs that the covered pixels experience on one timeslot.
Packet Throughput and Quality Analysis: Max BLER (%)
A pixel of the coverage area is coloured if the calculated highest BLER from all the transmitters exceeds the defined minimum
threshold. The pixel colour depends on the BLER. Coverage consists of several independent layers whose visibility in the map
window can be managed. There are as many layers as BLER display thresholds. Each layer shows the BLER that the covered
pixels experience on one timeslot.
3.4 Codec Mode Selection and CQI Calculations

Atoll supports FR, HR, EFR, and AMR codec modes. A codec configuration contains codec mode adaptation thresholds and
quality graphs for circuit quality indicators. Atoll has the following circuit quality indicators included by default:
FER or Frame Erasure Rate: The number of frames in error divided by the total number of frames. These frames are
usually discarded, in which case this can be called the Frame Erasure Rate.
BER or Bit Error Rate: BER is a measurement of the raw bit error rate in reception before the decoding process begins.
Any factor that impacts the decoding performance, such as frequency hopping, will impact the correlation between
BER and FER, or the perceived end-user voice quality.
MOS or Mean Opinion Score: Voice quality can be quantified using mean opinion score (MOS). MOS values can only
be measured in a test laboratory environment. MOS values range from 1 (bad) to 5 (excellent). Different voice codecs
have slightly different FER to MOS correlation since the smaller the voice codec bit rate is, the more sensitive it
becomes to frame erasures.
The default codec configurations in Atoll include default FER, BER, and MOS quality graphs with respect to the carrier to
interference ratio, and codec mode adaptation thresholds (calculated from the FER vs. C/I graphs for all codec modes at 5 %
FER).
145

Figure 3.1: FER vs. C/I Graphs
Figure 3.2: BER vs. C/I Graphs
Figure 3.3: MOS vs. C/I Graphs
146

AT332_TRR_E0
The graphs are based on:

[1] T. Halonen, J. Romero, J. Melero; GSM, GPRS and EDGE performance Evolution
towards 3G/UMTS, John Wiley and Sons Ltd.
[2] J. Wigard, P. Mogensen; A simple mapping from C/I to FER and BER for a GSM type of
air interface.
[3] 3GPP Specifications TR 26.975 V6.0.0; Performance characterization of the Adaptive
Multi-Rate (AMR) speech codec (Release 6)
3.4.1 Circuit Quality Indicator Calculations

Circuit quality indicator calculations include codec mode selection and CQI calculation. Codec modes may be selected using
ideal link adaptation or without it. Once codec modes have been selected, CQI corresponding to these codec modes are
determined from the look-up tables.
The following sections describe the two categories of calculations, i.e., with and without ideal link adaptations. Ideal link
adaptation implies that the selected codec mode corresponds to the best value of the reference CQI under the given radio
conditions. Without ideal link adaptation, the codec mode is selected based on the codec adaptation thresholds.
CQI calculations may be based on C/N or on C/(I+N). For calculating the noise, either the noise figure defined for the
calculations or that of the selected terminal type is used.
Different codec configurations may be defined for transmitter and terminals. In this case, Atoll only selects the coding
schemes that are common in the two, and gives priority to the thresholds defined in the terminal configuration. If no terminal
type is defined for the calculation, or if the terminal type does not have any codec configuration assigned to it, Atoll only uses
the codec configuration of the transmitter. If the transmitter does not have any codec configuration assigned to it, no codec
mode selection and CQI calculation is carried out.
If more than one codec modes satisfy the C/N or C/I conditions, Atoll selects the higher priority codec mode.
Txi
P rec TRX is the signal level received from the selected TRX type (tt) or on all the TRXs of Txi on each pixel of the Txi
coverage area,
CM is the set of all available codec modes,
Adaptation Threshold CM are the values of adaptation thresholds for the codec modes available in the codec
configuration,
The computed noise N is compared to the codec configuration reference noise N Ref . If the values are the same, the
defined graphs are used as is, otherwise the graphs are downshifted by the difference N N Ref .
When the calculations are based on C/(I+N):
Atoll calculates the carrier-to-interference ratio for all the TBC transmitters with codec configurations assigned, but
takes into account all the TBC transmitters (with and without codec configurations) to evaluate the interference.
The selection of codec modes is mainly based on the radio conditions mentionned above.
Nevertheless, you can optionally define some specific codec mode graphs accoding to a
specific hopping mode, mobility type, frequency band and MAL. As an example, you can
model the gain due to longer MALs in codec mode selection.
For more information on interference (I) calculation, see "DL Carrier-to-Interference Ratio Calculation" on page 131.
Ideal link adaptation for circuit quality indicator studies is defined at the codec configuration level. If the ideal link adaptation
option is checked, Atoll will select the codec mode, for the transmitter under study, according to the codec quality graphs (CQI
= f(C/N) and CQI = f(C/I)) related to the defined reference CQI, which may be different from the CQI being calculated.
Otherwise, Atoll will use the adaptation thresholds defined in the Adaptation Thresholds tab to determine the codec mode
to be used in the studies.
3.4.2 CQI Calculation Without Ideal Link Adaptation

3.4.2.1 Calculations Based on C/N
Atoll selects the highest priority codec mode, cm, from among the codec modes available in the codec configuration:
147

For each TRX type, tt, cm = Highest Priority CM
Txi
P
TRX
rec
--------------------------- Adaptation Threshold
CM
N
Txi
P rec TRX
For ------------------------ , Atoll determines the CQI from the CQI=f(C/N) graph associated to the selected codec mode, cm.
N

Txi
P rec TRX
--------------------------- Adaptation Threshold
CM
I+N
Txi
P rec TRX
- , Atoll determines the CQI from the CQI=f(C/I) graph associated to the selected codec mode, cm.
For ----------------------I+N
3.4.3 CQI Calculation With Ideal Link Adaptation

3.4.3.1 Calculations Based on C/N
Ideal link adaptation is used by a codec configuration according to a defined reference CQI (MOS by default).
Atoll calculates signal level received from Txi on each pixel of Txi coverage area and converts it into C/N values as described
earlier. Then, Atoll filters all the codec modes that satisfy the C/N criterion (defined by the CQI = f(C/N) graphs for the
reference CQI) and are common between the transmitter and the terminal type codec configuration.
The selected codec mode among these filtered codec modes will be,
Or, cm = Highest Priority CM
, for MOS
Txi
P
TRX
C rec
CQI Ref = Highest CQI=f ---= ---------------------------

N
N tot
, for BER and FER
Txi
TRX
C rec
CQI Ref = Lowest CQI=f ---= ---------------------------
N
N tot
, cm is the codec mode with the highest priority among the set of codec modes CM for which the reference CQI gives the
Txi
P rec TRX
-.
highest or the lowest value at the received C/N level, ----------------------N tot
If more than one codec mode graphs give the same value for reference CQI, then Atoll selects the codec mode with the highest
priority.
From the CQI = f(C/N) graph associated to the selected codec mode cm, Atoll evaluates the CQI for which the study was
Txi
P rec TRX
performed corresponding to ------------------------ for the selected codec mode.
N tot

Ideal link adaptation is used by a codec configuration according to a defined reference CQI (MOS by default).
Atoll calculates the C/I level received from the transmitter on each pixel of Txi coverage area, for each TRX and converts it into
C/(I+N). Then, Atoll filters all the codec modes that satisfy the C/(I+N) criteria (defined by the CQI = f(C/I) graphs for the
reference CQI) and are common between the transmitter and the terminal type codec configuration.
The selected codec mode among these filtered codec modes will be,
148

AT332_TRR_E0
Or, cm = Highest Priority CM
, for MOS
P Txi TRX
rec
C
CQI
= Highest CQI=f --- = ---------------------------
Ref
I I+N
tot
, for BER and FER
Txi
P
TRX
C
rec
CQI Ref = Lowest CQI=f --- = ---------------------------
I I+N
tot
, cm is the codec mode with the highest priority among the set of codec modes CM for which the reference CQI gives the
Txi
P rec TRX
-.
highest or the lowest value at the received C/(I+N) level, ----------------------I + N tot
If more than one codec mode graphs give the same value for reference CQI, then Atoll selects the codec mode with the highest
priority.
From the CQI = f(C/I) graph associated to the selected codec mode cm (indexed with the C/(I+N) values), Atoll evaluates the
Txi
P rec TRX
- for the selected codec mode.
CQI for which the study was performed corresponding to ----------------------I + N tot
3.4.4 Circuit Quality Indicators Coverage Predictions

The Circuit Quality Indicators coverage predictions show the areas BER, FER, and MOS values in the transmitter coverage
areas.
The interference and quality indicator conditions to meet for a pixel to be covered, and

We can distinguish seven cases as below. Let us assume that:
3.4.4.1.1
Each transmitter, Txi, has a codec configuration assigned.
All Servers
Txi
SubcellReceptionThreshold Prec BCCH
3.4.4.1.2

Txi

Txi
Txj
ji
149

3.4.4.1.3
the highest.

Txi

nd
Txi
Txj
ji
3.4.4.1.4

Txi

Txi
Txj
ji
3.4.4.1.5
the highest.
Txi

nd
Txi
Txj
ji
3.4.4.1.6

Txi

Txi
Txj
ji
Txi
And the received P rec BCCH exceeds the reception threshold defined per HCS layer.
150

AT332_TRR_E0
3.4.4.1.7
the highest.

Txi

Txi
Txj
ji
Txi
the highest.

3.4.4.2.1
Coverage Resolution
3.4.4.2.2
Display Types
It is possible to display the coverage predictions with colours depending on criteria such as:
BER
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the BER value. Coverage consists of
several independent layers whose visibility in the map window can be managed. There are as many layers as transmitter
coverage areas and BER display thresholds. Each layer shows the BER in the transmitter coverage area.
FER
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the FER value. Coverage consists of
coverage areas and FER display thresholds. Each layer shows the FER in the transmitter coverage area.
MOS
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the MOS value. Coverage consists of
coverage areas and MOS display thresholds. Each layer shows the MOS in the transmitter coverage area.
151

Max BER
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the highest BER value among the BER
values for all the transmitters covering the pixel. Coverage consists of several independent layers whose visibility in the map
window can be managed. There are as many layers as BER display thresholds. Each layer shows the BER value.
Max FER
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the highest FER value among the FER
window can be managed. There are as many layers as FER display thresholds. Each layer shows the FER value.
Max MOS
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the highest MOS value among the MOS
window can be managed. There are as many layers as MOS display thresholds. Each layer shows the MOS value.
3.5 UL Coverage Predictions

For each TBC transmitter, Txi, Atoll calculates the selected parameter at each Txi inside its calculation area. Results are shown
on each pixel, each pixel acting as a transmitting terminal. Hence, transmitters are here (non-interfering) receivers.
The coverage conditions in order to determine the DL service area of each TBC transmitter, and
Two interfaced predictions are available:
One prediction which shows on each pixel UL losses or UL signal levels

One prediction which shows on each pixel UL C/I levels.
Additional studies such as codec modes and coding schemes predictions are used during simulations but are not graphically
available.
3.5.1 DL Service Area Determination

Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialog box to determine the service
areas of the TBC transmitters.
3.5.1.1 All Servers

Txi

Txi
Txi
Txi
3.5.1.2 Best Signal Level and an Overlap Margin

Txi

Txi
Txi
Txi
152

AT332_TRR_E0
Txi
Txj
ji
the highest.
3.5.1.3 Second Best Signal Level and an Overlap Margin

The service area of Txi corresponds to the pixels:
Txi

Txi
Txi
Txi
nd
Txi
Txj
ji
3.5.1.4 Best Signal Level per HCS Layer and an Overlap Margin
Txi

Txi
Txi
Txi
Txi
Txj
ji
the highest.
3.5.1.5 Second Best Signal Level per HCS Layer and an Overlap Margin
Txi
153

Txi
Txi
Txi
nd
Txi
Txj
ji
3.5.1.6 HCS Servers and an Overlap Margin

Txi

Txi
Txi
Txi
Txi
Txj
ji
Txi
the highest.
3.5.1.7 Highest Priority HCS Server and an Overlap Margin

Txi

Txi
Txi
Txi
Txi
Txj
ji
Txi
154

AT332_TRR_E0
the highest.
3.5.1.8 Best Idle Mode Reselection Criterion (C2)

Such type of coverage is useful :
Txi

Txi
MinimumThreshold P rec BCCH MaximumThreshold

Txi
Txi
Txi
And C2
Txi
Txj
BCCH = Best C2 BCCH
j
integer.
3.5.2 Coverage by UL Signal Level

3.5.2.1 Coverage Resolution
3.5.2.2 Display Types

UL signal levels and UL losses calculations are explained in "UL Signal Level" on page 125. It is possible to display the coverage
by UL signal level with colours depending on any transmitter attribute or other criteria such as:
3.5.2.2.1
UL Signal Level (in dBm, dBV, dBV/m)

Atoll calculates the signal level received at each transmitter on its service area from surrounding pixels. A pixel of a service
area is coloured if the UL signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on
signal level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as
many layers as transmitter service areas. Each layer shows the different UL signal levels at the transmitter on its service area.
3.5.2.2.2
Best UL Signal Level (in dBm, dBV, dBV/m)

Atoll calculates the signal level received at each transmitter on its service area from surrounding pixels. When other service
areas overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the UL signal level is
155

greater than or equal to the defined thresholds (the pixel colour depends on the signal level). Coverage consists of several
independent layers whose visibility in the workspace can be managed. There are as many layers as transmitter service areas.
Each layer shows the different UL signal levels at the transmitter on its service area.
3.5.2.2.3
UL Total Losses (dB)

Atoll calculates total losses from the terminal at each transmitter on its service area. A pixel of a service area is coloured if UL
total losses are greater than or equal to the defined minimum thresholds (pixel colour depends on UL total losses). Coverage
areas. Each layer shows the different UL total losses at the transmitter on its service area.
3.5.2.2.4
Minimum UL Total Losses (dB)

Atoll calculates total losses from the transmitter on each pixel of each transmitter service area. When other service areas
overlap the studied one, Atoll chooses the lowest value. A pixel of a service area is coloured if UL total losses are greater than
or equal to the defined minimum thresholds (pixel colour depends on total losses). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as service areas. Each layer shows the
different total losses levels in the transmitter service area.
3.5.3 Coverage by UL C/I

An UL C/I coverage predictions is available. It provides the UL C/I level at the transmitter level caused by surrounding uplink
traffic.
3.5.3.1 Coverage Resolution

3.5.3.2 UL C/I Evaluation

The UL C/I level can be computed as follows, for a given MSA
C
--I
MSA
Term
Tx
= P rec N tot
UL
,
Tx
Tx
MSA DL
N tot = N thermal + NF
P rec
N thermal is the thermal noise (-121 dBm by default or user-defined)
NF
MSA DL
NRIntra techno log y
Term
Tx
+ NR Intra techno log y is the UL total noise at transmitter on the considered MSA
is the received signal level at the transmitter,
is the transmitter noise figure

is the intra-technology UL noise rise at the considered MSA. Since UL noise rise are defined per
TRX, Atoll takes the TRX UL noise rise in case of non-hopping or extracts a mean noise rise from the several TRXs
composing the MSA in case of Base Band Hopping or Synthesized Frequency Hopping.
MSA
For a given transmitter having several MSAs, all possible C
are displayed in case the detailed results box is selected. If
--I UL
not, the worst results (the min C/I per transmitter) are retained.
3.5.3.3 Coverage Area Determination

For each MSA, coverage area corresponds to pixels where C
--I
coverage prediction properties.
MSA
is between the lower and upper thresholds defined in the
UL
3.5.3.4 Display Types

3.5.3.4.1
C/I Level
Each pixel of the transmitter coverage area is coloured if the calculated UL C/I level is greater than or equal to the specified
minimum thresholds (pixel colour depends on UL C/I level). Coverage consists of several independent layers whose visibility
156

AT332_TRR_E0
in the workspace can be managed. There are as many layers as transmitter service areas. Each layer shows the different UL
C/I levels available in the transmitter coverage area.
3.5.3.4.2
Max C/I Level

Atoll compares calculated UL C/I levels received from transmitters on each pixel of each transmitter coverage area coverage
areas overlap the studied one and chooses the highest value. A pixel of a coverage area is coloured if the UL C/I level is greater
than or equal to the specified thresholds (the pixel colour depends on the UL C/I level). Coverage consists of several
layer corresponds to an area the highest received UL C/I level exceeds a defined minimum threshold.
3.5.3.4.3
Min C/I Level

Atoll compares UL C/I levels received from transmitters on each pixel of each transmitter coverage area the coverage areas
overlap the studied one and chooses the lowest value. A pixel of a coverage area is coloured if the UL C/I level is greater than
or equal to the specified thresholds (the pixel colour depends on the UL C/I level). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer
corresponds to an area the lowest received UL C/I level exceeds a defined minimum threshold.
3.5.4 Coverage by UL Coding Schemes

An UL Coding Scheme coverage prediction is implemented in order to be used in simulations. The prediction itself does not
have any interface. for the simulations, settings are hard coded and are described hereafter.
These calculations are based on C/(I+N). Coding schemes are selected without using ideal link adaptation.
Different GPRS/EDGE configurations may be defined for transmitter and terminals. In this case, Atoll only selects the coding
schemes that are common in the two, and gives priority to the thresholds defined in the transmitter configuration. If no
terminal type is defined for the calculation, or if the terminal type does not have any GPRS/EDGE configuration assigned to it,
Atoll only uses the GPRS/EDGE configuration of the transmitter. If the transmitter does not have any GPRS/EDGE
configuration assigned to it, no coding scheme selection and throughput calculation is carried out.
Txi
P rec TRX is the DL signal level received from the BCCH of Txi on each pixel of the Txi coverage area,
P rec
CS is the set of all available coding schemes,

Reception Threshold CS are the values of reception thresholds for the coding schemes available in the GPRS/EDGE
Term
is the UL the signal level received at each transmitter on its service area from surrounding pixels
configuration,
C ---------Threshold
I + N
CS are the values of C/(I+N) thresholds for the coding schemes available in the GPRS/EDGE
configuration,
The priorities of the coding scheme lists are as follows: DBS > DAS > MCS > CS.
Since the calculations are based on C/I and C/(I+N):
Atoll calculates the UL C/I to all the GPRS/EDGE TBC transmitters.

The reception thresholds given for signal level C are internally converted to C/N thresholds (N is the thermal noise
defined in the document database at -121 dBm by default) in order to be indexed by C/(I+N) values. C/I thresholds are
also indexed by the C/(I+N) value.
For more information on UL C/I calculation, see "Coverage by UL C/I" on page 156.

Atoll uses hard-coded parameters for simulations. In that case, the DL service area is based on the option "HCS servers" with
a margin of 4 dB.
Txi
Txi
Txj
And P rec BCCH Best P rec BCCH 4dB
ji
Txi
157

3.5.4.2 Coding Scheme Selection

For each MSA , cs C N = Lowest CS
And, cs C I + N = Lowest CS
Term
P rec
----------------- ----------- Threshold

I + N
CS
N
Term
P rec
----------------- ----------- Threshold

I + N
CS
I+N
csC/N is the coding scheme determined from the C/N, and csC/(I+N) is the coding scheme determined from the C/(I+N) level.
Both coding schemes are the coding schemes with the lowest coding scheme numbers from the lowest priority coding scheme
list.
The selected coding scheme, cs, is the coding scheme with the higher coding scheme number among csC/N and csC/(I+N):
cs = Max cs C N cs C I + N .
3.5.4.3 Throughput Calculation

For the coding scheme csC/N determined above, the TP = f(C) graph is internally converted to TP = f(C/N) graph. A throughput
value, TPC/N, corresponding to the C/(I+N) is determined from the TP = f(C/N) graph.
For the coding scheme csC/(I+N) determined above, the TP = f(C/I) graph is internally converted to TP = f(C/(I+N)) graph. A
throughput value, TPC/(I+N), corresponding to the C/(I+N) is determined from the TP = f(C/(I+N)) graph.
values.
pN = ------------------, pN is the thermal noise power (value in Watts), and p(I+N) is the interferences + thermal noise power (value
pI + N
in Watts).
3.5.5 Coverage by UL Codec Modes

An UL Codec Mode coverage prediction is implemented in order to be used in simulations. The prediction itself does not have
any interface. for the simulations, settings are hard coded and are described hereafter.
Circuit quality indicator calculations include codec mode selection and CQI calculation. Codec modes are selected according
to C/(I+N) quality without using ideal link adaptation. Once codec modes have been selected, CQI and number of used
timeslots (0.5 in case of HR) corresponding to these codec modes are determined from the look-up tables.
Different codec configurations may be defined for transmitter and terminals. In this case, Atoll only selects the coding
schemes that are common in the two, and gives priority to the thresholds defined in the transmitter configuration. If no
terminal type is defined for the calculation, or if the terminal type does not have any codec configuration assigned to it, Atoll
only uses the codec configuration of the transmitter. If the transmitter does not have any codec configuration assigned to it,
no codec mode selection and CQI calculation is carried out.
If more than one codec modes satisfy the quality conditions, Atoll selects the higher priority codec mode.
Txi
P rec TRX is the DL signal level received from the BCCH of Txi on each pixel of the Txi coverage area,
P rec
CM is the set of all available codec modes,

Adaptation Threshold CM are the values of adaptation thresholds for the codec modes available in the codec
Term
is the UL the signal level received at each transmitter on its service area from surrounding pixels
configuration,
The computed noise N is compared to the codec configuration reference noise N Ref . If the values are the same, the
defined graphs are used as is, otherwise the graphs are downshifted by the difference N N Ref .
158

AT332_TRR_E0
Since the calculations are based on C/I and C/(I+N):
Atoll calculates the UL C/I to all the GPRS/EDGE TBC transmitters.
For more information on UL C/I calculation, see "Coverage by UL C/I" on page 156.

Atoll uses hard-coded parameters for simulations. In that case, the DL service area is based on the option "HCS servers" with
a margin of 4 dB.
Txi

Txi
Txj
And P rec BCCH Best P rec BCCH 4dB
ji
Txi
3.5.5.2 Codec Mode Selection

For each MSA, cm = Highest Priority CM
Term
P rec
----------------- Adaptation Threshold
CM
I+N
Term
P rec
- , Atoll determines the CQI from the CQI=f(C/I) graph associated to the selected codec mode, cm.
For -----------I+N
3.6 Traffic Analysis

When starting a traffic analysis, Atoll distributes the traffic from maps to transmitters of each layer according to the
compatibility criteria defined in the transmitter, services, mobility type, terminal type properties. Transmitters considered in
traffic analysis are the active and filtered transmitters that belong to the focus zone.
If no focus zone exists in the .atl document, Atoll takes into account the computation
zone.
For details of the average timeslot capacity calculation, see the Network
Dimensioning section (calculation of minimum reduction factor).
3.6.1 Traffic Distribution

3.6.1.1 Normal Cells (Nonconcentric, No HCS Layer)
3.6.1.1.1
Circuit Switched Services

A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the BCCH and TCH
subcells of a transmitter if:
3.6.1.1.2
The terminal, t, works on the frequency band used by the BCCH subcell,
The terminal, t, works on the frequency band used by the TCH subcell.
Packet Switched Services

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the BCCH and TCH
subcells of a transmitter if:
The transmitter is an GPRS/EDGE station (option specified in the transmitter property dialog box),
The terminal, t, is technologically compatible with the transmitter,
The terminal, t, works on the frequency band used by the TCH subcell.
159

3.6.1.2 Concentric Cells

In case of concentric cells, TCH_INNER TRX type has the highest priority to carry traffic.
3.6.1.2.1

A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the TCH_INNER, BCCH
and TCH subcells of a transmitter if:
3.6.1.2.2
The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells.

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the TCH_INNER,
BCCH and TCH subcells of a transmitter if:
The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells.
3.6.1.3 HCS Layers

For each HCS layer, k, you may specify the maximum mobile speed supported by the transmitters of the layer.
3.6.1.3.1

A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the BCCH and TCH
subcells (and TCH_INNER in case of concentric cells) of a transmitter if:
3.6.1.3.2
The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells,
The users mobility, m, is less than the maximum speed supported by the layer, k.

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the BCCH and TCH
subcells (and TCH_INNER in case of concentric cells) of a transmitter if:
The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells,
The user mobility, m, is less than the maximum speed supported by the layer, k.
3.6.2 Calculation of the Traffic Demand per Subcell

Here we assume that:
Users considered for evaluating the traffic demand fulfil the compatibility criteria defined in the transmitter, services,
mobility, terminal properties as explained above.
Atoll distributes traffic on subcell service areas, which are determined using the option Best signal level per HCS
layer with a 0 dB overlap margin and the subcell reception threshold as lower threshold.
Same traffic is distributed to the BCCH and TCH subcells.
3.6.2.1 User Profile Traffic Maps

3.6.2.1.1
Normal Cells (Nonconcentric, No HCS Layer)

Number of subscribers ( X up m ) for each TCH subcell (Txi, TCH), per user profile up with a given mobility m, is inferred as:
X up m Txi TCH = S up m Txi TCH D
Sup,m is the TCH service area containing the user profile up with the mobility m and D is the user profile density.
For each behaviour described in the user profile up, Atoll calculates the probability for the user to be connected with a given
service using a terminal t.
For a circuit switched service c, we have:
160

AT332_TRR_E0
N call d
p up c t = ------------------3600
Ncall is the number of calls per hour and d is the average call duration (in seconds).
Then, Atoll evaluates the traffic demand, D up c t m , in Erlangs for the subcell (Txi, TCH) service area.
D up c t m Txi TCH = X up m Txi TCH p up c t
Packet Switched Services (Max Rate)
For a max rate packet switched service p, we have:
N call V 8
p up p t = ----------------------------3600
Ncall is the number of calls per hour and V is the transmitted data volume per call (in Kbytes).
Then, Atoll evaluates the traffic demand, D up p t m , in kbits/s for the subcell (Txi, TCH) service area.
D up p t m Txi TCH = X up m Txi TCH p up p t
Packet Switched Services (Constant Bit Rate)
For a constant bit packet switched service p, we have:
N call d
p up p t = ------------------3600
Ncall is the number of calls per hour and d is the average call duration (in seconds).
Then, Atoll evaluates the traffic demand, D up p t m , in kbits/s for the subcell (Txi, TCH) service area.
D up p t m Txi TCH = X up m Txi TCH p up p t
3.6.2.1.2
Concentric Cells
In case of concentric cells, Atoll distributes a part of traffic on the TCH_INNER service area (TCH_INNER is the highest priority
traffic carrier) and the remaining traffic on the outer ring served by the TCH subcell. The traffic spread over the TCH_INNER
subcell may overflow to the TCH subcell. In this case, the traffic demand is the same on the TCH_INNER subcell but increases
on the TCH subcell.
Traffic overflowing from the TCH_INNER to the TCH is not uniformly spread over the
TCH service area. It is still located on the TCH_INNER service area.
Number of subscribers ( X up m ) for each TCH_INNER (Txi, TCH_INNER) and TCH (Txi, TCH) subcell, per user profile up with a
given mobility m, is inferred as:
X up m Txi,TCH_INNER = S up m Txi,TCH_INNER D
X up m Txi,TCH = S up m Txi,TCH S up m Txi,TCH_INNER D
S up m Txi,TCH_INNER and S up m Txi,TCH respectively refer to the TCH_INNER and TCH subcell service areas containing the
user profile up with the mobility m. D is the user profile density.
Figure 3.4: Representation of a Concentric Cell TXi
161


For each user of the user profile up using a circuit switched service c with a terminal t, Atoll calculates the probability ( p up c t )
of the user being connected. Calculations are detailed in "Circuit Switched Services" on page 159.
Then, Atoll evaluates the traffic demand, D up c t m , in Erlangs in the (Txi, TCH_INNER) and (Txi, TCH) subcell service areas.
D up c t m Txi,TCH_INNER = X up m Txi,TCH_INNER p up c t
D up c t m Txi,TCH = X up m Txi,TCH p up c t + D up c t m Txi,TCH_INNER O max Txi,TCH_INNER
O max Txi,TCH_INNER is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell.
Packet Switched Services (Max Rate)
For each user of the user profile up using a max rate packet switched service p with a terminal t, probability of the user being
connected ( p up p t ) is calculated as explained in "Packet Switched Services" on page 159.
Atoll evaluates the traffic demand, D up p t m , in kbits/s in the (Txi, TCH_INNER) and (Txi, TCH) subcell service areas.
D up p t m Txi,TCH_INNER = X up m Txi,TCH_INNER p up p t
D up p t m Txi,TCH = X up m Txi,TCH p up p t + D up p t m Txi,TCH_INNER O max Txi,TCH_INNER
Packet Switched Services (Constant Bit Rate)
For each user of the user profile up using a constant bit packet switched service p with a terminal t, probability of the user
being connected ( p up p t ) is calculated as explained in "Packet Switched Services" on page 159.
Atoll evaluates the traffic demand, D up p t m , in kbits/s in the (Txi, TCH_INNER) and (Txi, TCH) subcell service areas.
D up p t m Txi,TCH_INNER = X up m Txi,TCH_INNER p up p t
D up p t m Txi,TCH = X up m Txi,TCH p up p t + D up p t m Txi,TCH_INNER O max Txi,TCH_INNER
3.6.2.1.3
HCS Layers
We assume two HCS layers: the micro layer has a higher priority than the macro layer. Txi belongs to the micro layer and Txj
to the macro. The traffic contained in the input traffic map can be assigned to all the HCS layers.
Normal Cells
Atoll distributes traffic on the TCH service areas. The traffic capture is calculated with the option Best signal level per HCS
macro
layer meaning that there is an overlap between HCS layers service areas. Let S overlapping Txj TCH denote this area (TCH
service area of the macro layer overlapped by the TCH service area of the micro layer). Traffic on the overlapping area is
distributed to the TCH subcell of the micro layer because it has a higher priority. On this area, traffic of the micro layer may
overflow to the macro layer. In this case, the traffic demand is the same on the TCH subcell of the micro layer but increases
on the TCH subcell of the macro layer.
Traffic overflowing to the macro layer is not uniformly spread over the TCH service area
of Txj. It is only located on the overlapping area.
162

AT332_TRR_E0
Figure 3.5: Representation of Micro and Macro Layers

Atoll evaluates the traffic demand on the micro layer (higher priority) as explained above. For further details, please refer to
formulas for normal cells. Then, it proceeds with the macro layer (lower priority).
macro
Number of subscribers ( X up m ) for each TCH subcell (Txj, TCH) of the macro layer, per user profile up with the mobility m, is
inferred as:
macro
macro
macro
X up m Txj TCH = S up m Txj TCH S up m overlapping Txj TCH D

macro
S up m Txj TCH is the TCH service area of Txj containing the user profile up with the mobility m and D is the profile density.
For each user described in the user profile up with the circuit switched service c and the terminal t, the probability for the user
being connected ( p up c t ) is calculated as explained in "Circuit Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up c t m , in Erlangs in the subcell (Txj, TCH) service area.
macro
S upm overlapping Txj TCH

macro
macro
micro
D up c t m Txj TCH = X up m Txj TCH p up c t + D up c t m Txi TCH ----------------------------------------------------------------- Omax Txi TCH
micro
S up m Txi TCH
For each user described in the user profile up with the packet switched service p and the terminal t, probability for the user
to be connected ( p up p t ) is calculated as explained in "Packet Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up p t m , in kbits/s in the subcell (Txj, TCH) service area.
macro
S upm overlapping Txj TCH

macro
macro
micro
D up p t m Txj TCH = X up m Txj TCH p up p t + D up p t m Txi TCH ----------------------------------------------------------------- Omax Txi TCH
micro
S up m Txi TCH
O max Txi TCH is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi (micro
micro
layer) and S up m Txi TCH is the TCH service area of Txi containing the user profile up with the mobility m.
Concentric Cells
Atoll evaluates the traffic demand on the micro layer (higher priority HCS layer) as explained above. For further details, please
refer to formulas given in case of concentric cells. Then, it proceeds with the macro layer (lower priority HCS layer).
The traffic capture is calculated with the option Best signal level per HCS layer. It means that there are overlapping areas
between HCS layers traffic is spread according to the layer priority. On these areas, traffic of the higher priority layer may
overflow.
The TCH_INNER service area of the macro layer is overlapped by the micro layer. This area consists of two parts: an area
macro
overlapped by the TCH service area of the micro layer S overlapping Txi TCH Txj,TCH_INNER and another overlapped by the
macro
TCH_INNER service area of the micro layer S overlapping Txi,TCH_INNER Txj,TCH_INNER .

Let us consider three areas, S1, S2 and S3.
macro
macro
S 1 = S up m Txj,TCH_INNER S up m overlapping Txi TCH Txj,TCH_INNER

macro
S 2 = S up m overlapping Txi,TCH_INNER Txj,TCH_INNER

macro
S 3 = S up m overlapping Txi TCH Txj,TCH_INNER S 2
163

Figure 3.6: Concentric Cells

macro
S up m Txj,TCH_INNER is the TCH_INNER subcell service area of Txj containing the user profile up with the mobility m. We
only consider the overlapping areas containing the user profile up with the mobility m.
macro
On S1, the number of subscribers per user profile up with a given mobility m ( X up m ) is inferred:
macro
X up m Txj,TCH_INNER = S 1 D
D is the user profile density.
The traffic spread over the TCH_INNER service area of the micro layer may overflow on the TCH subcell. The traffic overflowing
to the TCH subcell is located on the TCH_INNER service area. On S2, the TCH subcell traffic coming from the TCH_INNER subcell
traffic overflow may overflow proportional to R2.
S2
R 2 = ------------------------------------------------------micro
S up m Txi,TCH_INNER
The traffic spread over the ring served by the TCH subcell of the micro layer only may overflow on S3 proportional to R3.
S3
R 3 = -------------------------------------------------------------------------------------------------micro
micro
S up m Txi,TCH S up m Txi,TCH_INNER
micro
micro
S up m Txi,TCH and S up m Txi,TCH_INNER are the TCH and TCH_INNER service areas of Txi respectively containing the user
profile up with the mobility m.
For each user described in the user profile up with a circuit switched service c and a terminal t, the probability for the user
being connected ( p up c t ) is calculated as explained in "Circuit Switched Services" on page 159. Then, Atoll evaluates the
macro
traffic demand, D up c t m , in Erlangs in the subcell (Txj, TCH_INNER) service area.

macro
X up m Txj,TCH_INNER p up c t +
macro
D up c t m Txj,TCH_INNER = R D micro Txi,TCH_INNER O Txi,TCH_INNER O Txi,TCH +

2
up c t m
max
max
micro
R 3 X up m Txi TCH p up c t O max Txi TCH

For each user described in the user profile up with a packet switched service p and a terminal t, probability for the user to be
connected ( p up p t ) is calculated as explained in "Packet Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up p t m , stated in kbits/s in the subcell (Txj, TCH_INNER) service area.
macro
X up m Txj,TCH_INNER p up p t +
macro
D up p t m Txj,TCH_INNER = R D micro
2
up p t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi,TCH +
micro
R 3 X up m Txi TCH p up p t O max Txi TCH
164

AT332_TRR_E0
O max Txi TCH and O max Txi,TCH_INNER are the maximum rates of traffic overflow (stated in %) specified for the TCH
and TCH_INNER subcells of Txi respectively.
The area of the TCH ring of the macro layer is overlapped by the micro layer. There are two parts: an area overlapped by the
macro
TCH service area of the micro layer S overlapping Txi TCH Txj,TCH -- TCH_INNER and another one by the TCH_INNER service
macro
area of the micro layer S overlapping Txi,TCH_INNER Txj,TCH -- TCH_INNER .

macro
macro
macro
S' 1 = S up m Txj,TCH S up m Txj,TCH_INNER S up m overlapping Txi TCH Txj,TCH -- TCH_INNER

macro
S' 2 = S up m overlapping Txi,TCH_INNER Txj,TCH -- TCH_INNER

macro
S' 3 = S up m overlapping Txi TCH Txj,TCH -- TCH_INNER S' 2

macro
macro
S up m Txj,TCH and S up m Txj,TCH_INNER are the TCH and TCH_INNER subcell service areas of Txj respectively. We only
consider the overlapping areas containing the user profile up with the mobility m.
macro
On S1, the number of subscribers per user profile up with a given mobility m ( X up m ) is inferred:
macro
X up m Txj,TCH = S' 1 D
D is the user profile density.
The traffic spread over the TCH_INNER service area of the micro layer may overflow on the TCH subcell. The traffic overflowing
on the TCH subcell is located on the TCH_INNER service area. On S2, the TCH subcell traffic coming from the TCH_INNER
subcell traffic overflow may overflow proportionally to R2.
S' 2
R' 2 = ------------------------------------------------------micro
S up m Txi,TCH_INNER
The traffic spread over the ring served by the TCH subcell of the micro layer only may overflow on S3 proportional to R3.
S' 3
R' 3 = -------------------------------------------------------------------------------------------------micro
micro
S up m Txi,TCH S up m Txi,TCH_INNER
micro
micro
S up m Txi,TCH and S up m Txi,TCH_INNER are the TCH and TCH_INNER service areas of Txi respectively containing the user
profile up with the mobility m.
For each user described in the user profile up with a circuit switched service c and a terminal t, the probability for the user
being connected ( p up c t ) is calculated as explained in "Circuit Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up c t m , in Erlangs in the subcell (Txj, TCH) service area.
macro
X up m Txj TCH p up c t +
macro
D up c t m Txj TCH =
macro
D up c t m Txj,TCH_INNER O max Txj,TCH_INNER +

micro
R' 2 D up c t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi,TCH +

micro
R' 3 X up m Txi TCH p up c t m O max Txi TCH

For each user described in the user profile up with a packet switched service p and a terminal t, the probability for the user
being connected ( p up p t ) is calculated as explained in "Packet Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up p t m , in kbits/s in the subcell (Txj, TCH) service area.
macro
X up m Txj TCH p up p t +
macro
D up p t m Txj TCH =
macro
D up p t m Txj,TCH_INNER O max Txj,TCH_INNER +

micro
R' 2 D up p t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi,TCH +

micro
R' 3 X up m Txi TCH p up p t m O max Txi TCH
165

O max Txi,TCH is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi (micro layer),
O max Txi,TCH_INNER the maximum rate of traffic overflow indicated for the TCH_INNER subcell of Txi (macro layer),
O max Txj,TCH_INNER the maximum rate of traffic overflow indicated for the TCH_INNER subcell of Txj (macro layer) and
micro
X up m Txi TCH the number of subscribers with the user profile up and mobility m on the TCH service area of Txi (as
explained in "Concentric Cells" on page 160).
3.6.2.2 Sector Traffic Maps

We assume that the traffic map is built from a coverage by transmitter prediction calculated for the TCH subcells with options:
HCS Servers and no margin if the network only consists of normal cells and concentric cells,
Highest Priority HCS Server and no margin in case of HCS layers.
When creating the traffic map, you have to specify the traffic demand per transmitter and per service (throughput for a max
rate packet switched service and Erlangs for a circuit switched or constant bit rate packet switched service) and the global
distribution of terminals and mobility types.
Let E c Txi TCH denote the Erlangs for the circuit switched service, c, on the TCH subcell of Txi.
Let T p Txi TCH denote the throughput of the packet switched service (Max Bit Rate), p, on the TCH subcell of Txi.
Let E p Txi TCH denote the Erlangs for the packet switched service (Constant Bit Rate), p, on the TCH subcell of Txi.
We assume that 100% of users have the terminal, t, and the mobility type, m.
3.6.2.2.1
Normal Cells (Nonconcentric, No HCS Layer)

For each circuit switched service, c, Atoll evaluates the traffic demand, Dc,t,m, in Erlangs in the subcell (Txi, TCH) service area.
D c t m Txi TCH = E c Txi TCH
For each packet switched service (Max Bit Rate), p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell (Txi, TCH)
service area.
D p t m Txi TCH = T p Txi TCH
For each packet switched service (Constant Bit Rate), p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell (Txi,
TCH) service area.
D p t m Txi TCH = E p Txi TCH TP p GBR
TP p GBR is the guaranteed bit rate of the constant bit rate packet switched service p.
3.6.2.2.2
Concentric Cells
In case of concentric cells, Atoll distributes a part of traffic on the TCH_INNER service area (TCH_INNER is the highest priority
traffic carrier) and the remaining traffic, on the ring served by the TCH subcell only. The traffic spread over the TCH_INNER
subcell may overflow to the TCH subcell. In this case, the traffic demand is the same on the TCH_INNER subcell and rises on
the TCH subcell.
Traffic overflowing from the TCH_INNER to the TCH is not uniformly spread over the TCH
service area. It is only located on the TCH_INNER service area.
For each circuit switched service, c, Atoll evaluates the traffic demand, Dc,t,m, in Erlangs in the subcell, (Txi, TCH_INNER) and
(Txi, TCH), service areas.
S Txi,TCH_INNER
D c t m Txi,TCH_INNER = -------------------------------------------- E c Txi TCH
S Txi TCH
and
D c t m Txi,TCH =
-------------------------------------------------------------------------------S Txi,TCH S Txi,TCH_INNER E c Txi TCH +

S Txi TCH
D c t m Txi,TCH_INNER O max Txi,TCH_INNER
For each packet switched service (Max Bit Rate), p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell, (Txi,
TCH_INNER) and (Txi, TCH), service areas.
166

AT332_TRR_E0
S Txi,TCH_INNER
D p t m Txi,TCH_INNER = -------------------------------------------- T p Txi TCH
S Txi TCH
and
D p t m Txi,TCH =
S Txi,TCH S Txi,TCH_INNER
--------------------------------------------------------------------------------- T p Txi TCH +
S Txi TCH
D p t m Txi,TCH_INNER O max Txi,TCH_INNER
O max Txi,TCH_INNER is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell, S Txi,TCH
and S Txi,TCH_INNER are the TCH and TCH_INNER service areas of Txi respectively.
For each packet switched service (Constant Bit Rate), p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell, (Txi,
TCH_INNER) and (Txi, TCH), service areas.
S Txi,TCH_INNER
D p t m Txi,TCH_INNER = -------------------------------------------- E p Txi TCH TP p GBR
S Txi TCH
and
D p t m Txi,TCH =
-------------------------------------------------------------------------------S Txi,TCH S Txi,TCH_INNER E p Txi TCH TP p GBR +

S Txi TCH
D p t m Txi,TCH_INNER O max Txi,TCH_INNER
O max Txi,TCH_INNER is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell, S Txi,TCH
and S Txi,TCH_INNER are the TCH and TCH_INNER service areas of Txi respectively.
3.6.2.2.3
HCS Layers
We assume we have two HCS layers: the micro layer has a higher priority and the macro layer has a lower one. Txi belongs to
the micro layer and Txj to the macro one. The traffic contained in the input traffic map can be assigned to all the HCS layers.
Normal Cells
Atoll distributes traffic on the TCH service areas. The traffic capture is calculated with the option HCS Servers. It means that
macro
there is an overlapping area between HCS layers. Let S overlapping Txj TCH denote the TCH service area of the macro layer
overlapped by the TCH service area of the micro layer. Traffic on the overlapping area is distributed to the TCH subcell of the
micro layer (higher priority layer). On this area, traffic of the micro layer may overflow to the macro layer. In this case, the
traffic demand is the same on the TCH subcell of the micro layer but rises on the TCH subcell of the macro layer.
Traffic overflowing on the macro layer is not uniformly spread over the TCH service area
of Txj. It is only located on the overlapping area.
Atoll starts evaluating the traffic demand on the micro layer (highest priority HCS layer).
micro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txi, TCH) service area.
micro
D c t m Txi TCH = E c Txi TCH

micro
For each packet switched service (Max Bit Rate), p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txi,
TCH) service area.
micro
D p t m Txi TCH = T p Txi TCH

micro
For each packet switched service (Constant Bit Rate), p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell
(Txi, TCH) service area.
micro
D p t m Txi TCH = E p Txi TCH TP p GBR

Then, Atoll proceeds with the macro layer (lower priority HCS layer).
macro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txj, TCH) service area.
167

macro
S overlapping Txj TCH

macro
micro
- O max Txi TCH
D c t m Txj TCH = E c Txj TCH + D c t m Txi TCH ---------------------------------------------------micro
S
Txi TCH
macro
For each packet switched service (Max Bit Rate), p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj,
TCH) service area.
macro

macro
micro
- O max Txi TCH
D p t m Txj TCH = T p Txj TCH + D p t m Txi TCH ---------------------------------------------------micro
S
Txi TCH
O max Txi TCH is the maximum rate of traffic overflow (in %) specified for the TCH subcell of Txi (micro cell) and
S
micro
Txi TCH the TCH service area of Txi.

macro
(Txj, TCH) service area.
macro

macro
micro
- O max Txi TCH
D p t m Txj TCH = E p Txi TCH TP p GBR + D p t m Txi TCH ---------------------------------------------------micro
S
Txi TCH
O max Txi TCH is the maximum rate of traffic overflow (in %) specified for the TCH subcell of Txi (micro cell) and
S
micro
Txi TCH the TCH service area of Txi.

You can restrict the traffic assignement of each traffic map to a specific HCS layer in the
running options of the traffic capture. If you do so, no overflow occurs between HCS
layers and the only overflow which is considered occurs within concentric cells (See
"Concentric Cells" on page 160).
Concentric Cells
Atoll evaluates the traffic demand on the micro layer as explained above in case of concentric cells and then proceeds with
the macro layer (lower priority layer).
The traffic capture is calculated with the option HCS Servers. It means that there is overlapping areas between HCS layers
traffic is spread over according to the layer priority. On these areas, traffic of the higher priority layer may overflow.
Figure 3.7: Concentric Cells

The TCH_INNER service area of the macro layer is overlapped by the micro layer. This area consists of two parts: an area
macro
overlapped by the TCH service area of the micro layer S overlapping Txi TCH Txj,TCH_INNER and another overlapped by the
macro
TCH_INNER service area of the micro layer S overlapping Txi,TCH_INNER Txj,TCH_INNER .

168

AT332_TRR_E0
S1 = S
macro
macro
Txj,TCH_INNER S overlapping Txi TCH Txj,TCH_INNER
macro
S 2 = S overlapping Txi,TCH_INNER Txj,TCH_INNER

macro
S 3 = S overlapping Txi TCH Txj,TCH_INNER S 2

S
macro
Txj,TCH_INNER is the TCH_INNER subcell service area of Txj.
The traffic specified for Txj in the map description ( E c Txj TCH ) is spread over S1 proportionally to R1.
S1
R 1 = ------------------------------------map
S
Txj TCH
S
map
Txj TCH is the TCH service area of Txj in the traffic map with the option Best signal level of the highest priority layer.
The traffic spread over the TCH_INNER service area of the micro layer may overflow to the TCH subcell. The traffic overflowing
to the TCH subcell is located on the TCH_INNER service area. On S2, the TCH subcell traffic coming from the TCH_INNER subcell
traffic overflow may overflow proportional to R2.
S2
R 2 = ------------------------------------------------------micro
S
Txi,TCH_INNER
The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S3 proportional to R3.
S3
R 3 = -------------------------------------------------------------------------------------------------micro
micro
S
Txi,TCH S
Txi,TCH_INNER
macro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txj, TCH_INNER)
service area.
R 1 E c Txj TCH +
macro
D c t m Txj,TCH_INNER =
micro
R 2 D c t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi TCH +

micro
micro
S
Txi TCH S
Txi,TCH_INNER
- E c Txi TCH O max Txi TCH
R 3 ---------------------------------------------------------------------------------------------------------micro
S
Txi TCH
macro
For each packet switched service (Max Bit Rate), p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj,
TCH_INNER) service area.
R 1 T p Txj TCH +
macro
D p t m Txj,TCH_INNER =
micro
R 2 D p t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi TCH +

micro
micro
S
Txi TCH S
Txi,TCH_INNER
- T p Txi TCH O max Txi TCH
R 3 ---------------------------------------------------------------------------------------------------------micro
S
Txi TCH
O max Txi TCH is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi,
O max Txi,TCH_INNER is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi and
S
micro
Txi TCH is the TCH subcell service area of Txi.

macro
(Txj, TCH_INNER) service area.
R 1 E p Txi TCH TP p GBR +
micro
R 2 D p t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi TCH +

macro
D p t m Txj,TCH_INNER =
micro
micro
S
Txi TCH S
Txi,TCH_INNER
---------------------------------------------------------------------------------------------------------micro
Txi
TCH
R3
Txi
TCH
TP
p
p GBR O max Txi TCH
169

O max Txi,TCH_INNER is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi and
S
micro
Txi TCH is the TCH subcell service area of Txi.
The area of the TCH ring of the macro layer is overlapped by the micro layer. There are two parts: an area overlapped by the
macro
TCH service area of the micro layer S overlapping Txi TCH Txj,TCH -- TCH_INNER and another overlapped by the TCH_INNER
macro
service area of the micro layer S overlapping Txi,TCH_INNER Txj,TCH -- TCH_INNER .

S' 1 = S
macro
Txj TCH S
macro
macro
Txj,TCH_INNER S overlapping Txi TCH Txj,TCH -- TCH_INNER
macro
S' 2 = S overlapping Txi,TCH_INNER Txj,TCH -- TCH_INNER

macro
S' 3 = S overlapping Txi TCH Txj,TCH -- TCH_INNER S' 2

S
macro
Txj TCH and S
macro
Txj,TCH_INNER are the TCH and TCH_INNER subcell service areas of Txj respectively.
The traffic specified for Txj in the map description ( E c Txj TCH ) is spread over S1 proportional to R1.
S' 1
R' 1 = ------------------------------------map
S
Txj TCH
S
map
Txj TCH is the TCH service area of Txj in the traffic map with the option Best signal level of the highest priority layer.
The traffic spread over the TCH_INNER service area of the micro layer may overflow to the TCH subcell. The traffic overflowing
to the TCH subcell is located on the TCH_INNER service area. On S2, the TCH subcell traffic coming from the TCH_INNER
subcell traffic overflow may overflow proportional to R2.
S' 2
R' 2 = ------------------------------------------------------micro
S
Txi,TCH_INNER
The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S3 proportional to R3.
S' 3
R' 3 = -------------------------------------------------------------------------------------------------micro
micro
S
Txi,TCH S
Txi,TCH_INNER
macro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txj, TCH) service area.
R' 1 E c Txj TCH +

macro
macro
D c t m Txj
TCH =
D c t m Txj,TCH_INNER O max Txj,TCH_INNER +

micro
R' 2 D c t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi TCH +

micro
micro
S
Txi,TCH S
Txi,TCH_INNER
- E c Txi TCH O max Txi TCH
R' 3 ------------------------------------------------------------------------------------------------------micro
S
Txi,TCH
macro
For each packet switched service (Max Bit Rate), p,Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj,
TCH) service area.
R' 1 T p Txj TCH +

macro
macro
D p t m Txj
TCH =

micro
R' 2 D p t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi TCH +

micro
micro
S
Txi,TCH S
Txi,TCH_INNER R' 3 ------------------------------------------------------------------------------------------------------ T p Txi TCH O max Txi TCH
micro
Txi,TCH
S
170

AT332_TRR_E0
O max Txj,TCH_INNER is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txj,
O max Txi,TCH_INNER is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi,
S
micro
Txi,TCH is the TCH subcell service area of Txi and S
micro
Txi,TCH_INNER is the TCH_INNER subcell service area of Txi.

macro
For each packet switched service (Constant Bit Rate), p,Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell
(Txj, TCH) service area.
R' 1 E p Txi TCH TP p GBR +

macro

micro
macro
D p t m Txj
TCH =
R' 2 D p t m Txi,TCH_INNER O max Txi,TCH_INNER O max Txi TCH +

micro
micro
S
Txi,TCH S
Txi,TCH_INNER ------------------------------------------------------------------------------------------------------micro
Txi,TCH
S
R' 3
Txi
TCH
TP
p
p GBR O max Txi TCH
O max Txj,TCH_INNER is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txj,
O max Txi,TCH_INNER is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi,
S
micro
Txi,TCH is the TCH subcell service area of Txi and S
micro
Txi,TCH_INNER is the TCH_INNER subcell service area of Txi.
3.7 Network Dimensioning

Atoll is capable of dimensioning a GSM GPRS EDGE network with a mixture of circuit and package switched services. This
section describes the technical details of Atolls dimensioning engine.
3.7.1 Dimensioning Models and Quality Graphs

In Atoll, a dimensioning model is an entity utilized by the dimensioning engine along with other inputs (traffic, limitations,
criteria, etc.) in the process of dimensioning. A dimensioning model defines the QoS KPIs to be taken into account when
dimensioning a network for both circuit and packet switched traffic. The user can define either to use Erlang B or Erlang C
queuing model for circuit switched traffic and can define which KPIs to consider when dimensioning the network for packet
switched traffic. The dimensioning engine will only utilize the quality curves of the KPI selected. The KPIs not selected are
supposed to be either already satisfactory or not relatively important.
3.7.1.1 Circuit Switched Traffic

The network dimensioning for circuit switched traffic is performed using the universally accepted and adopted Erlang B and
Erlang C formulas. The dimensioning criterion in these formulas is the Grade of Service or the allowed blocking probability of
the circuit switched traffic.
In the Erlang B approach, this Grade of Service is defined as the percentage of incoming circuit switched calls that are blocked
due to lack of resources or timeslots. This formula implies a loss system. The blocked calls are supposed to be lost and the
caller has to reinitiate it.
In the Erlang C approach, the Grade of Service is the percentage of incoming calls that are placed in a waiting queue when
there are no resources available, until some resources or timeslots are liberated. This queuing system has no lost calls. As the
load on the system increases, the average waiting time in the queue also increases.
These formulas and their details are available in many books. For example, Wireless Communications Principles and Practice
by Theodore S. Rappaport, Prentice Hall.
Following the common practice, network dimensioning in Atoll is based on the principle that a voice or GSM call has priority
over data transmission. Therefore, as explained later in the network dimensioning steps, Atoll first performs network
dimensioning according to the circuit switched traffic present in the subcell in order to ensure the higher priority service
availability before performing the same for the packet switched traffic.
171

3.7.1.2 Packet Switched Traffic

Since packet switched traffic does not occupy an entire timeslot the whole time, it is much more complicated to study than
circuit switched traffic. Packet traffic is intermittent and bursty. Whenever there is packet data to be transferred, a Temporary
Block Flow (TBF) is initiated for transferring these packets. Multiple TBFs can be multiplexed on the same timeslot. This implies
that there can be many packet switched service users that have the same timeslots assigned for packet data transfer but at
different intervals of time.
This multiplexing of a number of packet switched service users over the same timeslots incurs a certain reduction in the
throughput (data transfer rate) for each multiplexed user. This reduction in the throughput is more perceivable when the
system traffic load is high. The following parts describe the three most important Key Performance Indicators in GPRS/EDGE
networks and how they are modelled in Atoll.
3.7.1.2.1
Throughput
Throughput is defined as the amount of data delivered to the Logical Link Control Layer in a given unit of time. Each temporary
block flow (TBF), and hence each user, has an associated measured throughput sample in a given network. Each network will
have a different throughput probability distribution depending on the load and network configuration. Instead of using the
precise probability distributions, it is more practical to compute the average and percentile throughput values.
In GPRS, the resources are shared between the users being served, and consequently, the throughput is reduced as the
number of active users increases. This reduction in user perceived throughput is modelled through a reduction factor. The
throughput experienced by a user accessing a particular service can be calculated as:
User throughput = Number of allocated timeslots x Timeslot capacity x Reduction Factor
Or
User throughput per allocated timeslot = Timeslot capacity x Reduction Factor
Timeslot Capacity
The timeslot capacity is the average throughput per fully utilized timeslot. It represents the average throughput from the
network point of view. It mainly depends on the networks propagation conditions and criteria in the coverage area of a
transmitter (carrier power, carrier-to-interference distribution, etc.). It is a measure of how much data the network is able to
transfer with 1 data Erlang, or in other words, how efficiently the hardware resources are being utilized by the network. It may
also depend on the RLC protocol efficiency.
Atoll computes the average timeslot capacity during the traffic analysis and is used to determine the minimum throughput
reduction factor. But since this information is displayed in the network dimensioning results (only due to relevance), this
information has been considered as a part of the network dimensioning process in this document.
Timeslot Utilisation
Timeslot utilization takes into account the average number of timeslots that are available for packet switched traffic. It is a
measure of how much the network is loaded with data services. Networks with timeslot utilisation close to 100% are close to
saturation and the end-user performance is likely to be very poor.
In Atoll this parameter is termed as the Load (Traffic load for circuit switched traffic and packet switched traffic load for packet
switched traffic). It is described in more detail in the Network dimensioning steps section.
Reduction Factor
Reduction factor takes into account the user throughput reduction due to timeslot sharing among many users. The figure
below shows how the peak throughput available per timeslot is reduced by interference and sharing.Reduction factor is a
function of the number of timeslots assigned to a user (Nu), number of timeslots available in the system (Ns) and the average
system packet switched traffic load (Lp) (utilization of resources in the system). Data Erlangs or data traffic is given by:
Data Erlangs = L P N S
172

AT332_TRR_E0
Figure 3.8: Reduction of Throughput per Timeslot

More precisely, the reduction factor is a function of the ratio Ns/Nu (Np). Np models the equivalent timeslots that are available
for the packet switched traffic in the system. For example, a 24-timeslot system with each user assigned 3 timeslots per
connection can be modelled by a single timeslot connection system with 8 timeslots in total.
The formula for reduction factor can be derived following the same hypotheses followed by Erlang in the derivation of the
blocking probability formulas (Erlang B and Erlang C).
Let X be a random variable that measures the reduction factor in a certain system state:
0 if n = 0
1 if 0 < n N P
NP
------ if n > N P
n
n is the instantaneous number of connections in the system. The throughput reduction factor is defined as:
P X= n
X -------------------P X= 0
RF
n=0
Or,
RF =
PX= n
X --------------------------n=0
P X= i
i
Here, P(X=n) is the probability function of having n connections in the system. Under the same assumptions as those of the
Erlang formulas, the probability function can be written as:
n
LP NP
----------------------n!
P X = n = --------------------------------------------------------------------------------------N
P
LP NP
LP NP
----------------------------- i NP
+ 1 N P! N P
+
--------------------i!
i=0
i = NP
if 0 n N P
LP NP
----------------------------- i NP
N P! N P
P X = n = --------------------------------------------------------------------------------------N
P
LP NP
LP NP
----------------------------- i NP
+ 1 N P! N P
+
--------------------i!
i=0
i=N
if n > N P
Hence the reduction factor can finally be written as:
173

NP
LP NP
--------------------- +
i!
N
LP NP
-----P-
-----------------------------i N i
P
i=1
i = N + 1 N P! N P
P
RF = ------------------------------------------------------------------------------------------------------N
LP NP
--------------------- +
i!
i=1
LP NP
-----------------------------i N
P
+ 1 N P! N P
i = NP
This formula is not directly applicable in any software application due to the summations up to infinity. Atoll uses the following
version of this formula that is exactly the same formula without the summation overflow problem.
NP
NP + 1
P
n
n
NP
L
LP NP
----------------------- ------------------- ln 1 L P +
----P
N P!
n!
n
n
=
1
n
=
1
RF = -----------------------------------------------------------------------------------------------------------------------N
P
n=1
NP
n
LP
LP NP
LP NP
----------------------- + -------------------------- -------------N P!
1 LP
n!
The default quality curves for the Reduction Factor have been derived using the above formula. Each curve is for a fixed
number of timeslots available for packet switched traffic (Np) describing the reduction factor at different values of packet
switched traffic load (Lp). The figure below contains all the reduction factor quality curves in Atoll. The Maximum reduction
factor can be 1, implying a maximum throughput, and the minimum can be 0, implying a saturated system with no data
throughput.
Figure 3.9: Reduction Factor for Different Packet Switched Traffic Loads (Lp, X-axis)
Each curve in the above figure represents an equivalent number of packet switched timeslots, NP.
3.7.1.2.2
Delay
Delay is the time required for an LLC PDU to be completely transferred from the SGSN to the MS, or vice versa. As the delay
is a function of the delays and the losses incurred at the packet level, the network parameters, such as the packet queue
length, and different protocol properties, such as the size of the LLC PDU, become important. It is also quite dependent upon
the radio access round trip time (RA RTT) and has a considerable impact on the application level performance viewed by the
user.
The delay parameter is a user level parameter rather than being a network level quantity, like throughput per cell, timeslot
capacity, TBF blocking and reduction factor, hence it is difficult to model and is currently under study. Hence, no default curve
is presently available for delay in Atoll.
174

AT332_TRR_E0
3.7.1.2.3
Blocking Probability
In GPRS, there is no blocking as in circuit switched connections. If a new temporary block flow (TBF) establishment is requested
and there are already M users per timeslot, M being the maximum limit of multiplexing per timeslot (Multiplexing factor), the
request is queued in the system to be established later when resources become available.
Supposing that M number of users can be multiplexed over a single timeslot (PDCH), we can have a maximum of M * Np users
in the system. This implies that if a new TBF is requested when there are already M * Np users active, it will be blocked and
placed in a queue. So the blocking probability is the probability of having M * Np + 1 users in the system or more, meaning,
P X= n
for n = M N P + 1
as in this case n is always greater than Np, we have,

n
LP NP
----------------------------- i NP
N P! N P
P X = n = --------------------------------------------------------------------------------------N
P
LP NP
--------------------- +
i!
i=0
LP NP
----------------------------- i NP
+ 1 N P! N P
i = NP
So, the Blocking Probability can be given as:
BP =
LP NP
------------------------------ 1 NP
N
!
N
n = MN +1 P
P
P
P X = n = ---------------------------------------------------------------------------------------N
n = MN+1
LP NP
--------------------- +
i!
i=0
i = NP
LP NP
------------------------------ 1 NP
+ 1 N P! N P
Eliminating the summations to infinity, the blocking probability can be stated in a simpler form:
M NP
LP NP
LP
----------------------------------------- ------------ M NP NP 1 L
P
N P! N P
BP = ---------------------------------------------------------------------------------N
P
i=0
NP
i
LP
LP NP LP NP
--------------------- + -------------------------- ------------N P!
1 LP
i!
The above formula has been used to generate the default quality curves for blocking probability in Atoll.
These graphs are generated for a user multiplexing factor of 8 users per timeslot. Each curve represents an equivalent number
of packet switched timeslots, NP.
The curves depict the blocking probabilities for different number of available connections (Np) at different packet switched
traffic loads (Lp) for a fixed user multiplexing factor of 8. The figure below contains all the blocking probability curves for
packet switched traffic dimensioning in Atoll. The blocking probability increases with the packet switched traffic load, which
implies that as the packet switched traffic increases for a given number of timeslots, the system starts to get more and more
loaded, hence there is higher probability of having a temporary block flow placed in a waiting queue.
175

Figure 3.10: Blocking Probability for Different Packet Switched Traffic Loads (Lp, X-axis)
Reference: T. Halonen, J. Romero, J. Melero; GSM, GPRS and EDGE performance
Evolution towards 3G/UMTS, John Wiley and Sons Ltd.
3.7.2 Network Dimensioning Process

The network dimensioning process is described below in detail. As the whole dimensioning process is in fact a chain of small
processes that have there respective inputs and outputs, with outputs of a preceding one being the inputs to the next, the
best method is to detail each process individually in form of steps of the global dimensioning process.
3.7.2.1 Network Dimensioning Engine

During the dimensioning process, Atoll first computes the number of timeslots required to accommodate the circuit switched
traffic. Then it calculates the number of timeslots to add in order to satisfy the demand of packet switched traffic. This is
performed using the quality curves entered in the dimensioning model used. If the dimensioning model has been indicated to
take all three KPIs in to account (throughput reduction factor, delay and blocking probability), the number of timeslots to be
added is calculated such that:
The throughput reduction factor is greater than the minimum throughput reduction factor,
Delay is less than the maximum permissible delay defined in the service properties, and
The blocking probability is less than the maximum allowable blocking probability defined in the service properties.
The figure below depicts a simplified flowchart of the dimensioning engine in Atoll.
Figure 3.11: Network Dimensioning Process
176

AT332_TRR_E0
On the whole, following are the inputs and outputs of the network dimensioning process:
3.7.2.1.1
Inputs
3.7.2.1.2
Circuit switched traffic demand

Packet switched traffic demand
Timeslot configurations defined for each subcell
Target traffic overflow rate and Half-rate traffic ratio for each subcell
Service availability criteria: minimum required throughput per user, maximum permissible delay, maximum allowable
blocking probability etc.
Dimensioning model parameters: Maximum number of TRXs per transmitter, dimensioning model for circuit switched
traffic, number of minimum dedicated packet switched timeslots per transmitter, maximum number of TRXs added
for packet switched services, KPIs to consider, and their quality curves.
Outputs
Number of required TRXs per transmitter

Number of required shared, circuit switched and packet switched timeslots
Traffic load
Served circuit switched traffic
Served packet switched traffic
Effective rate of traffic overflow
Actual KPI values: throughput reduction factor, delay and blocking probability
3.7.2.2 Network Dimensioning Steps

This section describes the entire process step by step as it is actually performed in Atoll. Details of the calculations of the
parameters that are calculated during each step are described as well.
3.7.2.2.1
Step 1: Timeslots Required for CS Traffic

Atoll computes the number of timeslots required to accommodate the circuit switched traffic assigned to each subcell. Atoll
takes the circuit switched traffic demand (Erlangs) either user-defined or calculated in the traffic analysis and assigned to the
current subcell and the maximum blocking probability defined for the circuit switched service, and computes the required
number of timeslots to satisfy this demand using the Erlang B or Erlang C formula (as defined by the user).
If the user-defined target rate of traffic overflow per subcell, OTarget, is greater than the maximum blocking rate defined in the
services properties, it is going to be taken as the Grade of Service required for that subcell instead of the maximum blocking
rate of the service.
For the blocking probability GoS and circuit switched traffic demand TDC, Atoll determines the required number of timeslots
TSreq. C for each subcell using formulas described below. In fact, Atoll searches for TSreq. C value until the defined grade of
service is reached.
For Erlang B, we have:
TS
reqC
TD C
--------------------------- TS reqC !
GoS = -----------------------------TS
reqC
TD C
--------------k!
For Erlang C, we have:

TS
reqC
TD C
GoS = -----------------------------------------------------------------------------------------------------------------------------------TS
1
reqC
TD C
TS reqC
TD C
+ TS reqC ! 1 --------------
TS reqC
TD C
---------------k!
Atoll considers the effect of half-rate circuit switched traffic by taking into account a user-defined percentage of half-rate
traffic. Atoll computes the effective equivalent number of full-rate timeslots that will be required to carry the total traffic with
the defined percentage of half-rate traffic.
If the number of timeslots required to accommodate the full-rate circuit switched traffic is TSreq. FR, and the percentage of
half-rate traffic within the subcell is defined by HR, then the effective number of equivalent full-rate circuit switched timeslots
TSeff. that can carry this traffic mix is calculated by:
TS eff = TS reqFR 1 HR
-------
2
177

Atoll employs this simplified approach to integrating half-rate circuit switched traffic, which provides approximately the same
results as obtained by using the half-rate traffic charts.
3.7.2.2.2
Step 2: TRXs Required for CS Traffic and Dedicated PS Timeslots

This stage of the network dimensioning process computes the number of TRXs required to carry the circuit switched traffic
demand through the number of required timeslots calculated above and the timeslot configuration defined by the user in the
network settings. Atoll distributes the number of required circuit switched timeslots calculated in Step 1 taking into account
the presence of dedicated packet switched timeslots in each TRX according to the timeslot configurations.
If a timeslot configuration defines a certain number of dedicated packet switched timeslots pre-allocated in certain TRXs,
those timeslots will not be considered capable of carrying circuit switched traffic and hence will not be allocated. For example,
if 4 timeslots have been marked as packet switched timeslots in the first TRX and Atoll computes 8 timeslots for carrying a
certain circuit switched traffic demand, then the number of TRXs to be allocated cannot be 1 even if there is no packet
switched traffic considered yet.
The total numbers of timeslots that carry circuit switched and packet switched traffic respectively are the sums of respective
dedicated and shared timeslots:
TS P = TS S + TS P dedicated and TS C = TS S + TS C dedicated
3.7.2.2.3
Step 3: Effective CS Blocking, Effective CS Traffic Overflow and Served CS Traffic

In this step, the previously calculated number of required TRXs is used to compute the effective blocking rate for the circuit
switched traffic. This is performed by using the Erlang B or Erlang C formula with the circuit switched traffic demand and the
number of required TRXs as inputs and computing the Grade of Service (or blocking probability). It then calculates the
effective traffic overflow rate, Oeff..
In case of Erlang B formula, the effective rate of traffic overflow for the circuit switched traffic is the same as the circuit
switched blocking rate. While in case of the Erlang C model, the circuit switched traffic is supposed to be placed in an infinitelength waiting queue. This implies that there is no overflow in this case.
From this data, it also computes the served circuit switched traffic. This is the difference of the circuit switched traffic demand
and the percentage of traffic that overflows from the subcell to other subcells calculated above. Hence, for an effective traffic
overflow rate of Oeff. and the circuit switched traffic demand of TDC, the served circuit switched traffic STC is computed as:
ST C = TD C 1 O eff
3.7.2.2.4
Step 4: TRXs to Add for PS Traffic

This step is the core of the dimensioning process for packet switched services. First of all, Atoll computes the number of TRXs
to be added to carry the packet switched traffic demand. This is the number of TRXs that contain dedicated packet switched
and shared timeslots.
To determine this number of TRXs, Atoll calculates the equivalent average packet switched traffic demand in timeslots by
studying each pixel covered by the transmitter. This calculation is in fact performed in the traffic analysis process or is userdefined in the subcells table. Knowing the traffic demand per pixel of the covered area in terms of kbps and the maximum
attainable throughput per pixel (according to the C and/or C/I conditions and the coding scheme curves in the GPRS/EDGE
configuration), Atoll calculates the average traffic demand in packet switched timeslots by:
TD P
Timeslots
Traffic
demand per pixel (kbps)-------------------------------------------------------------------------Throughput per pixel (kbps)
pixel
The average timeslot capacity of a transmitter is calculated by dividing the packet switched traffic demand over the entire
coverage area (in kbps) by the packet switched traffic demand in timeslots calculated above.
With the number of timeslots required to serve the circuit switched traffic, the timeslots required for packet switched traffic
and their respective distributions according to the timeslot configurations being known, Atoll calculates the number of
timeslots available for carrying the packet switched traffic demand. These timeslots can be dedicated packet switched
timeslots and the shared ones. So, following the principle that shared timeslots are potential carriers of both traffic types,
TS P = TS S + TS P dedicated
TS C = TS S + TS C dedicated
The packet switched traffic load is calculated by the formula:
ST C TS C dedicated + TD P
Timeslots
L P = -----------------------------------------------------------------------------------TS P
The second important parameter for the calculation of Reduction Factor, Delay and Blocking Probability is the equivalent
number of available timeslots for packet switched traffic, i.e. NP. This is computed by dividing the total number of timeslots
178

AT332_TRR_E0
available for carrying packet switched traffic by the number of downlink timeslots defined in the mobile terminal properties.
So, NP is calculated at this stage as:
TS P
N P = ----------------------TS Terminal
, TSTerminal is the number of timeslots that a terminal will use in packet switched calls.
The number timeslots that a terminal can use in packet switched calls is the product of the number of available DL timeslots
for packet-switched services (on a frame) and the number of simultaneous carriers (in case of EDGE evolution).
The number of timeslots that a terminal will use in packet switched calls is determined by taking the lower of the maximum
number of timeslots on a carrier for packet switched service defined in the service properties and the maximum number of
timeslots that a mobile terminal can use for packet switched services (see above) on acarrier.
TS Terminal = min TS Max Service TS Max TerminalType
and TS Max TerminalType = TS DL TerminalType Carriers DL TerminalType
Here, the min(X,Y) function yields the lower value among X and Y as result.
Now, knowing the packet switched traffic load, LP, and the equivalent number of available timeslots, NP, Atoll finds out the
KPIs that have been selected before launching the dimensioning process using the quality curves stored in the dimensioning
model.
This particular part of this step can be iterative if the KPIs to consider in dimensioning are not satisfied in the first try. If the
KPIs calculated above are within acceptable limits as defined by the user, it means that the dimensioning process has
acceptable results. If these KPIs are not satisfied, then Atoll increases the number of TRXs calculated for carrying packet
switched traffic by 1 (each increment adding 8 more timeslots for carrying packet switched traffic as the least unit that can be
physically added or removed is a TRX) and resumes the computations from Step 3. It then recalculates the packet switched
traffic load, LP, and the equivalent number of available timeslots, NP. Then it recomputes the KPIs with these new values of LP
and NP. If the KPIs are within satisfactory limits the results are considered to be acceptable. Otherwise, Atoll performs another
iteration to find the best possible results.
The calculated values of all the KPIs are compared with the ones defined in the service properties. The values for maximum
Delay and Blocking probability are defined directly in the properties but the minimum throughput reduction factor is
calculated by Atoll using the users inputs: minimum throughput per user and required availability. This calculation is in fact
performed during the traffic analysis process, but since it is relevant to the dimensioning procedure, it is displayed in a column
in the dimensioning results so that the user can easily compare the minimum requirement on the reduction factor KPI with
the resulting one. If dimensioning is not based on a traffic analysis, the minimum throughput reduction factor is a user-defined
parameter.
Minimum Throughput Reduction Factor Calculation
The minimum throughput reduction factor is computed using the input data: minimum required throughput per user defined
in the service properties, the average throughput per timeslot deduced from the throughput curves stored in the GPRS/EDGE
configuration properties for each coding scheme, the total number of downlink timeslots defined in the properties of the
mobile terminal (See TS Max TerminalType defintion above) and the required availability defined in the service properties.
It is at the stage of calculating the average timeslot capacity per transmitter that Atoll studies each covered pixel for carrier
power or carrier-to-interference ratio. According to the measured carrier power or carrier-to-interference ratio, Atoll deduces
the maximum throughput available on that pixel through the throughput vs. C or throughput vs. C/I curves of the GPRS/EDGE
configuration.
The throughput per timeslot per pixel TPTS, Pixel can be either a function of carrier power C, or carrier power C and the carrierto-interference ratio C/I, depending on the user-defined traffic analysis RF conditions criteria. Therefore,
TP TS Pixel = f C
Or
C
TP TS Pixel = f C and TP TS Pixel = f ---
i
The required availability parameter defines the percentage of pixels within the coverage area of the transmitter that must
satisfy the minimum throughput condition. This parameter renders user-manageable flexibility to the throughput
requirement constraint.
To calculate the minimum throughput reduction factor for the transmitter, Atoll computes the minimum throughput
reduction factor for each pixel using the formula:
179

TP user min
RF min Pixel = ---------------------------------------------------TP TS Pixel TS Terminal
Once the minimum reduction factor for each pixel is known, Atoll calculates the global minimum reduction factor that is
satisfied by the percentage of covered pixels defined in the required availability. The following example may help in
understanding the concept and calculation method.
Example: Let the total number of pixels, covered by a subcell S, be 1050. The reliability level set to 90%. This implies that the
required minimum throughput for the given service will be available at 90% of the pixels covered. This, in turn, implies that
there will be a certain limit on the reduction factor, i.e. if the actual reduction factor in that subcell becomes less than a
minimum required, the service will not be satisfactory.
Atoll computes the minimum reduction factor at each pixel using the formula mentioned above, and outputs the following
results:
RFmin
Number of pixels
0.3
189
0.36
57
0.5
20
0.6
200
0.72
473
0.9
23
0.98
87
So for a reliability level of 90%, the corresponding RFmin will be the one provided at least 90% of the pixels covered, i.e. 945
pixels. The corresponding value of the resulting RFmin in this example hence turns out to be 0.9, since this value covers 962
pixels in total. Only 87 of the covered pixels imply an RFmin of 0.98. These will be the pixels that do not provide satisfactory
service.
This calculation is performed for each service type available in the subcell coverage area. The final minimum throughput
reduction factor is the highest one amongst all calculated for each service separately.
The minimum throughput reduction factor RFmin value is a minimum requirement that must be fulfilled by the network
dimensioning process when the Reduction Factor KPI is selected in the dimensioning model.
Figure 3.12: Minimum Throughput Reduction Factor
3.7.2.2.5
Step 5: Served PS Traffic

Atoll calculates the served packet switched traffic using the number of timeslots available to carry the packet switched traffic
demand. As the result of the above iterative step, Atoll always finds the best possible answer in terms of number of timeslots
required to carry the packet switched traffic demand unless the requirement exceeds the maximum limit on the number of
the packet switched traffic timeslots defined in the dimensioning model properties. Hence, there is no packet traffic overflow
unless the packet switched traffic demand requires more TRXs than the maximum allowed
3.7.2.2.6
Step 6: Total Traffic Load

This step calculates the final result of the dimensioning process, i.e. the total traffic load. The total traffic load L is calculated as:
180

AT332_TRR_E0
ST C + ST P
L = --------------------------------------------------------------------------------TS C dedicated + TS P dedicated + TS S
,
STC is the served circuit switched traffic

STP is the served packet switched traffic
TSC, dedicated is the number of dedicated circuit switched timeslots
TSP, dedicated is the number of dedicated packet switched timeslots
TSS is the number of shared timeslots
3.8 Key Performance Indicators Calculation

This feature calculates the current values for all circuit switched and packet switched Key Performance Indicators as a measure
of the current performance of the network. It can be used to evaluate an already dimensioned network in which recent traffic
changes have been made in limited regions to infer the possible problematic areas and then to improve the network
dimensioning with respect to these changes.
The concept of this computation is the inverse of that of the dimensioning process. In this case, Atoll has the results of the
dimensioning process already committed and known. Atoll then computes the current values for all the KPIs knowing the
number of required TRXs, the respective numbers of shared and dedicated timeslots and the circuit switched and packet
switched traffic demands.
The computation algorithm utilizes the parameters set in the dimensioning model properties and the quality curves for the
throughput reduction factor, delay and the blocking probability.
The following conventional relations apply:
If,
TSC, dedicated is the number of timeslots dedicated to the circuit switched traffic,
TSP, dedicated is the number of timeslots dedicated to the packet switched traffic,
TSS is the number of shared timeslots for a transmitter,
Then, the number of timeslots available for the circuit switched traffic, TSC, is defined as:
TS C = TS S + TS C dedicated
And the number of timeslots available for the packet switched traffic, TSP, is given by:
TS P = TS S + TS P dedicated
3.8.1 Circuit Switched Traffic

For each subcell, Atoll has already calculated the effective traffic overflow rate and the blocking rate during the dimensioning
process. Also knowing the circuit switched traffic demand, TDC, and the number of timeslots available for circuit switched
traffic, TSC, the blocking probability can be easily computed using the Erlang formulas or tables.
3.8.1.1 Erlang B
Under the current conditions of circuit switched traffic demand, TDC, and the number of timeslots available for the circuit
switched traffic, TSC, the percentage of blocked circuit switched traffic can be computed through:
TS C
TD C
-------------------- TS C !
% of blocked traffic = -------------------------TS
C
TD C
---------------k!
In a network dimensioning based on Erlang B model, the circuit switched traffic overflow rate, OC, is the same as the
percentage of traffic blocked by the subcell calculated above.
3.8.1.2 Erlang C
Similarly, under the current conditions of circuit switched traffic demand, TDC, and the number of timeslots available for the
circuit switched traffic, TSC, the percentage of delayed circuit switched traffic can be computed through:
181


TS
TD C
% of traffic delayed = ---------------------------------------------------------------------------------------------------------TS 1
C
TD C
TS
TD
+ TS C ! 1 --------C-
TS C
TD C
--------------k!
If the circuit switched traffic demand, TDC, is higher than the number of timeslots available to accommodate circuit switched
traffic, the column for this result will be empty signifying that there is a percentage of circuit switched traffic actually being
rejected rather than just being delayed under the principle of Erlang C model.
The circuit switched traffic overflow rate, OC, will be 0 if the circuit switched traffic demand, TDC, is less than the number of
timeslots available for the circuit switched traffic, TSC.
If, on the other hand, the circuit switched traffic demand, TDC, is higher than the number of timeslots available to carry the
circuit switched traffic, TSC, then there will be a certain percentage of circuit switched traffic that will overflow from the
subcell. This circuit switched traffic overflow rate, OC, is calculated as:
TD C TS C
O C = ----------------------TD C
3.8.1.3 Served Circuit Switched Traffic

The result of the above two processes will be a traffic overflow rate for the circuit switched traffic for each subcell, OC. The
served circuit switched traffic, STC, is calculated as:
ST C = TD C 1 O C
3.8.2 Packet Switched Traffic

Identifying the total traffic demand, TDT, (circuit switched traffic demand + packet switched traffic demand) as:
TD T = TD C + TD P
The following two cases can be considered.
3.8.2.1 Case 1: Total Traffic Demand > Dedicated + Shared Timeslots

In the case the total number of timeslots available is less than the total traffic demand, there will be packet switched data
traffic that will be rejected by the subcell as it will not be able to accommodate it. The following results are expected in this
case:
3.8.2.1.1
Traffic Load
The traffic load will be 100%, as the subcell will have more traffic to carry than it can. This implies that the system will be
loaded to the maximum and even saturated. Hence the user level quality of service is bound to be very unsatisfactory.
3.8.2.1.2
Packet Switched Traffic Overflow

In a 100% loaded, or even saturated subcell, the packet switched data calls will start being rejected because of shortage of
available resources. Hence there will be a perceptible packet switched traffic overflow in this subcell, OP. This overflow rate
is calculated as show below:
TS C dedicated + TS P dedicated + TS S ST C
O P = 1 ----------------------------------------------------------------------------------------------------------- 100
TD P
3.8.2.1.3
Throughput Reduction Factor

The resulting throughput reduction factor for a 100% loaded or saturated subcell will be 0. Hence, the throughput perceived
by the packet switched service user will be 0, implying a very bad quality of service.
3.8.2.1.4
Delay
Again for a 100% loaded or saturated subcell, the delay at the packet switched service user end will be infinite as there is no
data transfer (throughput = 0).
3.8.2.1.5
All the data packets will be rejected by the system since it is saturated and has no free resources to allocate to incoming data
packets. Hence, the blocking probability will be 100%.
182

AT332_TRR_E0
3.8.2.1.6
Served Packet Switched Traffic

With the packet switched data traffic overflowing from the subcell, there will be a part of that traffic that is not served. The
served packet switched data traffic, STP, is calculated on the same principle as the served circuit switched traffic:
ST P = TD P 1 O P
3.8.2.2 Case 2: Total Traffic Demand < Dedicated + Shared Timeslots

In the case the total traffic demand is less than the number of timeslots available to carry the traffic, the subcell will not be
saturated and there will be some deducible values for all the data KPIs. In a normally loaded subcell, the packet switched data
traffic will have no overflow percentage. This is due to the fact that the packet switched data traffic is rather placed in a waiting
queue than be rejected.
Therefore, there will be a within limits packet switched traffic load, LP, calculated as under:
ST C TS C dedicated + TD P
Timeslots
L P = -----------------------------------------------------------------------------------TS P
The second parameter for computing the KPIs from the quality curves of the dimensioning model is the number of equivalent
timeslots available for the packet switched data traffic, NP, which is calculated in the same manner as in the dimensioning
process as well:
TS P
N P = ----------------------TS Terminal
These parameters calculated, now Atoll can compute the required KPIs through their respective quality curves.
3.8.2.2.1
Traffic Load
The traffic load is computed knowing the total traffic demand and the total number of timeslots available to carry the entire
traffic demand:
TD T
Traffic Load = --------------------------------------------------------------------------------TS C dedicated + TS P dedicated + TS S
3.8.2.2.2
Packet Switched Traffic Overflow

In a normally loaded subcell, no packet switched data calls will be rejected. The packet switched traffic overflow will,
therefore, be 0.
3.8.2.2.3
Throughput Reduction Factor

The resulting throughput reduction factor for a normally loaded subcell is calculated through the throughput reduction factor
quality curve for given packet switched traffic load, LP, and number of equivalent timeslots, NP.
3.8.2.2.4
Delay
The resulting delay the subcell is calculated through the delay quality curve for given packet switched traffic load, LP, and
number of equivalent timeslots, NP.
3.8.2.2.5
The resulting blocking probability for a normally loaded subcell is calculated through the blocking probability quality curve for
given packet switched traffic load, LP, and number of equivalent timeslots, NP.
3.8.2.2.6
Served Packet Switched Traffic

As there is no overflow of the packet switched traffic demand from the subcell under consideration, the served packet
switched traffic will be the same as the packet switched traffic demand:
ST P = TD P
183

3.9 Simulations
Once you have modelled the network services and users and have created traffic maps, you can create simulations. The simulation process consists of two steps:
1. Obtaining a realistic user distribution: Atoll generates a user distribution using a Monte Carlo algorithm; this user
distribution is based on the traffic database and traffic maps and is weighted by a Poisson distribution between simulations of the same group.
Each user is assigned a service, a mobility type, and an activity status by random trial, according to a probability law
that uses the traffic database.
The user activity status is an important output of the random trial and has direct consequences on the next step of
the simulation and on the network interferences. A user can be either active or inactive. Both active and inactive users
consume radio resources and create interference.
Finally, another random trial determines user positions in their respective traffic zone (possibly according to the clutter weighting and the indoor ratio per clutter class).
2. Modelling network regulation mechanisms: Atoll manages the GSM resources as described in "Radio Resource Management in GSM" on page 184
3.9.1 Radio Resource Management in GSM

In this section, the following are explained:
"MSA (Mobile Station Allocation) Definition" on page 131

"GSM Simulation Process" on page 184.
3.9.1.1 GSM Simulation Process

Figure 3.13 shows the GSM simulation algorithm. The specific simulation process in GSM consists of the following steps:
184

AT332_TRR_E0
Figure 3.13: GSM simulation algorithm

For each simulation, the simulation process,
1. It sets initial values for the following parameters:
a. Mobile transmission power is set to the maximum mobile power.
b. Cell traffic loads for each MSA and transmitter are set to their average current value in the Transmitters table (one
traffic load value per subcell).
For each iteration k, the simulation process
2. For each circuit-switched mobile
a. Determines the server and the MSA to which the circuit-switched mobile is attached.
b. Determines the downlink C/(I+N) for each of these mobiles as explained in "DL Carrier-to-Interference Ratio Calculation" on page 131.
c. Determines the uplink C/(I+N) for each of these mobiles as explained "UL C/I Evaluation" on page 156
d. Determines MSA codec modes in downlink as explained in "Calculations Based on C/(I+N)" on page 148 part of
"CQI Calculation Without Ideal Link Adaptation" on page 147.
e. Determines MSA codec modes in uplink as explained in "Codec Mode Selection" on page 159.
f.
Performs the corresponding target power controls.
185

See detailed information in "Servers Selection" on page 187 and "Codec Mode Assignment and DL Power Control" on
page 187.
3. For each packet-switched mobile
a. Determines the server and the MSA to which the packet-switched mobile is attached.
b. Determines the downlink C/(I+N) for each of these mobiles as explained in "DL Carrier-to-Interference Ratio Calculation" on page 131.
c. Determines the uplink C/(I+N) for each of these mobiles as explained "UL C/I Evaluation" on page 156
d. Determines MSA coding scheme in downlink as explained in "Calculations Based on C/(I+N)" on page 137 part of
"Coding Scheme Selection and Throughput Calculation Without Ideal Link Adaptation" on page 137.
e. Determines MSA coding scheme in uplink as explained in "Coding Scheme Selection" on page 158.
f. Evaluates the number of necessary timeslots to reach the minimum downlink and uplink throughput demands
(defined in the requested service) of the users randomly ranked.
g. Performs the corresponding target power controls. The number of timeslots in DL and UL are obviously not linked.
See detailed information in "Servers Selection" on page 187 and "Coding Scheme Assignment, Throughput Evaluation
and DL Power Control" on page 188.
4. It equally shares the remaining resources to packet-switched users who did not reach their maximum throughput
demands. Resources and throughputs are finally assigned to each packet-switched user.
See detailed information in "Codec Mode Assignment and DL Power Control" on page 187.
5. It updates the traffic loads, Half-Rate traffic ratios, DL power control gains and DTX gains of all the subcells according
to the resources in use and the total resources.
See detailed information in "Subcell Traffic Loads Management" on page 189, "Half-Rate Traffic Ratio Management"
on page 189, "DL Power Control Gain Management" on page 189 and "DTX DL Gain Management" on page 190.
6. It updates the UL traffic loads of all the subcells and the UL noise rises of all the TRXs according to the resources in use
and the total resources.
7. Performs the convergence test to see whether the differences between the current and the new loads are within the
convergence thresholds.
The convergence criteria are evaluated at the end of each iteration k, and can be written as follows:
Subcell i
TL DL GSM
Subcell i
TL UL GSM
Subcell
Max
Subcell
Subcell
i
i
TL
TL UL GSM
UL GSM k
All Subcell
Max
TRX
TRX
i
i
Max NR UL GSM NRUL GSM
k
All TRX
i
Subcell i
Req
, PCG DL GSM
k 1
Subcell i
Req
k 1
k 1
Subcell i
If TL DL GSM
Subcell
Subcell
k 1
i
i
PCG
DL GSM k PCG DL GSM
All Subcell
TRX i
NR UL GSM
Max
Subcell i
PCG DL GSM
Subcell
i
i
TL
TL DL GSM
DL GSM k
All Subcell
, TL UL GSM
TRX i
Req
and NR UL GSM
are the simulation convergence
thresholds defined when creating the simulation, Atoll stops the simulation in the following cases.
Convergence: Simulation has converged between iteration k - 1 and k if:
Subcell i
Subcell i
TL DL GSM TL DL GSM
k
TRX i
Subcell i
Req
Subcell i
AND PCG DL GSM PCG DL GSM

k
Subcell i
Req
Subcell i
AND TL UL GSM TL UL GSM

k
Req
AND
TRX i
NR UL GSM NRUL GSM

k
Req
No convergence: Simulation has not converged even after the last iteration, i.e., k = Max Number of Iterations defined
Subcell
Subcell
Subcell i
Subcell i
TL UL GSM TL UL GSM
k
186
TRX i
Req
Subcell
when creating the simulation, if: TL DL GSM TL DL GSM

TRX i
OR NR UL GSM NRUL GSM

k
Req
Subcell
OR PCG DL GSM PCGDL GSM
Req
Req
OR

AT332_TRR_E0
8. Repeats the above steps (from step 2.) for the iteration k+1 using the new calculated loads as the current loads until
convergence.
3.9.1.2 Servers Selection

For a given network, the service areas of each transmitter are evaluated in the same way than an HCS server study with 0 dB
margin. In other words, each pixel, is covered by the best server of each HCS layer, assuming the received signal strength is
greater than the reception threshold defined on that layer.
In addition to the coverage condition above, for a given mobile distribution, a mobile might be served by a transmitter if its
mobility (as assigned by Atoll at the beginning of the simulation) does not exceed the maximum speed permitted on that layer.
Finally the frequency band(s) in use in the transmitter have to be supported by the user terminal.
In none of these conditions are fulfilled, the mobile is rejected with the condition "No Coverage".
If these conditions are fulfilled, as a result, each mobile then has a list of potential servers, each server being on a different
HCS layer. For each mobile list, Atoll sorts the potential servers according to their HCS layer priority in decreasing order.
On the very first iteration of the simulation, the mobile selects the highest priority transmitter. During the iterative process,
if the mobile is regularly rejected from the highest priority transmitter, it will select the second highest priority transmitter
and so on, until convergence.
In addition, if the mobile is rejected from a layer, even after convergence, the algorithm will try to attach this mobile to a lower
priority layer until no solution can be found.
3.9.1.3 Codec Mode Assignment and DL Power Control

Two types of services can be assigned to users: circuit-switched and packet-switched ones. The network has been set up and
dimensioned in order to first serve circuit services, and then to serve packet services with the remaining resources.
When serving a circuit-switched user, depending on the computed radio conditions at the server location, a codec mode is
assigned to a user. Depending on this codec mode, the user will use either an entire timeslot (any AMR, EFR or FR codec mode)
or half a timeslot (HR codec mode).
As explained in "GSM Simulation Process" on page 184, the resource element assigned to a mobile station is an MSA.
Depending on the assigned MSA, the level of quality at the mobile might be different, and consequently, its served codec
mode so as the required number of timeslots.
Assuming a server is selected for each mobile, several MSAs are candidate. For each candidate MSA, a codec mode study is
run, using the computed C/(I+N) and based on the user terminal and mobility (See "Calculations Based on C/(I+N)" on
page 148 for more information). For each MSA, a codec mode is obtained. For each mobile, the list of candidate codec modes
is saved.
At the beginning of a simulation iteration, no traffic is attached to MSAs. Their load starts from 0 and is increased as traffic
increases and mobiles are attached to them. For a given user, within his MSA list, the MSA having currently the lowest load is
selected and, as a consequence, the load of this MSA is now increased. The effect of this mechanism results in a load balancing
of MSAs within a transmitter.
When MSAs are almost full, Atoll selects the MSAs the most optimised in term of timeslot occupancy. As an example, to
optimise the resource allocation, a codec mode costing half a timeslot might be chosen instead of a codec mode costing an
entire timeslot in the case the MSA with the lowest cost would have been chosen.
This mechanism is then reproduced for all the users requesting a circuit-switched service.
i
For each MSA k, the assigned codec mode i corresponds to a quality target: C I Target . Due to the radio conditions, and using
k
the victim max power, a C I Max is obtained.

k
If C I Max C I Target , no codec mode can be served and the mobile is rejected with the condition "No Service".
If C I Max C I Target , the corresponding codec mode is assigned to the mobile. If the MSA is on the BCCH, no power
k
control is applied. For any other TRX type, Atoll evaluates the minimum required power P Min in order to reduce the quality
i
at the users terminal to C I Target for the assigned MSA k.

The maximum allowed power reduction is set at 30 dB by default. This means that the
power cannot be reduced by more than 30 dB from the initial to final C/(I+N), after power
control.
The power control is considered achieved when the final C/(I+N) is at less than 1 dB from
i
the C I Target .
187

To summarise, at this step, each circuit-switched user is assigned a MSA, a codec mode, a corresponding number of timeslots
i
(0.5 or 1) and a corresponding minimum required power to get the C I Target of the served MSA.
Then, if the user has been dropped as inactive at the beginning of the simulation, his corresponding number of timeslots is
consumed but no DL power is considered for this specific user. Inactive users only participate in the timeslot management but
do not affect DL power.
Finally, if the user has been dropped as active at the beginning of the simulation, both timeslots and powers have to be
considered to make him connected.
3.9.1.4 Coding Scheme Assignment, Throughput Evaluation and DL Power

Control
After having served the circuit traffic over one iteration, the algorithm now tries to serve packet-switched traffic.
When serving a packet-switched user, depending on the computed radio conditions at the server location, a coding scheme
is assigned to a user and a throughput per timeslot is obtained. Then some timeslots are assigned to each packet-switched
service user in order to obtain a throughput between the min and the max DL throughput demand per user defined in the
considered service properties.
As explained in "MSA (Mobile Station Allocation) Definition" on page 131, the resource element assigned to a mobile station
is an MSA. Depending on the assigned MSA, the level of quality at the mobile might be different, and consequently, its served
coding scheme so as the required number of timeslots to get a certain throughput demand. For packet-switched traffic, the
timeslot Assignment is realised in two steps. In the first step, Atoll tries to allocate the minimum throughput demand of the
service. In the second step, using remaining resources (timeslots), Atoll tries to allocate more throughput up to the maximum
throughput demand of the service. If a user cannot get its minimum throughput demand for insufficient number of available
timeslots, the user is rejected with the condition "Resource Saturation".
Assuming a server is selected for each mobile, several MSAs are candidate. For each candidate MSA, a coding scheme study
is run, using the computed C/(I+N) and based on the user terminal and mobility (See "Calculations Based on C/(I+N)" on
page 139 for more information). For each MSA, a coding scheme is obtained, from which we get a throughput per timeslot.
As explained in "Packet Throughput and Quality Analysis: Application Throughput (kbps)" on page 144, the maximum of
timeslots the user can benefit is the minimum between the number of DL timeslots defined in the selected terminal and
service. Considering the minimum DL throughput demand for the service, one can estimate how many timeslots are needed
to get that throughput on each MSA. Then, Atoll only keeps the MSAs for which this number of timeslots is lower than the
number of timeslots supported (see above) and for which there is enough remaining timeslots. Then, for each mobile, the list
of candidate coding schemes is saved.
For a given user, within his MSA list, the MSA having currently the lowest load is selected and, as a consequence, the load of
this MSA is now increased. In the same way than for circuit traffic, the effect of this mechanism results in a load balancing of
MSAs within a transmitter.
This mechanism is then reproduced for all the users requesting a packet-switched service. At this step, each packet-switched
service has a coding scheme and, ideally, is supposed to be served his DL minimum throughput demand.
The second step of resources allocation for packet-switched traffic is to share the remaining resources between connected
users in order they get their maximum throughput demand. As an example, lets imagine than a MSA is already occupied as
follows:
2 TS for circuit-switched service users (3 users: 2 HR codec modes + 1 FR codec mode)

2.4 TS for packet-switched service users after the first step (2 users).
If this MSA is defined over a TCH subcell, its capacity is 8 TS. In other words, 4.4 TS have been used, and 3.6 TS remain. The
two packet-switched users have obtained their minimum throughput demand. In order to reach their maximum throughput
demand, the remaining TS are equally shared between the two connected users: 1.8 TS per user. If the first user can get his
maximum throughput demand with only 1.5 TS, the remaining 0.3 TS will be able to be used by the user. As a consequence,
this second user could benefit of 2.1 TS in order to get his maximum demand. If, finally, he only needs 1.3 TS to get this
demand, 0.8 TS remain unused for that MSA.
This mechanism of equally share of remaining resources is then applied for all the connected packet-switched service users
over all their MSAs.
j
For each MSA k, the assigned coding scheme j corresponds to a quality target: C I Target . Due to the radio conditions, and
k
using the victim max power, a C I Max is obtained.

k
If C I Max C I Target , no coding scheme can be served and the mobile is rejected with the condition "No Service".
188

AT332_TRR_E0
k
If C I Max C I Target , the corresponding coding scheme is assigned to the mobile. If the MSA is on the BCCH, no power
k
control is applied. For any other TRX type, Atoll evaluates the minimum required power P Min in order to reduce the quality
j
at the users terminal to C I Target for the assigned MSA k.

The maximum allowed power reduction is set at 30 dB by default. This means that the
power cannot be reduced by more than 30 dB from the initial to final C/(I+N), after power
control.
The power control is considered achieved when the final C/(I+N) is at less than 1 dB from
i
the C I Target .
To summarise, at this step, each packet-switched user is assigned a MSA, a coding scheme, a corresponding number of
j
timeslots (which might not be an integer value) and a corresponding minimum required power to get the C I Target of the
served MSA.
3.9.1.5 Subcell Traffic Loads Management

When circuit-switched and packet-switched traffic have been served or rejected, Atoll performs an update on several
parameters. The first parameter to be updated are the subcell DL and UL traffic loads. Considering that subcell loads are values
which are unique per traffic pool (e.g. BCCH and TCH subcells belong to the same traffic pool because they are in charge of
the same traffic area) in DL and in UL, the number of timeslots necessary to connect the traffic have to be summed up over
the several MSAs over a same traffic pool.
For the traffic pool TP i , the subcell DL traffic load is computed as follows:
TL TP
TS DL used
MSA TP
i DL
i
= ------------------------------------------------- where the number of DL TS available for a BCCH subcell is 7 and 8 for any other subcell.
TS DL available
MSA
TP i
The DL traffic load value is then assigned to all the subcells of a same traffic pool.
For the traffic pool TP i , the subcell UL traffic load is computed as follows:
TL TP
TS UL used
MSA TP
i UL
i
= ------------------------------------------------- where the number of UL TS available for a BCCH subcell is 7 and 8 for any other subcell.
TS UL available
MSA TP
The UL traffic load value is then assigned to all the subcells of a same traffic pool.
3.9.1.6 Half-Rate Traffic Ratio Management

The second parameter at the end of an iteration is the Half-rate traffic ratio. This is the percentage of half-rate voice traffic in
the subcell. This value is used to calculate the number of timeslots required to respond to the voice traffic demand and is
evaluated per traffic pool. This value referring to voice traffic only, circuit-switched users only are taken into account in its
evaluation.
HR users
MSA TP
i
HR RATIO TP = ------------------------------------.
users represents HR and FR circuit-switched service users.
i
users MSATP
MSA TP
The Half-Rate traffic ratio is then assigned to all the subcells of a same traffic pool.
3.9.1.7 DL Power Control Gain Management

At the end of each iteration, the subcell DL power control gain is evaluated by taking into account all the connected users:
active and inactive circuit-switched service users (assuming each inactive user does not cost any DL power but only
some timeslots)
all packet users
189

From the minimum required powers evaluated at the end of "Codec Mode Assignment and DL Power Control" on page 187
and "Coding Scheme Assignment, Throughput Evaluation and DL Power Control" on page 188 in order to get respectively the
appropriate codec modes and coding schemes without any excess of unneeded power, an average minimum required power
is obtained for each mobile connected to the subcell S as follows:
k
PMin TSi
i
i-------------------------------S
- = P Moy
TS i
where i are the mobiles connected to the subcell S, over its MSAs
The ratio PCG
P Max
= ---------------S (in dB), where P Max
P Moy
is the max power of the considered subcell, represents the mean power
control gain, due to active and inactive users, which can be assigned to the subcell.
It is essential to note that there is no power control on the BCCH and, consequently, the mean power control gain on the BCCH
is 0.
3.9.1.8 DTX DL Gain Management

A certain gain representing inactive circuit-switched service users has also to be evaluated. In "DL Power Control Gain
Management" on page 189, the mean DL power control gain concerns both active and inactive users. The DTX gain models
the fact that inactive circuit-switched users, even if they are connected to the network, do not produce the same level of
interference than active circuit-switched users.
From the minimum required powers evaluated at the end of "Codec Mode Assignment and DL Power Control" on page 187 in
order to get the appropriate codec modes without any excess of unneeded power, an average minimum required power is
obtained for each circuit-switched active mobile connected to the subcell S as follows:
ki
active
P Min
TS i
i active S
active
------------------------------------------------------------ = P Moy
TS i
active
where i active are the circuit-switched active mobiles connected to the subcell S, over
active
i active
its MSAs
P Moy
S
The ratio -------------------------- (in dB), where P Moy
P Moy
S active
is average requested power defined in "DL Power Control Gain Management" on
page 189 above, represents the DTX gain, due to circuit-switched active users, which can be assigned to the subcell.
3.9.1.9 GSM Simulation Results

At the end of the simulations, an active user can be connected in DL if:
he has a serving cell assigned,

For a circuit-switched (resp. packet-switched) service, he has a codec mode (resp. coding scheme) corresponding to
his activity status,
he is not rejected due to resource saturation.
If a user is rejected during server determination, the cause of rejection is "No Coverage". If a user is rejected because quality
is too low to obtain any codec mode or coding scheme, the cause of rejection is "No Service". If a user is rejected because he
cannot be allocated a sufficient number of resources to obtain its codec mode or coding scheme, the cause of rejection is
"Resource Saturation," i.e., all of the cells resources were used up by other users.
Considering only the connected traffic at the end of the GSM part of the simulation process, the main results obtained are:
At the subcell level
Subcell traffic loads (UL and DL)

DL Power control gains
DTX gains
Half-rate traffic ratios
At the TRX level
intra-technology UL noise rises
Subcell traffic loads, DL Power control gains and intra-technology UL noise rises can be used as input for GSM quality-based
coverage predictions.
190

AT332_TRR_E0
3.10 Automatic Neighbour Allocation

The intra-technology neighbour allocation algorithm takes into account all the TBC transmitters. It means that all the TBC
transmitters of the .atl document are potential neighbours.
The transmitters to be allocated will be called TBA transmitters. They must fulfil the following conditions:
They are active,

They satisfy the filter criteria applied to the Transmitters folder,
They are located inside the focus zone,
They belong to the folder on which allocation has been executed. This folder can be either the Transmitters folder or
a group of transmitters or a single transmitter.
Only TBA transmitters may be assigned neighbours.

zone.
3.10.1 Neighbour Allocation for All Transmitters

We assume a reference transmitter A and a candidate neighbour, transmitter B.
When automatic allocation starts, Atoll checks following conditions:
The distance between both transmitters must be less than the user-definable maximum inter-site distance. If the
distance between the reference transmitter and the candidate neighbour is greater than this value, then the
candidate neighbour is discarded.
When allocation is based on distance only, Atoll considers the effective distance, which corresponds to the real intertransmitter distance weighted by the azimuths of antennas. For information on the effective distance calculation, see
"Appendix: Calculation of the Inter-Transmitter Distance" on page 195. Otherwise, it takes the real distance.
You can force Atoll to compare the Max inter-site distance with the real inter-site
distance by adding the following lines in Atoll.ini:
[Neighbours]
RealInterSiteDistanceCondition=1
The calculation options:

Force co-site transmitters as neighbours: This option enables you to force transmitters located on the reference
transmitter site in the candidate neighbour list. This constraints can be weighted among the others and ranks the
neighbours through the importance field.
Force adjacent transmitters as neighbours: This option enables you to force transmitters geographically adjacent
to the reference transmitter in the candidate neighbour list. This constraint can be weighted among others and
ranks the neighbours through the importance field.
191

Adjacency criterion:
Geographically adjacent transmitters are determined on the basis of their Best Server
coverages in 2G (GSM GPRS EDGE) projects. More precisely, a transmitter TXi is
considered adjacent to another transmitter TXj if there exists at least one pixel of TXi Best
Server coverage area TXj is the 2nd Best Server. The ranking of the adjacent neighbour
transmitter increases with the number of these pixels. The figure below shows the above
concept.
When the adjacency option is checked, adjacent cells are sorted and listed from the
most adjacent to the least, depending on the above criterion. Adjacency is relative
to the number of pixels satisfying the criterion.
Force adjacent layers as neighbours: If selected, Atoll adds all the transmitters adjacent across network layers to
the reference transmitter to the candidate neighbour list. The weight of this constraint is always the average of
the Min and Max values defined for the adjacency factor. This weight is used to calculate the rank of each
neighbour and its importance. Transmitters are considered adjacent across layers if they belong to different layers
and have a coverage overlap of at least one pixel.
Force symmetry: This option enables user to force the reciprocity of a neighbourhood link. Therefore, if the
reference transmitter is a candidate neighbour of another transmitter, the latter will be considered as candidate
neighbour of the reference transmitter.
If the neighbours list of a transmitter is full, the reference transmitter will not be added
as a neighbour of that transmitter and that transmitter will be removed from the
reference transmitters neighbours list. You can force Atoll to keep that transmitter in
the reference transmitters neighbours list by adding the following option in the Atoll.ini
file:
[Neighbours]
DoNotDeleteSymmetrics = 1
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore,
you may force/forbid a transmitter to be candidate neighbour of the reference transmitter.
Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current
neighbours and carries out a new neighbour allocation. If not selected, the existing neighbours are kept.
If the Use Coverage Conditions check box is selected, there must be an overlapping zone ( S A S B ) with a given cell
edge coverage probability. Otherwise, only the distance criterion is taken into account.
The overlapping zone ( S A S B ) is defined as follows:
192
SA is the area where the received signal level from transmitter A is greater than a minimum signal level. SA is the
coverage area of reference transmitter A restricted between two boundaries; the first boundary represents the
start of the handover area (best server area of A plus handover margin named handover start) and the second
boundary shows the end of the handover area (best server area of A plus the margin called handover end).
SB is the coverage area where the candidate transmitter B is the best server.

AT332_TRR_E0
SA SB
- 100 ) if the option Take into account Covered Area is
Atoll calculates either the percentage of covered area ( ----------------SA
selected, or the percentage of traffic covered on the overlapping area S A S B for the option Take into account Covered
Traffic. Then, it compares this value to the % minimum covered area (minimum percentage of covered area for the option
Take into account Covered Area or minimum percentage of covered traffic for the option Take into account Covered
Traffic). If this percentage is not exceeded, the candidate neighbour B is discarded.
The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see
number 4 below).
Figure 3.14: Overlapping Zones

Atoll uses traffic map(s) selected in the default traffic analysis in order to determine the
percentage of traffic covered in the overlapping area.
The importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason,
and to quantify the neighbour importance.
Atoll lists all neighbours and ranks them by importance value so as to eliminate some of them from the neighbour list if the
maximum number of neighbours to be allocated to each transmitter is exceeded. If we consider the case for which there are
15 candidate neighbours and the maximum number of neighbours to be allocated to the reference transmitter is 8. Among
these 15 candidate neighbours, only 8 (having the highest importances) will be allocated to the reference transmitter.
In the Results part, Atoll only displays the transmitters for which it finds new neighbours. For these transmitters, it provides
the list of neighbours, the number of neighbours and the maximum number of neighbours allowed for each transmitter. In
addition, it indicates the importance (in %) of each neighbour and the allocation reason. Therefore, a neighbour may be
marked as exceptional pair, co-site, intra-HCS and inter-HCS adjacent, coverage or symmetric. For neighbours accepted for
co-site, intra-HCS or inter-HCS adjacency and coverage reasons, Atoll displays the percentage of area meeting the coverage
conditions (or the percentage of covered traffic on this area) and the corresponding surface area (km2) (or the traffic covered
on the area in Erlangs), the percentage of area meeting the adjacency conditions and the corresponding surface area (km2).
Finally, if cells have previous allocations in the list, neighbours are marked as existing.
3.10.2 Neighbour Allocation for a Group of Transmitters or One

Transmitter
In this case, Atoll allocates neighbours to:
TBA transmitters,
193

Neighbours of TBA transmitters marked as exceptional pair, intra-HCS or inter-HCS adjacent and symmetric,
Neighbours of TBA transmitters that satisfy coverage conditions.
Automatic neighbour allocation parameters are described in "Neighbour Allocation for All Transmitters" on page 191.
3.10.3 Neighbour Importance Calculation

The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf. table
below); this value varies between 0 and 100%.
Neighbourhood cause
When
Importance value
Existing neighbour
Only if the Delete existing neighbours option is not selected

and in case of a new allocation
Existing importance
Exceptional pair
Only if the Force exceptional pairs option is selected
100 %
Co-site transmitter
Only if the Force co-site cells as neighbours option is

selected
Importance Function (IF)
Adjacent transmitters
Only if the Force intra-HCS adjacents as neighbours option is

selected
Adjacent layer
Only if the Force inter-HCS adjacents as Neighbours option is

selected
Neighbourhood relationship that

fulfils coverage conditions
Only if the % minimum covered area is exceeded
Symmetric neighbourhood
relationship
Only if the Force neighbour symmetry option is selected
Except the case of forced neighbours (importance = 100%), priority assigned to each neighbourhood cause is determined
using the Importance Function (IF). The IF considers the following factors for calculating the importance:
The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference
transmitter.
d Di = 1 ---------d max
d is the real inter-transmitter distance if coverage conditions are considered. For calculations based on distance only,
it corresponds to the effective distance (in m), which is the real inter-transmitter distance weighted by the azimuths
of antennas. For information on the effective distance calculation, see "Appendix: Calculation of the Inter-Transmitter
Distance" on page 195.
d max is the maximum inter-site distance defined in the Neighbour Importance Evaluation dialogue.
You can force Atoll to consider the individual distances between reference cells and their
respective neighbour candidates by adding the following lines in Atoll.ini:
[Neighbours]
CandidatesMaxDistanceInImportanceCalculation=1
In this case, the maximum inter-site distance is the highest distance value between the
reference cell and its potential neighbours. It is different for each reference cell.
The co-site factor (C): a Boolean,

The adjacency factor (A): the percentage of adjacency,
The overlapping factor (O): the percentage of overlapping.
The minimum and maximum importance assigned to each of the above factors can be defined.
Factor
Min importance
Default value
Max importance
Default value
Distance factor (Di)
Min(Di)
1%
Max(Di)
10%
Overlapping factor (O)
Min(O)
10%
Max(O)
30%
Adjacency factor (A)
Min(A)
30%
Max(A)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%
The Importance Function is evaluated as follows:
194

AT332_TRR_E0
Neighbourhood cause
Importance Function
Resulting IF using the default values

from the table above
Coverage
Min(O)+Delta(O){Max(Di)(Di)+(100%Max(Di))(O)}+Min(Di)+Delta(Di)(Di)
10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
Adjacent layer
(Min(A)+Max(A))/2
45%
Adjacent transmitters
Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+
(100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di)
30%+30%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Co-site transmitters
Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+
60%+40%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Delta(X)=Max(X)-Min(X)
Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance
factor in the importance calculation.
If the Min and Max value ranges of the importance function factors do not overlap,
the neighbours will be ranked by neighbour cause. With the default values for
minimum and maximum importance fields, neighbours will be ranked in this order:
co-site neighbours, adjacent neighbours, and neighbours allocated based on coverage
overlapping.
If the Min and Max value ranges of the importance function factors overlap, the
neighbours may be ranked differently. There can be a mix of the neighbourhood
causes.
The default value of Min(O) = 1% ensures that neighbours selected for symmetry will
have an importance greater than 0%. With a value of Min(O) = 0%, neighbours
selected for symmetry will have an importance field greater than 0% only if there is
some coverage overlapping.
3.10.4 Appendix: Calculation of the Inter-Transmitter Distance

Atoll takes into account the real distance ( D in m) and azimuths of antennas in order to calculate the effective intertransmitter distance ( d in m).
d = D 1 + x cos x cos
x = 0.3% so that the maximum D variation does not exceed 1%.
Figure 3.15: Inter-Transmitter Distance Computation

The formula above implies that two cells facing each other will have a smaller effective distance than the real physical
distance. It is this effective distance that will be taken into account rather than the real distance.
3.11 AFP Appendices

3.11.1 The AFP Cost Function
The notations listed hereafter are used to describe the cost function:
TRG:
TRGs:
:
g :
Group of TRXs
Set of all the TRGs
If and only if
Size of any group g
195

ARFCN:
ARFCN
x :
A i g :
Set of all the frequencies

:
Set of all the subsets of frequencies

The largest integer
Number of times a group g 2
ARFCN
is assigned to TRGi in the assignment A
For example:
When i is NH, A i g = 1 g is a single member group containing one of the frequencies assigned at TRGi.
If |g| is not 1 or if g does not contain a frequency assigned at i, then A i g = 0 .
When i is BBH, A i g can be either 0 or equal to the number of TRXs in TRGi.

A i g = Number of TRXs in TRGi g is the set of frequencies assigned to TRXs of TRGi. (|g| = number of TRXs in
TRGi).
When we talk about "TRXs of i using g", and in the case of BBH, then there are |g| such virtual TRXs, each using
the entire group g and having a virtual MAIO [0, |g| - 1].
When i is SFH, A i g must be less than or equal to the umber of TRXs in TRGi. A i g = n g is the set of frequencies
assigned to n TRXs of TRGi.
We assume all the groups assigned to TRGi to have the same length.
TSi:
TLi:
Number of timeslots available for each TRX in TRGi

Traffic load of TRGi (calculated or user-defined)
TL i = #Erlangs of a single TRX in TRGi divided by TSi
TSUi:
CFi:
QMINi:
PMAXi:
REQi:
Downlink timeslot use ratio (due to DTX) at TRGi

Cost factor of TRGi (AFP Weight)
Minimum required quality (in C/I) at TRGi
Percentage permitted to have quality lower than QMINi at TRGi
Required number of TRXs at TRGi
A communication uses the group g in TRGi if its mobile allocation is g. The probability to be interfered is denoted by P i i' g A
(i is the TRX index). Different TRX indexes may have different MAIOs. P i i' g A is a function of the whole frequency
assignment. The precise definition of the term to be interfered is provided afterwards. The probability penalty due to
violating a separation constraint is P i i' g A . It is a function of the whole frequency assignment as well.
The term Atom will be used in the following context:
For two TRGs, i and k,
ATOM i ATOM k
i and k are synchronised, have the same HSN, the same MAL length and the same hopping mode.
NH TRGs or BBH TRGs are always in separate atoms. If two TRGs interfere but are not in the same atom, these can be taken
as unsynchronised. The quality of unsynchronised TRGs is a function of all possible frequency combinations. For synchronised
TRGs, pairs of frequencies emitted at the same time are known.
3.11.1.1 Cost Function

The Atoll AFP cost function is a TRX based cost and not an interference matrix entry based cost. It counts the impaired traffic
of the network TRXs in weighted Erlangs.
The cost function is reported to the user during the AFP progress with the help of its 5 components: mis , sep , comp ,
corr and dom .
= mis + sep + comp + corr + dom
,
mis represents the missing TRX cost component
sep represents the separation component
comp represents the additional cost component (interference, cost of changing a TRX)
196

AT332_TRR_E0
corr represents the corrupted TRX cost component

dom represents the out-of-domain frequency assignment cost component
mis =
MIS_TRX i TL i CF i TS i
CORR_TRX i TL i CF i TS i
DOM_TRX i TL i CF i TS i
i TRGs
corr =
i TRGs
dom =
i TRGs
sep
i TRGs
comp
ARFCN
g2
i' TRXs of i using g
i TRGs
' i i' g A TL i CF i TS i
ARFCN
g2
i' TRXs of i using g
'' i i' g A TL i CF i TS i
In the above equations,
i is the TRX index belonging to 0 1 ... A i g 1 .
MIS_TRX i is the number of missing TRXs for the subcell i.
MIS_TRX i = MAX 0 REQi
g2
A i g
ARFCN
is the cost value for a missing TRX. This value can vary between 0 and 10. The default cost value is set to 1 and can
be modified in the AFP module properties dialog box.
CORR_TRX i is the number of corrupted TRXs for the subcell i.
is the cost value of a corrupted TRX. This value can vary between 0 and 10. The default cost value is set to 10 and
can be modified in the AFP module properties dialog box.
DOM_TRX i is the number of TRXs, for the subcell i, having out-of-domain frequencies assigned.
is the cost value of a TRX with out-of-domain frequencies assigned. This value can vary between 0 and 1. The default
cost value is set to 0.5 and can be modified in the AFP module properties dialog box.
And, as mentioned earlier, a virtual TRX is considered in case of BBH.
If i is valid, the algorithm evaluates the cost of a valid TRX. This cost has two components, ' i i' g A and '' i i' g A .
' i i' g A is the separation violation probability penalty.
'' i i' g A is complementary probability penalty due to interference and the cost of modifying a TRX.
If the option Take into account the cost of all the TRXs available in the AFP module properties dialog box is selected,
then,
' i i' g A = P' i i' g A and '' i i' g A = P'' i i' g A
Or if the option Do not include the cost of TRXs having reached their quality target available in the AFP module
properties dialog box is selected, the algorithm compares P' i i' g A + P'' i i' g A with the quality target specified for
i, P MAX :
If P' i i' g A + P'' i i' g A P MAX ,
Then ' i i' g A = P' i i' g A and '' i i' g A = P'' i i' g A .
197

Otherwise,
Both ' i i' g A and '' i i' g A will be equal 0.
P' i i' g A is the same as ' i i' g A (separation violation probability penalty) and P'' i i' g A the same as '' i i' g A
(complementary probability penalty due to interference and the cost of modifying a TRX) in most cases. These are explained
in detail in the next sections.
3.11.1.2 Cost Components

Separation violation and interference cost components are described hereafter. Parameters considered in the cost function
components can be fully controlled by the user. Some of these parameters are part of the general data model (quality
requirements, percentage of interference allowed per subcell), while others (such as separation costs and diversity gains) can
be managed through the properties dialog box of the Atoll AFP module.
3.11.1.2.1
Separation Violation Cost Component

The separation violation cost component is evaluated for each TRX. Estimation is based on costs specified for the required
separations.
Let SEP_CONSTR i k denote the required separation constraint between TRGi and TRGk. Let Cost s z denote the user defined
separation penalty for a required separation s and actual separation z. SEP i k v is used instead of Cost SEP_CONSTR
i k z
as
abbreviation.
The AFP module properties dialog box takes probability percentages as inputs while this
document deals in probability values.
ii'kgg'k' is considered to be the effect of a separation violation on the i' th TRX of TRGi assigned the group g, caused by the
k' th TRX of TRGk assigned the group g' .
denotes the overall weight of the separation violation cost component. This value can be between 0 and 1, set to 1 by
default. It can be modified in the AFP module properties dialog box.
ik represents the weight of the specific separation constraint between i and k. This specific weight depends on the type of
separation violation and follows the following priority rule:
1. Exceptional pairs
2. Co-transmitters
3. Co-site
4. Neighbours
For example, if a pair of subcells are co-site and neighbours at the same time, they will be considered as co-site because higher
priority. Hence, ik of these subcells will be the weight of co-site relations. If only a neighbour relation exists between two
subcells, then ik will be further weighted by the neighbour relation importance. The value of ik remains between 0 and 1.
The default weights of each type of separation are available in the Separation cost tab.
If ATOM i ATOM k
SEP i k
f f'
fg
f' g'
Then ii'kgg'k' = ik ----------------------------------------- , which is same for all values of k.

g g'
If ATOM i = ATOM k
Then ii'kgg'k' = ik
SEPi k g g'
f_n
...
F_N
----------------------------------------------------------------------------F_N
In the above equations, F_N g is the number of frames in the MAL g. F_N g = g .
198

AT332_TRR_E0
Since F_N g = F_N g' , we shortly denote the two as F_N .
Let f_n denote the instantaneous frame number from 0 to F_N .

While = f_n + MAIO A
And = f_n + MAIO A
th
i g i'
k g' k'
modulo F_N and g is the frequency in g,
modulo F_N and g' is the
th
frequency in g.
In addition, frequencies belonging to a MAL with a low fractional load, and breaking a separation constraint, should not be
weighted equally as in a non-hopping separation breaking case. Therefore, the cost is weighted by an interferer diversity gain.
1
G i k g g' = ----------------------------------------------------------- 0,1 SEP_GAIN i k g g'
10
The separation gain, denoted by SEP_GAIN i k g g' is basically a function of the MAL length (and, of course, of the
hopping mode). With frequency hopping, the effects of DTX and traffic load become more significant (due to the consideration
of the average case instead of the worst case). For this reason, it is possible to consider these effects in
SEP_GAIN i k g g' through the relevant option available in the Advanced tab of the AFP module properties dialog box.
Without this option, the SEP_GAIN i k g g' is:
SEP_GAIN i k g g' = I_DIV g
I_DIV g is the user defined interferer diversity gain (dB) for a given MAL length. It is used in P i i' g A definition as well.
On the other hand, if this option is selected, the SEP_GAIN i k g g' becomes,
2 + ASYN_GAIN i k g'
SEP_GAIN i k g g' = I_DIV g +0,5 TSU_GAIN k min 10 4 + 2 + I_DIV g -----------------------------------------------------------------
4
1
TSU_GAIN k = log 10 -------------------------- ,
TL k TSU k
And ASYN_GAIN i k g' =
0
I_DIV( g'
if ATOM(i) = ATOM(k)
Otherwise
More than one separation violations may exist for a TRX. Many small G i k g g' and ' ii'kgg' have to be combined to form one
cost element, the P' i i' g A . This is done through iterating over all violating assignments and by summing up an equivalent
to the probability of not being violated while considering each separation violation as an independent probability event. This
sum is naturally limited to 100% of the TRX traffic, and is given by,
1 ii'kgg'k' G i k g g'
1
P' i i' g A =
k TRGs
ARFCN
g' 2
k' TRXs of k using g'
In the above formula, if k = i , then k' i' , so that interference with itself is not taken into account.
3.11.1.2.2
Interference Cost Component

The interference cost component is evaluated for each TRX. Its estimation is based on interference histograms calculated for
pairs of subcells. In addition, it takes into account frequency and interferer diversity gains and models frequency hopping and
gain due to DTX.
Interference histograms are described in User Manual (GSM GPRS EDGE project
management, GSM GPRS EDGE network optimisation, GSM GPRS EDGE generic AFP
management). Interference histograms can also be exported to files. For further
description, refer to "Interference" on page 204.
199

When estimating P'' i i' g A , the following problems are encountered:
The QMINi C/I quality indicator corresponds to the accumulated interference level of all interferers while the C/I
interference histograms correspond to pair-wise interferences.
Both QMINi and the histograms correspond to a single frequency. In case of a MAL containing more than one
frequencies, interferences on several different frequencies of a MAL must be combined.
This estimation, presented below, is the simplest possible as it solves the first problem by linear summation and
truncation at the value of 1 and it solves the second problem by averaging and adding the two diversity gains:
F_DIV g , the frequency diversity gain, and
I_DIV g , the interferer diversity gain.
Hereafter, denotes the global weight of interference cost component. This value can vary between 0 and 1 and is set to
0.35 by default, which can be modified in the AFP module properties dialog box.
Let F_N g be the number of frames in the MAL g. F_N g = g .
Let f_n denote the instantaneous frame number from 0 to F_N .
Let MAIO A
k g' j
be the jth MAIO of A k g' , j is one of the 0 1 ... A k g' 1 TRXs.
The value of MAIO A
k g' j
is one of 0 1 ... g'
If TRGk is NH, then MAIO A
k g' j
If TRGk is BBH, then MAIO A
k g' j
= 0.
= j.
As said earlier, in case of BBH, we consider g' virtual TRXs, the jth TRX has the MAIO j.
Let g i be the ith frequency in the group g.
Similar to the definition of ii'kgg'k' , ' ii'kgg'k' is defined as an interference event. ' ii'kgg'k' is the effect interference on the i' th
TRX of TRGi assigned the group g, caused by the k' th TRX of TRGk assigned the group g' .
If ATOM i ATOM k
Then ' ii'kgg'k' =
f g f' g'
C
Probability ----- Q_UB i k f f'
I ik
-------------------------------------------------------------------------g g'
f f' ADJ_SUP + INTERF_GAIN i k g g'
Q_UB i k f f' = QMIN i

If ATOM i = ATOM k
Then,
Since F_N g = F_N g' , these are both represented by F_N .
ii'kgg'k'
C
Probability ----- Q_UB i k f f'
I ik
-
=
------------------------------------------------------------------------F_N
f_n 0 1 ... F_N 1
,
f = g ,
f' = g' ,
= f_n + MAIO A
= f_n + MAIO A
i g i'
modulo F_N ,
k g' k'
Q_UB i k f f' = QMIN i
200
modulo F_N ,
f f' ADJ_SUP + INTERF_GAIN i k g g'

AT332_TRR_E0
Therefore, we have, P'' i i' g A = 1 1 P' i i' g A
1 ii'kgg'k' P' i i' g A
k TRGs
ARFCN
g' 2
k' TRXs of k using g'
In the above formula, if i = k , then k' i' , so that interference with itself is not taken into account.
The sum is limited to 100% of the TRX traffic. INTERF_GAIN i k g g' is quite similar to SEP_GAIN i k g g' . The only
difference is the frequency diversity gain, F_DIV g , added to SEP_GAIN i k g g' .
3.11.1.2.3
I_DIV, F_DIV and Other Advanced Cost Parameters

When combining interference effects (or separation violation effects) on different frequencies belonging to a MAL, the
following considerations should be taken into account:
1. Non-linearity of Frame Error Rate (FER) with respect to average C/I conditions and MAL length.
2. Interference Diversity Gain. This factor represents that the effect of average negative effects over user geographic
location are directly proportional to the MAL length.
3. Frequency Diversity Gain. This factor models the gain due to diversity of multi-path effects and should be applied to
the interference cost component only.
4. The fact that long MALs with synthesized hopping permit discarding the worst case estimation and include a gain due
to DTX and low traffic load at the interferer end.
The Advanced properties tab shown in the figure below facilitates modelling these effects.
Figure 3.16: The Advanced tab of the AFP module Properties dialog box
The Interference Diversity Gain table lists the values of I_DIV provided as a functions of MAL length. This gain is applied to the
interference cost component and to the separation constraint violation cost component. Therefore, it provides a means to
model the non-linear FER effects and interference diversity both. The default values in this table correspond to the curve
y = 2 log 10 x . This equation generates values somewhat lower than empirical best-found values (this is because we
prefer a slightly pessimistic cost function to be on the safe side).
The other table contains the F_DIV values, which are the same as the I_DIV values by default.
201

3.11.2 The AFP Blocked Traffic Cost

This section provides additional information on the AFP cost components used for the optimisation of the number of TRXs.
This optimisation is performed for each traffic pool in the network. In most cases, the traffic pool is equivalent to a transmitter
and corresponds to the BCCH and TCH subcells. In more complex cases, a traffic pool may include additional subcells, and
more than one traffic pools may exist per transmitter.
The cost component described below, and the recalculation of traffic loads, is only used when the AFP performs the
oprimisation of the number of TRXs.
The notations listed hereafter are used for the description.
{BCCH, TCH(1), TCH(2), , TCH(n)}:
{d(0), d(1), d(2), , d(n)}:

{ts(0), ts(1), ts(2), , ts(n)}:
{L(0), L(1), L(2), , L(n)}:
{CF(0), CF(1), CF(2), , CF(n)}:
CS (Erlangs):
PS (Data Timeslots):
{nb(0), nb(1), nb(2), , nb(n)}:

{HR(0), HR(1), HR(2), , HR(n)}:
Subcells of a traffic pool.

For concentric cells, at least two traffic pools exist per transmitter.
The BCCH subcell may not always be part of the pools TRX types.
Number of required TRXs of each TRX type in the pool.
Numbers of traffic timeslots.
Traffic loads.
AFP cost factors.
Overall circuit-switched traffic demand of the traffic pool (Subcells table or
traffic analysis results).
Overall packet-switched traffic demand of the traffic pool (Subcells table or
traffic analysis results).
If CS or PS is less than 1, its value is set to 1 in order to avoid working
with transmitters carrying no traffic.
Number of TRXs in the frequency plan.
TCH HR use ratios.
3.11.2.1 Calculation of New Traffic Loads Including Blocked Traffic Loads

During the optimisation of the number of TRXs, traffic loads are calculated in order to determine the blocked traffic loads
BL nb . The blocked traffic load is then multiplied by the AFP cost weight and the number of timeslots to calculate the
blocked traffic cost.
Without the optimisation of the number of required TRXs, the networks weighted Erlangs are calculated as follows:
n
WE =
d i ts i L i CF i
i=0
With the optimisation of the number of TRXs, the networks weighted Erlangs are calculated as follows:
n
WE =
nb i ts i BL nb + L nb CF i
i=0
BL nb and L nb represent the load estimation and the blocked load estimation of the AFP. They are calculated at traffic
pool level for the vector {nb(0), nb(1), nb(2), , nb(n)} as follows:
HR
PS + CS 1 ------------
BL nb + L nb = --------------------------------------------------------------n
Max 1
nb i ts i
i 0
HR
= Max i = 0 HR i
BL nb is determined from the above equation once L nb is known. L nb is obtained from the Erlang B equation applied
to the traffic pool demand and the total number of timeslots (TTS):
n
nb
ts
TTS = Max 1
------------------------------
HR
i = 0 1 ------------2
The Max() function above gives 1 timeslot when there is no TRX.

P Blocking = ErlangB CS TTS
202

AT332_TRR_E0
The above equations give the number of served circuit-switched timeslots (SCS):
HR CS 1 P
SCS = 1 ------------Blocking
2
The number of served packet-switched timeslots (SPS) is obtained as follows:
n
SPS = Min PS Max 1

nb i ts i SCS
i=0
L nb is given by:
SCS + SPS
L nb = -------------------------------------------------------------n
Max 1
nb i ts i
i 0
BL nb is given by:
HR
PS + CS 1 ------------2
BL nb = --------------------------------------------------------------- L nb
n
Max 1
nb i ts i
i 0
Once L nb and BL nb are known, L nb replaces TLi in the cost function (See "The AFP Cost Function" on page 195), and
BL nb is used to generate a new cost component, the blocked Erlangs of the pool:
n
nb i ts i BL nb CF i
i=0
3.11.2.2 Recalculation of CS and PS From Traffic Loads

In earlier versions, the detailed traffic demand information is not available. In order to guide the AFP to generate it from the
loads, the following two equations with three variables must be solved. The equations are solvable due to the monotone
nature of the Erlang B function.
Inputs for a given traffic pool:
{d(0), d(1), d(2), , d(n)}:

L:
Number of required TRXs of each TRX type in the pool

Traffic load
TTS' :
ts
TTS' = Max 1
----------------------------
HR
i = 0 1 ------------2
MB:
Maximum blocking rate (between 0 and 1).
The ratio of packet-switched demand is given by:

PS
R = ---------------------------------------------------HR
PS + CS 1 ------------
2
Here, we assume that a traffic load of 1 is generated by a demand of (1+MB)*TTS which generates a blocking rate of MB. In
other words, the ratio is calculated so that the worst case blocking rate is BM, giving a load of 1.
The following equations are solved to find PS, CS, and R, which are calculated for a traffic load of 1.
MB = ErlangB CS' TTS'
PS'
R' = ------------------------------------------------------HR
PS' + CS' 1 ------------2
PS' - + CS'
1 + MB TTS' = --------------------------HR
1 ------------
2
203

When the traffic load of a pool is not 1, PS is different from PS and CS is different from CS. Here, however, we assume that
R = R. This assumption implies that R is more or less the same as MB for big traffic pools and considerably larger than MB for
smaller pools.
The following equations are solved to find PS, CS, and R, which are calculated for the actual traffic loads.
PS
R = ---------------------------------------------------HR
PS + CS 1 ------------
2
P Blocking = ErlangB CS TTS'
HR CS 1 P
SCS = 1 ------------Blocking
2
n
SPS = Min PS Max 1

d i ts i SCS
i=0
SCS + SPS =
d i ts i L i
i=0
The above five equations are solved to get the values of the five variables PS, PC, P Blocking , SCS, SPS, and calculate the cost.
3.11.2.3 Testing the Blocked Cost Using Traffic Analysis

As long as the conditions below hold truw, the blocked cost calculation in the AFP and the effective overflow calculation in the
KPI calculation and dimensioning use the same algorithm. The conditions are:
The AFP cost factors are 1,

The HR ratios are the same within the subcells of a traffic pool,
The dimensioning model is based on Erlang B,
The timeslot configurations are the default ones,
There exists at least one TRX in the traffic pool (and at least one Erlang of traffic),
All transmitters belong to the same HCS Layer.
L nb
Effective Overflow rate = 1 -------------------------------------L nb + BL nb
Output: New values for CS and PS.
3.11.3 Interference
This appendix provides a high-level overview of interference taken into account by the AFP.
3.11.3.1 Using Interferences

If interferences are to be taken into account by the AFP, they must be calculated or imported beforehand. In order to do this,
the user should previously decide to take interferences into account (enabling the loading of all the potential interferers).
Otherwise, Atoll does not allow performing their computation by disabling the histogram part in the corresponding dialog
box.
3.11.3.2 Cumulative Density Function of C/I Levels

For each [interfered subcell, interfering subcell] pair, Atoll calculates a C/I value on each pixel of the interfered subcell service
area (as if the two subcells share the same channel). Then, Atoll integrates these C/I values to determine a C/I distribution and
transforms this distribution function into a cumulative density function in the normal way.
In Atoll, both the IMco and IMadj are represented by this Cumulative Density function This implies that each query for the
probability to have C/I conditions worse than X dB requires a single memory access: the co-channel interference probability
at X dB. In order to deduce the adjacent interference probability value, Atoll looks up the cumulative density function at the
value corresponding to X - Y dB, Y dB being the adjacency suppression value. The following example may be helpful in further
clarifying this concept:
Example: Let [TX1, BCCH] and [TX2, BCCH] be the interfered and interfering subcells respectively. The service areas
for both have been defined by Best Server with 0 dB margin. The interference probability is stated in percentage of
interfered area.
204

AT332_TRR_E0
Figure 3.17: The cumulative density of C/I levels between [TX1, BCCH] and [TX2, BCCH]
In this case, we observe that the probability for C/I (BCCH of TX2 effecting the BCCH of TX1) being greater than 0 is
100% (which is normal because TX1 is the Best Server). The probability of having a C/I value at least equal to 31 dB is
31.1%. For a required C/I level of 12 dB on the BCCH of TX1, the interference probability is 6.5% (as this requirement
is fulfilled with a probability of 93.5%).
The subcell power offset does not enter the calculation results in the .clc file. It is added
later by the AFP interface. On the other hand, its influence on the subcell service zone is
taken into account in the .clc file.
3.11.3.3 Precise Definition

Pci v n C_I is defined to be the probability of a communication (call) occupying a timeslot in subcell v (victim) to have C/I
conditions of C_I with respect to a co-channel interference from the BCCH TRX of cell n (neighbour). We assume C_I values to
be discrete and in dB. CDF(Pci) is the cumulative density function of Pci:
CDF Pci v n C_I =
Pci v n x
xC I
3.11.3.4 Precise Interference Distribution Strategy

Why does Atoll calculate and maintain precise interference distributions, while the most common solution (used by most
other tools) is rather to compress the information into two values: the co-channel and adjacent-channel interference
probabilities?
The reason is simply that it,
3.11.3.4.1
improves the AFP result,

introduces very little (or no) overhead, and
creates more generic interference information.
Direct Availability of Precise Interference Distribution to the AFP

In the presence of frequency hopping, and when one or more frequencies are common (or adjacent) in two interfering MAL
sequences, the hopping gain depends on following factors:
the MAL length,

the traffic load on the interferer TRX,
DTX level, and
the number of common (and adjacent) frequencies in the two MALs.
All these factors cannot be pre-calculated since it is the AFP that determines the MAL length and the MAL frequencies.
3.11.3.4.2
Efficient Calculation and Storage of Interference Distribution

In the innermost loop of the calculation process Atoll increments a counter each time a C/I level has a certain value. In the
case of a two-entry IM, there are only two counters for each [interfered, interferer] pair. In the case of precise distribution
information, there are about 40 counters per pair. In both cases, the number of operations is the same: one increment of an
integer value. Once Atoll finishes the counting for an [interfered, interferer] pair, it compresses the information from the
205

counters to a Cumulative Density Function (CDF) representation. In this way, access to interference probability at a certain
level is instantaneous. Thus, the only overheads are the read / write times to the files and the memory occupation at running
time. These two overheads are negligible and do not affect the calculations, the heaviest part of the task.
3.11.3.4.3
Robustness of the IM
By having precise C/I distributions calculated and exported, the user is free to change the following settings without the need
for recalculating their interference distributions:
1. Quality requirements of network elements (required C/I, % Probability Max, ),
2. C/I weighting (the interference levels above and below the C/I target),
3. Separation requirements and/or neighbour relations,
4. Hopping gain values, DTX activities, traffic load levels, HSNs, synchronisation information,
5. Any frequency assignment setting (MAL length directives, frequency domains, assignment strategies, number of
required TRXs, cost function parameters, ), or
6. Remove equipment
By not mixing any of the elements above, the interference information keeps its original probability units and is easier to check
and validate. Therefore, the user spends less time on interference recalculations than in the case of a two-entry matrix (
everything is included).
3.11.3.5 Traffic Load and Interference Information Discrimination

Atoll maintains the traffic load separate from the interference information. The reasons for implementing this strategy are
explained here.
Let us look at the possible alternatives to this strategy:
1. The mixed option: The interference information contains the traffic information as well. In this way, each IM entry
will contain the quantity of traffic interfered if a co-channel / adjacent channel reuse exists.
2. The separated option: The AFP has separate access to traffic load information and to interference probabilities (As in
Atoll).
Knowing the difference between the two alternative solutions explains why the second strategy has been opted for for Atoll.
However, in detail, this has been done because:
Option 2 is a superset that contains option 1. But option 1, being a subset, does not contain option 2 (i.e. once the
information are mixed they cannot be separated).
It does not create any overhead (the size of the additional information is negligible compared to the size of the IM).
It helps keeping the unit definitions simpler.
It is facilitates merging IMs with different traffic units.
The traffic information can be used for weighting the separation violation component.
The traffic load can be used in deciding whether a TRX can be left uncreated.
For example, if there are too many TRXs at a site and the user wishes that the AFP remove one of them, in order to be
able to not violate site constraints, the AFP must know the traffic loads in order to choose a low load TRX to be
removed.
The gain introduced by the traffic load of the interferer depends on the hopping mode and the MAL length.
Incorporating this gain in the IM (as a result of the mixed option) means that the IMs become hopping-mode and MALsize dependent. This is a bad idea since the AFP should be able to change the MAL. And the user should be able to
change the hopping mode without recalculating the IM. In addition, an IM calculated externally to Atoll, with a nonhopping BCCH can be used for the hopping TCH.
A third option also exists. Though, this option is so practically useless due to its inefficiency. It consists in mixing IM and traffic
but still keeping the traffic in its isolated form. This is again a bad idea because of the unit definition and the variety of IM
sources. It involves less benefits than the option chosen in Atoll.
206
Chapter 4
UMTS HSPA Networks
"General Prediction Studies" on page 209
"Definitions" on page 212
"UMTS HSPA Prediction Studies" on page 282
"Primary Scrambling Code Allocation" on page 317
"Automatic GSM-UMTS Neighbour Allocation" on

page 328

208

Chapter 4: UMTS HSPA Networks
AT332_TRR_E0
4 UMTS HSPA Networks

This chapter describes all the calculations performed in Atoll UMTS HSPA documents.
All the calculations are performed on TBC (to be calculated) transmitters. For the definition of TBC transmitters please refer
to "Path Loss Matrices" on page 104.
4.1 General Prediction Studies

4.1.1 Calculation Criteria
Three criteria can be studied in point analysis (Profile tab) and in common coverage studies. Study criteria are detailed in the
table below:
Study criteria
Formulas
Signal level ( P rec ) in dBm
Signal level received from a transmitter on a carrier (cell)

P rec ic = EIRP ic L path M Shadowing model L Indoor + G term L term
Path loss ( L path ) in dBm

Total losses ( L total ) in dBm
Tx
L total = L path + L Tx + L term + L indoor + M Shadowing model G Tx + G term
where,
ic is a carrier rank,
L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model,
L ant
Tx
M Shadowing model is the shadowing margin. This parameter is taken into account when the option Shadowing taken into
account is selected,
L term are the receiver losses,
G term is the receiver antenna gain,
G Tx is the transmitter antenna gain,
L Tx is the transmitter loss ( L Tx = L total DL ). For information on calculating transmitter loss, see "UMTS, CDMA2000, TDSCDMA, WiMAX, and LTE Documents" on page 30.
EIRP ic = P pilot ic + G Tx L Tx ( P pilot ic is the cell pilot power).
When you make the prediction, you can consider the best carrier of all bands or the
best carrier of a particular frequency band (Best (All Bands/Specific Band) option). In
this case, Atoll takes the highest pilot power of carriers to calculate the signal level
received from a transmitter.
Atoll considers that G term and L term equal zero.

4.1.2.1 Profile Tab
Atoll displays either the signal level received from the selected transmitter on a carrier ( P rec ic ), or the highest signal level
received from the selected transmitter on the best carrier.
209

For a selected transmitter, it is also possible to study the path loss, L path , or the total
losses, L total . Path loss and total losses are the same on any carrier.

Analysis provided in the Reception tab is based on path loss matrices. So, you can study reception from TBC transmitters for
which path loss matrices have been computed on their calculation areas.
For each transmitter, Atoll displays either the signal level received on a carrier, ( P rec ic ), or the highest signal level received
on the best carrier.
Reception bars are displayed in a decreasing signal level order. The maximum number of reception bars depends on the signal
level received from the best serving cell. Only reception bars of cells whose signal level is within a 30 dB margin from the best
serving cell can be displayed.
You can use a value other than 30 dB for the margin from the best serving cell signal
level, for example a smaller value for improving the calculation speed. For more
information on defining a different value for this margin, see the Administrator
Manual.
4.1.3 Coverage Studies

For each TBC transmitter, Txi, Atoll determines the selected criterion on each pixel inside the Txi calculation area. In fact, each
pixel within the Txi calculation area is considered as a potential (fixed or mobile) receiver.
Coverage study parameters to be set are:
The study conditions in order to determine the service area of each TBC transmitter,
The display settings to select how to colour service areas.

Atoll uses parameters entered in the Condition tab of the coverage study property dialog box to predetermine areas where it
will display coverage.
We can distinguish three cases:
4.1.3.1.1
All Servers
The service area of Txi corresponds to the bins where:
Txi
Txi
Txi
MinimumThreshold P rec ic or L total or L path MaximumThreshold
4.1.3.1.2

Txi
Txi
Txi

And
Txi
Txj
P rec ic Best P rec ic M
ji
M is the specified overlap margin (dB).

Best function: considers the highest value.
210

AT332_TRR_E0
If the margin equals 0 dB, Atoll will consider bins where the signal level received from
Txi is the highest.
If the margin is set to 2 dB, Atoll will consider bins where the signal level received
from Txi is either the highest or 2dB lower than the highest.
If the margin is set to -2 dB, Atoll will consider bins where the signal level received
from Txi is 2dB higher than the signal levels from transmitters, which are 2nd best
servers.
4.1.3.1.3

Txi
Txi
Txi

And
nd
Txi
Txj
P rec ic 2 Best P rec ic M
ji

2nd Best function: considers the second highest value.
Txi is the second highest.
from Txi is either the second highest or 2dB lower than the second highest.
from Txi is 2dB higher than the signal levels from transmitters, which are 3rd best
servers.

4.1.3.2.1
Plot Resolution
Prediction plot resolution is independent of the matrix resolutions and can be defined on a per study basis. Prediction plots
are generated from multi-resolution path loss matrices using bilinear interpolation method (similar to the one used to
evaluate site altitude).
4.1.3.2.2
Display Types
It is possible to display the transmitter service area with colours depending on any transmitter attribute or other criteria such
as:
Atoll calculates signal level received from the transmitter on each pixel of each transmitter service area. A pixel of a service
area is coloured if the signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on signal
level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as transmitter service areas. Each layer shows the different signal levels available in the transmitter service area.
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where other service
areas overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the signal level is greater
than or equal to the defined thresholds (the pixel colour depends on the signal level). Coverage consists of several
layer corresponds to an area where the signal level from the best server exceeds a defined minimum threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if path loss is greater than or equal to the defined minimum thresholds (pixel colour depends on path loss). Coverage
areas. Each layer shows the different path loss levels in the transmitter service area.
211

Total Losses (dB)

Atoll calculates total losses from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if total losses is greater than or equal to the defined minimum thresholds (pixel colour depends on total losses).
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where other service
areas overlap the studied one, Atoll determines the best transmitter and evaluates path loss from the best transmitter. A pixel
of a service area is coloured if the path loss is greater than or equal to the defined thresholds (pixel colour depends on path
layers as defined thresholds. Each layer corresponds to an area where the path loss from the best server exceeds a defined
minimum threshold.
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where service areas
overlap the studied one, Atoll determines the best transmitter and evaluates total losses from the best transmitter. A pixel of
a service area is coloured if the total losses is greater than or equal to the defined thresholds (pixel colour depends on total
layers as defined thresholds. Each layer corresponds to an area where the total losses from the best server exceed a defined
minimum threshold.
Number of Servers
Atoll evaluates how many service areas cover a pixel in order to determine the number of servers. The pixel colour depends
on the number of servers. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the number of servers is greater than
or equal to a defined minimum threshold.
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the signal level from this
transmitter fulfils signal conditions defined in Conditions tab with different Cell edge coverage probabilities. There is one
coverage area per transmitter in the explorer.
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the best signal level received
fulfils signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability in the explorer.
4.2 Definitions
This section details the terms that describe the users and the services, the input parameters as well as the formulas used in
simulations and predictions (coverage predictions and point analysis).
4.2.1 Glossary
In this chapter, we will use the following terms to describe the users and the services:
R99 users: The Circuit (R99) and Packet (R99) service users. They require an R99 bearer.
HSDPA users: The users that only support HSDPA. They have an HSDPA-capable terminal and one of these services:
Packet (HSDPA - Best Effort),

Packet (HSDPA - Variable Bit Rate).
HSDPA users require an R99 bearer (i.e. the A-DPCH radio bearer) and an HSDPA bearer.
HSPA users: The users that support both HSDPA and HSUPA. They have an HSPA-capable terminal and one of these services:
Packet (HSPA - Best Effort),

Packet (HSPA - Variable Bit Rate),
Packet (HSPA - Constant Bit Rate).
HSPA users require an R99 bearer (i.e. the E-DPCCH/A-DPCH radio bearer), an HSDPA bearer and an HSUPA bearer.
DC-HSDPA users: The dual-cell HSDPA users. Users with dual-cell HSDPA-capable terminals that can simultaneously connect
to two HSDPA cells of the transmitter for data transfer. The R99 A-DPCH bearer is transmitted on one of the cells, which is
212

AT332_TRR_E0
called the anchor cell. The user can be assigned an HSDPA bearer in each of the cells.
MC-HSDPA users: The multi-cell HSDPA users. Users with multi-cell HSDPA-capable terminals that can simultaneously connect
to several HSDPA cells of the transmitter for data transfer. The R99 A-DPCH bearer is transmitted on one of the cells, which is
called the anchor cell. The user can be assigned an HSDPA bearer in each of the cells.
DB-MC-HSDPA users: The dual-band multi-cell HSDPA users. Users with dual-band multi-cell HSDPA-capable terminals that
can simultaneously connect to several HSDPA cells on co-site transmitters using different frequency bands. The R99 A-DPCH
bearer is transmitted on one of the cells, which is called the anchor cell. The user can be assigned an HSDPA bearer in each of
the cells.
BE services: Best Effort services.
VBR services: Variable Bit Rate services.
CBR services: Constant Bit Rate services. CBR services do not support multi-cell HSDPA mode.
4.2.2 Inputs
This table lists simulation and prediction inputs (calculation options, quality targets, active set management conditions, etc.).
Name
Value
Unit
Description
F ortho
Clutter parameter
None
Orthogonality factor
Tx
Site equipment parameter
None
MUD factor
F MUD
Terminal parameter - HSDPA properties
None
MUD factor
cn first
Frequency band parameter
None
First carrier number
cnlast
None
Last carrier number
cn
None
Carrier number step
F MUD
Term
Carrier rank of the current carrier

cn cn first
- cn lower
ic = ------------------------ cn
ic
None
Where cn lower is the number of

carrier numbers lower than cn
including excluded carriers and
carriers of other frequency bands
Cell parameter
Threshold for macro diversity

None specified for a transmitter on a given
carrier ic
Cell parameter
Handover margin for a transmitter

on a given carrier ic.
None
Used for best serving cell selection in
UMTS and HSPA specific predictions.
CIO Txi ic
Cell parameter
Cell Individual Offset for a

transmitter on a given carrier ic.
None
Used for best serving cell selection in
UMTS and HSPA specific predictions.
RSCP min Txi ic
Cell parameter or Global parameter
The minimum pilot RSCP required for

a user to be connected to the
transmitter on a given carrier
req
E----c
I 0 threshold Mobility parameter
None
Ec/I0 target on downlink for the best

serving cell
Global parameter
None
Pilot RSCP threshold for compressed

mode activation
Global parameter
None
Ec/I0 threshold for compressed

mode activation
AS_Th Txi ic
M HO Txi ic
Q pilot
CM activation
RSCP pilot
CM activation
Q pilot
213

Name
Value
Unit
Description
DL
E b
---(Reception equipment, R99 bearer, Mobility) parameter
N t req
None
Eb/Nt target on downlink
Global parameter
None
Downlink Eb/Nt target increase due

to compressed mode activation
E b
--- N t req (Reception equipment, R99 bearer, Mobility) parameter
None
Eb/Nt target on uplink
Global parameter
None
Uplink Eb/Nt target increase due to

compressed mode activation
Max
Site parameter
None
Number of channel elements

available for a site on uplink
N CE DL N I
Max
Site parameter
None

available for a site on downlink
N CE UL N I
Simulation result
None
Number of channel elements of a

site consumed by users on uplink
N CE DL N I
Simulation result
None

site consumed by users on downlink
Site equipment parameter - UL overhead resources for common

channels/cell
None
Number of channel elements used by

the cell for common channels on
uplink
Site equipment parameter - DL overhead resources for common

channels/cell
None

downlink
R99 T CH
(R99 bearer, site equipment) parameter
None
Number of channel elements used

for R99 traffic channels on uplink
R99 T CH
None

for R99 traffic channels on downlink
N CE
HSUPA
(HSUPA bearer, site equipment) parameter
None

consumed by the HSUPA bearer on
uplink
Max
Site parameter
kbps
Maximum Iub backhaul throughput

for a site in the uplink
TP Iub DL N I
Max
Site parameter
kbps
Maximum Iub backhaul throughput

for a site in the downlink
TP Iub UL N I
Simulation result
kbps
Iub backhaul throughput for a site in

the uplink
TP Iub DL N I
Simulation result
kbps
Iub backhaul throughput for a site in

the downlink
kbps
Iub throughput required by the cell

for common channels in the
downlink
HSDPA
HSDPA Iub backhaul overhead
E1 T1 Ethernet
kbps
Throughput carried by an E1/T1/

Ethernet link
R99 T CH
kbps
Iub backhaul throughput consumed

by the R99 bearer in the uplink
R99 T CH
kbps

by the R99 bearer in the downlink
HSUPA
(HSUPA bearer, site equipment) parameter
kbps

by the HSUPA bearer in the uplink
N Codes Txi ic
Simulation constraint
None
Maximum number of 512 bit-length

OVSF codes available per cell (512)
N Codes Txi ic
Simulation result
None
Number of 512 bit-length OVSF

codes used by the cell
DL
Q req
DL
Q req
UL
UL
Q req
UL
Q req
N CE UL N I
Overhead
N CE UL
Overhead
N CE DL
N CE UL
N CE DL
TP Iub UL N I
Overhead
TP Iub DL N I
Overhead Iub
TP
TPIub UL
TPIub DL
TP Iub
Max
214

AT332_TRR_E0
Name
Overhead
N Codes
HSPDSCH Min
N Codes
Unit
Description
Site equipment parameter - DL overhead resources for common

channels/cell
None
Number of 256 bit-length OVSF

codes used by the cell for common
channels
Txi ic
Cell parameter (for HSDPA only)
Maximum number of 16 bit-length

None OVSF codes available per cell for HSPDSCH
Txi ic
Minimum number of 16 bit-length

None OVSF codes available per cell for HSPDSCH
HSPDSCH Max
N Codes
Value
NF term
Terminal parameter
None
Terminal Noise Figure
NF Tx
Transmitter parameter (user-defined or calculated from transmitter

equipment characteristics)
None
Transmitter Noise Figure
1.38 10-23
J/K
Boltzman constant
293
Ambient temperature
3.84 MHz
Hz
Spreading Bandwidth
NRinter techno log y
Cell parameter
Tx DL
Cell parameter
Only used in uplink interference-based calculations of the MonteCarlo simulation
RF ic ic adj
Network parameter
If not defined, it is assumed that there is no inter-carrier
interference
Tx UL
Tx m
ICP ic ic
i
Network parameter
If not defined, it is assumed that there is no inter-technology
downlink interferences due to external transmitters
None Inter-technology downlink noise rise

None
Inter-technology uplink noise rise
None
Interference reduction factor

between two adjacent carriers ic
and ic adj
Inter-technology Channel Protection

between the signal transmitted by Tx
and received by m assuming the
None
frequency gap between ic i (external
network) and ic
UL
Cell parameter (user-defined or simulation result)
Total uplink load factor of the cell
UL
Simulation result
Uplink cell load contribution due to

R99 traffic
X HSUPA
Cell parameter
Uplink cell load contribution due to

HSUPA traffic
UL
Simulation constraint (global parameter or cell parameter)
Maximum uplink load factor of the

cell
Maximum percentage of used power
Thermal noise at transmitter
Thermal noise at terminal
bps
Chip rate
X R99
UL
X max
DL
%Power max
Tx UL
Tx
NF Tx K T W NR inter techno log y
Term
NF Term K T W NR inter techno log y
N0
N0
Rc
Tx DL
W 10
UL
f rake efficiency
DL
Terminal parameter
None
Downlink rake receiver efficiency

factor
TP P DL
R99
R99 bearer parameter
kbps
R99 bearer downlink peak

throughput
F spreading Active user
None
Downlink spreading factor for active

users
DL
None
Downlink spreading factor for

inactive users
f rake efficiency
DL
F spreading Inactive user
None Uplink rake receiver efficiency factor
215

Name
Value
Unit
Description
rc
DL
None
Ratio between DPCCH and DPCH

transmission duration on downlink
DPCCH and DPCH respectively refer
to the Dedicated Physical Control
Channel and Dedicated Physical
Channel
TP max DL
Cell parameter
kbps
Maximum connection throughput

per user on downlink
TP P UL
R99
kbps
R99 bearer uplink peak throughput
f act
UL
Service parameter
kbps
Uplink activity factor for the service
f act
DL
Service parameter
kbps
Downlink activity factor for the

service
f act ADPCH
UL
Service parameter
kbps
Uplink activity factor on E-DPCCH

channels
f act ADPCH
DL
Service parameter
kbps
Downlink Activity factor on A-DPCH

channel
TPD min UL
Service parameter
kbps
Minimum required bit rate that the

service should have in order to be
available in the uplink
TPD min DL
Service parameter
kbps

available in the downlink
TPD max UL
Service parameter
kbps
Maximum bit rate that the service

can require in the uplink
TPD max DL
Service parameter
kbps
Maximum bit rate that the service

can require in the downlink
Ratio between the DPCCH and DPCH

powers transmitted on uplink
DPCCH and DPCH respectively refer
None
to the Dedicated Physical Control
Channel and Dedicated Physical
Channel
rc
UL
TP max UL
Cell parameter
kbps
Maximum connection throughput

per user on uplink
DL
W ----------------R99
TP P DL
None
Service downlink processing gain
Gp
UL
W ----------------R99
TP P UL
None
Service uplink processing gain
I HSDPABearer
HSDPA bearer parameter
None
Index of the HSDPA bearer obtained

by the user in the cell (Txi,ic)
HSDPA bearer parameter
kbps
Peak RLC throughput supported by

the HSDPA bearer
kbps
Peak RLC throughput provided to the

user in the cell (Txi,ic) in the
downlink
Gp
DL
TP P RLC
I HSDPABearer
DL
Without MIMO: TP P RLC I HSDPABearer

DL
TPP RLC Tx ic
DL
With MIMO (transmit diversity): TP P RLC I HSDPABearer

With MIMO (spatial multiplexing):
DL
TP P RLC I HSDPABearer
216
Max
1 + f SM Gain G SM 1

AT332_TRR_E0
Name
Value
Unit
Description
kbps

user in the downlink
TPD min DL
----------------------------------------------------DL
TP P RLC I HSDPABearer
HSDPA bearer consumption for a

CBR service user
HSDPA study result
kbps
Effective RLC throughput supported

by the HSDPA bearer
HSDPA study result
kbps
Average effective RLC throughput

supported by the HSDPA bearer
DL
HSDPA study result
kbps
Peak MAC throughput supported by

the HSDPA bearer
DL
HSDPA study result
kbps
Effective MAC throughput supported

by the HSDPA bearer
DL
HSDPA study result
kbps
User application throughput on

downlink
TP A
UL
HSUPA study result
kbps
User application throughput on

uplink
TP Av A
UL
HSUPA study result
kbps
User average application throughput

on uplink
I HSUPABearer
HSUPA Bearer parameter
None
Index of the HSUPA bearer obtained

in the cell (Txi,ic)
HSDPA study result

For single-carrier HSDPA users
DL
DL
TP P RLC Tx ic
TP P RLC
For DC-HSDPA users
DL
TP P RLC Tx ic
ic Tx
C HSDPABearer
DL
TP E RLC
DL
TP Av E RLC
TP P MAC
TP E MAC
TP A
N Rtx
I HSUPABearer
HSUPA bearer selection parameter
Maximum number of
retransmissions a HARQ process will
kbps
perform for a block of data before
moving on to a new block of data, for
the HSUPA bearer index
UL
TP P RLC
I HSUPABearer
UL
HSUPA bearer parameter
kbps
Peak RLC throughput supported by

the HSUPA bearer
kbps

user in the cell (Txi,ic) in the uplink
HSUPA study result
TP P RLC
TP P RLC I HSUPABearer
C HSUPABearer
TPD min UL
----------------------------------------------------UL
HSUPA bearer consumption for a

CBR service user
HSUPA study result
kbps
Minimum effective RLC throughput

supported by the HSUPA bearer
TP Av E R LC
HSUPA study result
kbps
Average effective RLC throughput

supported by the HSUPA bearer
TP P M AC
UL
HSUPA study result
kbps
Peak MAC throughput supported by

the HSUPA bearer
TP Offset
Service parameter (for HSDPA only)
kbps
Throughput offset
f TP Scaling
Service parameter (for HSDPA only)
Scaling factor
P max Txi
Transmitter parameter
Maximum shared power

Available only if the inter-carrier
power sharing option is activated
P SCH Txi ic
Cell parameter
Cell synchronisation channel power
P OtherCCH Txi ic
Cell parameter
Cell other common channels (except

CPICH and SCH) power
UL
TP Min E R LC
UL
UL
217

Name
Value
Unit
Description
P pilot Txi ic
Cell parameter
Cell pilot power
P HSDPA Txi ic
Cell parameter (user-defined or simulation result) (for HSDPA only)

P HS PDSCH Txi ic + n HS SCCH P HS SCCH Txi ic
Available cell HSDPA power

HSDPA: High Speed Downlink Packet
Access
P HS PDSCH Txi ic
Simulation result (for HSDPA only)
Cell HS-PDSCH power

HS-PDSCH: High Speed Physical
Downlink Shared Channel
P HS SCCH Txi ic
Cell HS-SCCH power

HS-SCCH: High Speed Shared Control
Channel
n HS SCCH
P Headroom Txi ic
Cell headroom power
P max Txi ic
Cell parameter
Maximum Cell power
P tch Txi ic
Simulation result
R99 traffic channel power

transmitted on carrier ic
min
Minimum power allowed on R99

traffic data channel
P tch
max
Maximum power allowed on R99

traffic data channel
P HSUPA Txi ic
Cell parameter
Cell HSUPA power

HSUPA: High Speed Uplink Packet
Access
P tx H SDPA Txi ic
Simulation result
Transmitter HSDPA power

transmitted on carrier ic
Transmitter R99 power transmitted

on carrier ic
P tch
number of HS-SCCH channels

managed by the cell
Simulation result
P pilot Txi ic + P SCH Txi ic + P OtherCCH Txi ic +
P tx R99 Txi ic
P tch Txi ic +
tch(ic) used for

R99 users
DL
P tch Txi ic f act ADPCH
tch(ic) used for

HSUPA users
P tx Txi ic
Cell parameter (user-defined or simulation result)

P tx R99 Txi ic + P tx H SDPA Txi ic + P HSUPA Txi ic
Transmitter total power transmitted

on carrier ic
P term R99
Calculated in the simulation but not displayed
Terminal power transmitted to

obtain the R99 radio bearer
P term HSUPA
Calculated in the simulation but not displayed
Terminal power transmitted to

obtain the HSUPA radio bearer
Total power transmitted by the

terminal
Simulation result
P term
P term R99
UL
f act ADPCH
+ P term HSUPA for HSPA users
P term R99 for R99 and HSDPA users
218
P term
min
Terminal parameter
Minimum terminal power allowed
P term
max
Terminal parameter
Maximum terminal power allowed
BTS
BTS parameter
Percentage of BTS signal correctly

transmitted
term
Terminal parameter
Percentage of terminal signal

correctly transmitted
Clutter parameter
Percentage of pilot finger percentage of signal received by the

terminal pilot finger
G Tx
Antenna parameter
None
Transmitter antenna gain

AT332_TRR_E0
Name
Value
Unit
Description
G Term
Terminal parameter
None
Terminal gain
DL
R99 bearer parameter - Depends on the transmitter Tx diversity
None
Gain due to transmit diversity
UL
R99 bearer parameter - Depends on the transmitter Rx diversity
None
Gain due to receive diversity
G SM
Max
MIMO configuration parameter
dB
Maximum spatial multiplexing gain

for a given number of transmission
and reception antennas
G TD
DL
MIMO configuration parameter
dB
Downlink Transmit Diversity gain for

a given number of transmission and
reception antenna ports
f SM Gain
Clutter parameter
None
Spatial multiplexing gain factor
G TD
Clutter parameter
dB
Additional diversity gain in downlink
L Tx

None
Transmitter lossa
L body
Service parameter
None
Body loss
L Term
Terminal parameter
None
Terminal loss
L indoor
Clutter parameter
L path
Propagation model result
None
Path loss
M Shadowing model
Result calculated from cell edge coverage probability and model

standard deviation
None
Model Shadowing margin

Only used in prediction studies
M Shadowing Ec Io
Result calculated from cell edge coverage probability and Ec/I0

standard deviation
None
Ec/I0 Shadowing margin

None
DL gain due to availability of several

pilot signals at the mobile b.
DL
Result calculated from cell edge coverage probability and DL Eb/Nt

standard deviation
None
DL Eb/Nt Shadowing margin

UL
Result calculated from cell edge coverage probability and UL Eb/Nt

standard deviation
None
UL Eb/Nt Shadowing margin

None
UL quality gain due to signal diversity

in soft handoffc.
None
Random shadowing error drawn

during Monte-Carlo simulation
Only used in simulations
None
Transmitter-terminal total loss
G Div
G Div
DL
DL
M Shadowing Eb Nt
npaths
DL
G macro diversity
M Shadowing Eb Nt
n=2 or 3
UL
npaths
UL
G macro diversity
E Shadowing
Indoor loss
UL
M Shadowing Eb Nt
n=2 or 3
Global parameter (default value)
Simulation result
UL
In prediction studiesd
For Ec/I0 calculation
L path L Tx L term L body L indoor M Shadowing Ec Io
------------------------------------------------------------------------------------------------------------------------------------G Tx G term
LT
For DL Eb/Nt calculation

L path L Tx L term L body L indoor M Shadowing Eb Nt
DL
---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
For UL Eb/Nt calculation
UL
---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
In simulations
L path L Tx L term L body L indoor E Shadowing
-------------------------------------------------------------------------------------------------------------------G Tx G term
219

Name
Value
Unit
Description
P c Txi ic
P pilot Txi ic
-------------------------------LT
Chip power received at terminal
DL
P tch Txi ic
----------------------------LT
Bit power received at terminal on

carrier ic
DL
P tx Txi ic
--------------------------LT
Total power received at terminal

from a transmitter on carrier ic
P tch Txi ic
----------------------------LT

from traffic channels of a transmitter
on carrier ic
P term
-----------LT
Bit power received at transmitter on

carrier ic used by terminal
P term R99
-----------------------LT

carrier ic used by terminal

DPDCH from a terminal on carrier ic
P b Txi ic
P tot Txi ic
DL
P traf Txi ic
tch ic
UL
P b ic
UL
P b R99 ic
UL
UL
P b DPDCH ic
UL
P b R99 ic 1 r c
L Tx = L total UL on uplink and L Tx = L total DL on downlink. For information on calculating transmitter losses on uplink and downlink,
a.
see "UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents" on page 30.
npaths
M Shadowing Ec Io corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in case
b.
of downlink Ec/I0 modelling.

npaths
c.
M Shadowing Eb Nt
d.
case of uplink soft handoff modelling.

In uplink prediction studies, only carrier power level is downgraded by the shadowing margin ( M Shadowing Eb Nt
UL
corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in
UL
). In downlink
prediction studies, carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing Eb Nt
M Shadowing Ec Io ) while extra-cell interference level is not. Therefore, M Shadowing Eb Nt
DL
DL
or
or M Shadowing Ec Io is set to 1 in downlink
extra-cell interference calculation.
4.2.3 Ec/I0 Calculation

This table details the pilot quality ( Q pilot or Ec Io ) calculations.
Name
Value
Unit
Description
I intra txi ic
P SCH txi ic
DL
DL
-
P tot txi ic BTS P tot txi ic ----------------------------L
Downlink intra-cell interference at

terminal on carrier ic
Downlink extra-cell interference at

Downlink inter-carrier interference

at terminal on carrier ic
DL
DL
I extra ic
DL
I inter carrier ic
DL
I inter techno log y ic
txj j i
DL
P tot txj
ic adj
txj
j
---------------------------------------------
RF ic ic adj
ni
220
DL
P tot txj ic
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
Downlink inter-technology
interference at terminal on carrier ic
a

AT332_TRR_E0
Name
Value
Unit
Description
Total received noise at terminal on

carrier ic
None
Quality level at terminal on pilot for

carrier ic
Without Pilot:
DL
DL
DL
DL
I intra txi ic + I extra ic + I inter carrier ic + I inter techno log y ic

Term
+ N0
DL
I 0 ic
1 BTS P c txi ic
DL
Total noise:
DL
DL
P tot txi ic + I extra ic + I inter carrier ic

DL
Term
+ I inter techno log y ic + N 0

BTS P c txi ic
-------------------------------------------------DL
I 0 ic
Ec
Q pilot txi ic ----
I0
a.
In the case of an interfering GSM external network in frequency hopping, the ICP value is weighted according to the fractional load.
4.2.4 DL Eb/Nt Calculation

Eb
DL
This table details calculations of downlink traffic channel quality ( Q tch or ------ ).
Nt DL
Name
Value
Unit
Description
I intra txi ic
P SCH txi ic
DL
DL
-
P tot txi ic BTS F ortho P tot txi ic ----------------------------L



DL
DL
I extra ic
txj j i
DL
P tot txj
DL
I inter carrier ic
RF ic ic adj
ni
DL
ic adj
txj j
---------------------------------------------
DL
N tot ic
DL
P tot txj ic
DL
DL
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
DL
Term
DL
I intra ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0

carrier ic
None
Quality level at terminal on a traffic

channel from one transmitter on
carrier ic b
Without useful signal:

DL
E
DL
Q tch txi ic ----b-
N t DL
BTS P b txi ic
DL
------------------------------------------------------------------------------------------------------ G DL
Div G p
DL
DL
N tot ic 1 F ortho BTS P b txi ic
DL
BTS P b txi ic
DL
- G DL
Total Noise: ------------------------------------------Div G p
DL
N tot ic
DL
Q ic
DL
f rake efficiency
DL
Q tch tx k
tx ActiveSet
k
ic
Quality level at terminal using carrier

ic due to combination of all
None
transmitters of the active set
(Macro-diversity conditions).
DL
G SHO
Q ic
--------------------------------------------------DL
Q tch BestServer ic
req
P tch txi
Q req
----------------- P tch txi ic
DL
Q ic
DL
None
Soft handover gain on downlink
Required transmitter traffic channel

power to achieve Eb/Nt target at
DL
a.
b.
ic
Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account only in simulations. In point
analysis and coverage studies, Atoll uses the option Total noise to evaluate DL and UL Eb/Nt.
221

4.2.5 UL Eb/Nt Calculation

Eb
UL
This table details calculations of uplink traffic channel quality ( Q tch or ------ ).
Nt UL
Name
Value
Pb
UL
UL intra
I tot
txi ic
ic
term
Unit
Description
Total power received at transmitter

from intra-cell terminals using carrier
ic

from extra-cell terminals using
carrier ic
Uplink inter-carrier interference at

txi
UL extra
I tot
txi ic
UL
P b ic
term
txj j i
UL
P b ic adj
UL
term
I inter carrier txi ic
txj j
-----------------------------------
RF ic ic adj
UL
I tot txi ic
UL extra
I tot
UL intra
Tx
txi ic + 1 F MUD term I tot
UL
UL
N tot txi ic
UL
txi ic + I inter carrier txi W
ic
tx
I tot txi ic + N 0
Total received interference at

transmitter on carrier ic
Total noise at transmitter on carrier

ic
(Uplink interference)
None
Quality level at transmitter on a

traffic channel for carrier ic a

UL
E
UL
Q tch txi ic ----b-
N t UL
term P b DPDCH ic
UL
-------------------------------------------------------------------------------------------------------- G UL
Div G p
UL
Tx
UL
N tot txi ic 1 F MUD term P b ic
UL
term P b DPDCH ic
UL
- G UL
Total noise: --------------------------------------------------Div G p
UL
N tot txi ic
UL
No HO: Q tch txi ic
UL
Softer HO: f rake efficiency
UL
Q tch txk ic
tx ActiveSet
k
samesite
Soft, softer/soft HO (No MRC):

Max Q UL
tch tx k
tx k ActiveSet
UL
Q ic
ic
UL
G macro diversity
Softer/soft HO (MRC):
UL
UL
UL
Q tch tx k ic Q tch tx l ic
tx k ,tx l ActiveSet f rake efficiency
tx k samesite
tx k
Max
Quality level at site using carrier ic

due to combination of all
transmitters of the active set located
at the same site and taking into
account
increasing of the quality due
None
to macro-diversity (macro-diversity
gain).
UL
In simulations G macro diversity = 1 .
tx l othersite
UL
G macro diversity
UL
UL
G SHO
Q ic
-------------------------------------------------UL
Q tch BestServer ic
None
Soft handover gain on uplink
Required terminal power to achieve

Eb/Nt target at transmitter on carrier
ic
UL
req
P term ic
a.
222
Q req
----------------- P term
UL
Q ic

AT332_TRR_E0
4.3 Simulations
The simulation process consists of two steps:
1. Obtaining a realistic user distribution
Atoll generates a user distribution using a Monte-Carlo algorithm, which requires traffic maps and data as input. The
resulting user distribution complies with the traffic database and maps provided to the algorithm.
the simulation and on the network interferences. A user may be either active or inactive. Both active and inactive
users consume radio resources and create interference.
Then, Atoll randomly assigns a shadowing error to each user using the probability distribution that describes the
shadowing effect.
Finally, another random trial determines user positions in their respective traffic zone and whether they are indoors
or outdoors (according to the clutter weighting and the indoor ratio per clutter class defined for the traffic maps).
2. Power control simulation
4.3.1 Generating a Realistic User Distribution

During the simulation, a first random trial is performed to determine the number of users and their activity status. Four
activity status are modelled:
Active UL: the user is active on UL and inactive on DL
Active DL: the user is active on DL and inactive on UL
Active UL+DL: the user is active on UL and on DL
Inactive: the user is inactive on UL and on DL
The determination of the number of users and the activity status allocation depend on the type of traffic cartography used.
Atoll determines the total number of users attempting connection in each simulation
based on the Poisson distribution. This may lead to slight variations in the total numbers
of users in different simulations. To have the same total number of users in each
simulation of a group, add the following lines in the Atoll.ini file:
[Simulation]
RandomTotalUsers=0
4.3.1.1 Simulations Based on User Profile Traffic Maps

User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of
user profiles with an associated mobility type and a given density (number of subscribers with the same profile per km).
User profile traffic maps: Each polygon and line of the map is assigned a density of subscribers with given user profile and
mobility type. If the map is composed of points, each point is assigned a number of subscribers with given user profile and
mobility type.
The user profile models the behaviour of the different subscriber categories. Each user profile contains a list of services and
their associated parameters describing how these services are accessed by the subscriber.
From environment (or polygon) surface (S) and user profile density (D), a number of subscribers (X) per user profile is inferred.
X = SD
When user profile traffic maps are composed of lines, the number of subscribers (X)
per user profile is calculated from the line length (L) and the user profile density (D)
(nb of subscribers per km) as follows: X = L D
The number of subscribers (X) is an input when a user profile traffic map is composed
of points.
For each behaviour described in a user profile, according to the service, frequency use and exchange volume, Atoll calculates
the probability for the user being active in uplink and in downlink at an instant t.
223

4.3.1.1.1
Circuit Switched Service (i)

User profile parameters for circuit switched services are:
The used terminal (equipment used for the service (from the Terminals table)),
The average number of calls per hour N call ,
The average duration of a call (seconds) d .
The number of users and their distribution per activity status is determined as follows:
Calculation of the service usage duration per hour ( p 0 : probability of a connection):
N call d
p o = ------------------3600
Calculation of the number of users trying to access the service i ( n i ):
ni = X p0
Next, we can take into account activity periods during the connection in order to determine the activity status of each user.
Calculation of activity probabilities:

UL
DL
Probability of being inactive on UL and DL: p inactive = 1 f act 1 f act

UL
DL
DL
UL
Probability of being active on UL only: p UL = f act 1 f act

Probability of being active on DL only: p DL = f act 1 f act
UL
DL
Probability of being active both on UL and DL: p UL + DL = f act f act

UL
DL
Where, f act and f act are respectively the UL and DL activity factors defined for the circuit switched service i.
Calculation of number of users per activity status:

inactive
Number of inactive users on UL and DL: n i
= n i p inactive
Number of users active on UL and inactive on DL: n i UL = n i p UL

Number of users active on DL and inactive on UL: n i DL = n i p DL
Number of users active on UL and DL both: n i UL + DL = n i p UL + DL
Therefore, a user when he is connected can have four different activity status: either active on both links, or inactive on both
links, or active on UL only, or active on DL only.
4.3.1.1.2
Packet Switched Service (j)

User profile parameters for packet switched services are:
The used terminal (equipment used for the service (from the Terminals table)),
The average number of packet sessions per hour N sess ,
The volume (in kbytes) which is transferred on the downlink V DL and the uplink V UL during a session.
A packet session consists of several packet calls separated by a reading time. Each packet call is defined by its size and may be
divided in packets of fixed size (1500 Bytes) separated by an inter arrival time.
In Atoll, a packet session is described by following parameters:
UL
N packet c all : Average number of packet calls on the uplink during a session,
DL
N packet c all : Average number of packet calls on the downlink during a session,
UL
T packet call : Average time (millisecond) between two packets calls on the uplink ,
DL
T packet call : Average time (millisecond) between two packets calls on the downlink ,
UL
T packet : Average time (millisecond) between two packets on the uplink ,
224

AT332_TRR_E0
DL
T packet : Average time (millisecond) between two packets on the downlink ,

UL
S packet : Packet size (Bytes) on uplink,

DL
S packet : Packet size (Bytes) on downlink.
Figure 4.1: Description of a Packet Session

Calculation of the average packet call size (kBytes):
V UL
V DL
UL
DL
S packet c all = ---------------------------------------and S packet c all = ---------------------------------------UL
UL
DL
DL
N packet c all f eff
N packet c all f eff
UL
DL
Where f eff and f eff are the UL and DL efficiency factors defined for the packet switched service j.
UL
DL
For packet (HSDPA) and packet (HSPA) services, f eff and f eff are set to 1.
Calculation of the average number of packets per packet call:

UL
DL
S packet c all
S packet c all
UL
- + 1 and N DL
- + 1
N packet = int ------------------------------packet = int ------------------------------UL
S packet 1024
S DL
packet 1024
1kBytes = 1024Bytes.
Calculation of the average duration of inactivity within a packet call (s):

UL
UL
DL
DL
N packet 1 T packet
N packet 1 T packet
UL
DL
- and D Inactivity packet call = -------------------------------------------------------- D Inactivity packet call = --------------------------------------------------------1000
1000
Calculation of the average duration of inactivity in a session (s):
UL
UL
UL
DL
DL
DL
D Inactivity session = N packet c all D Inactivity packet call and D Inactivity session = N packet c all D Inactivity packet call
Calculation of the average duration of activity in a session (s):

UL
UL
DL
DL
N packet S packet 8
N packet S packet 8
UL
UL
DL
- and D DL
D Activity session = N packet c all -----------------------------------------------Activity session = N packet c all -----------------------------------------------UL
DL
TP Av 1000
TP Av 1000
UL
DL
Where TP Av and TP Av are the uplink and downlink average requested throughputs defined for the service j.
Therefore, the average duration of a connection (in s) is:
225

UL
UL
UL
DL
DL
DL
D Connection = D Activity session + D Inactivity session and D Connection = D Activity session + D Inactivity session
Calculation of the service usage duration per hour (probability of a connection):
N sess
N sess
UL
DL
DL
- D UL
p Connection = ----------Connection and p Connection = ------------ D Connection
3600
3600
Calculation of the probability of being connected:

UL
DL
p Connected = 1 1 p Connection 1 p Connection

Therefore, the number of users who want to get the service j is:
n j = X p Connected
As you can see on the picture above, we have to consider three possible cases when a user is connected:
1st case: At a given time, packets are downloaded and uploaded.
In this case, the probability of being connected is:

UL
DL
p Connection p Connection
UL + DL
p Connected = --------------------------------------------------------p Connected
2nd case: At a given time, packet are uploaded (no packet is downloaded).
Here, the probability of being connected is:

UL
DL
p Connection 1 p Connection
UL
p Connected = ----------------------------------------------------------------------p Connected
3rd case: At a given time, packet are downloaded (no packet is uploaded).
In this case, the probability of being connected is:

DL
UL
DL
p Connected = ----------------------------------------------------------------------p Connected
Now, we have to take into account activity periods during the connection in order to determine the activity status of each
user.
UL
Calculation of the probability of being active:

UL
DL
D Activity session
D Activity session
DL
= -------------------------------------------------------------------------------------------and f = -------------------------------------------------------------------------------------------UL
UL
DL
DL
D Inactivity session + D Activity session
D Inactivity session + D Activity session
Therefore, we have:
1st case: At a given time, packets are downloaded and uploaded.
The user can be active on UL and inactive on DL; this probability is:
1
p UL = f
UL
DL
UL + DL
1 f p Connected
The user can be active on DL and inactive on UL; this probability is:
1
p DL = f
DL
UL
UL + DL
1 f p Connected
The user can be active on both links; this probability is:

1
p UL + DL = f
UL
DL
UL + DL
p Connected
The user can be inactive on both links; this probability is:

UL
DL
UL + DL
p inactive = 1 f 1 f p Connected
2nd case: At a given time, packet are uploaded (no packet is downloaded).
The user can be active on UL and inactive on DL; this probability is:
2
p UL = f
UL
UL
p Connected
226

AT332_TRR_E0
UL
UL
p inactive = 1 f p Connected
3rd case: At a given time, packet are downloaded (no packet is uploaded).
The user can be active on DL and inactive on UL; this probability is:
3
p DL = f
DL
DL
p Connected

DL
DL
p inactive = 1 f p Connected
Calculation of number of users per activity status

inactive
Number of inactive users on UL and DL: n j
= n j p inactive + p inactive + p inactive

1
Number of users active on UL and inactive on DL: n j UL = n j p UL + p UL

Number of users active on DL and inactive on UL: n j DL = n j p DL + p DL
1
Number of users active on UL and DL: n j UL + DL = n j p UL + DL

Therefore, a user when he is connected can have four different activity status: either active on both links, or inactive on both
links, or active on UL only, or active on DL only.
The user distribution per service and the activity status distribution between the users
are average distributions. And the service and the activity status of each user are
randomly drawn in each simulation. Therefore, if you compute several simulations at
once, the average number of users per service and average numbers of inactive, active
on UL, active on DL and active on UL and DL users, respectively, will correspond to
calculated distributions. But if you check each simulation, the user distribution between
services as well as the activity status distribution between users is different in each of
them.
4.3.1.2 Simulations Based on Sector Traffic Maps

Sector traffic maps can be based on live traffic data from OMC (Operation and Maintenance Centre). Traffic is spread over the
best server coverage area of each transmitter and each coverage area is assigned either the throughputs in the uplink and in
the downlink or the number of users per activity status or the total number of users (including all activity statuses).
4.3.1.2.1
Throughputs in Uplink and Downlink

When selecting Throughputs in Uplink and Downlink, you can input the throughput demands in the uplink and downlink for
each sector and for each listed service.
Atoll calculates the number of users active in uplink and in downlink in the Txi cell using the service (NUL and NDL) as follows:
UL
Rt
N UL = ---------- and
UL
TP Av
DL
Rt
N DL = ---------DL
TPAv
UL
is the kbits per second transmitted in UL in the Txi cell to supply the service.
DL
is the kbits per second transmitted in DL in the Txi cell to supply the service.
Rt
Rt
DL
TP Av is the downlink average requested throughput defined for the service,

UL
TP Av is the uplink average requested throughput defined for the service.

NUL and NDL values include:
Users active in uplink and inactive in downlink (ni(UL)),

Users active in downlink and inactive in uplink (ni(DL)),
And users active in both links (ni(UL+DL)).
Atoll takes into account activity periods during the connection in order to determine the activity status of each user.
Activity probabilities are calculated as follows:
227


UL
DL
Probability of being inactive in UL and DL: p inactive = 1 f act 1 f act

UL
DL
DL
UL
Probability of being active in UL only: p UL = f act 1 f act

Probability of being active in DL only: p DL = f act 1 f act
UL
DL
Probability of being active both in UL and DL: p UL + DL = f act f act

UL
DL
Where, f act and f act are respectively the UL and DL activity factors defined for the service i.
Then, Atoll calculates the number of users per activity status:
We have:
p UL + p UL + DL n j UL + n j DL + n j UL + DL = N UL
p DL + p UL + DL n j UL + n j DL + n j UL + DL = N DL
Therefore, we have:
N UL p UL + DL N DL p UL + DL
Number of users active in UL and DL both: n i UL + DL = min --------------------------------- --------------------------------
p UL + p UL + DL p DL + p UL + DL
Number of users active in UL and inactive in DL: n i UL = N UL n i UL + DL
Number of users active in DL and inactive in UL: n i DL = N DL n i UL + DL
inactive
Number of inactive users in UL and DL: n i
n j UL + n j DL + n j UL + DL
- p inactive
= -----------------------------------------------------------------------------1 p inactive
Therefore, a connected user can have four different activity status: either active in both links, or inactive in both links, or active
in UL only, or active in DL only.
4.3.1.2.2
Total Number of Users (All Activity Statuses)

When selecting Total Number of Users (All Activity Statuses), you can input the number of connected users for each sector
and for each listed service ( n i ).
UL
DL
Probability of being inactive in UL and DL: p inactive = 1 f act 1 f act

UL
DL
DL
UL
Probability of being active in UL only: p UL = f act 1 f act

Probability of being active in DL only: p DL = f act 1 f act
UL
DL
Probability of being active both in UL and DL: p UL + DL = f act f act

UL
DL
Where, f act and f act are respectively the UL and DL activity factors defined for the service i.
inactive
Number of inactive users in UL and DL: n i
= n i p inactive
Number of users active in UL and inactive in DL: n i UL = n i p UL

Number of users active in DL and inactive in UL: n i DL = n i p DL
Number of users active in UL and DL both: n i UL + DL = n i p UL + DL
228

AT332_TRR_E0
4.3.1.2.3
Number of Users per Activity Status

inactive
When selecting Number of Users per Activity Status, you can directly input the number of inactive users ( n i
), the
number of users active in the uplink ( n i UL ), in the downlink ( n i DL ) and in the uplink and downlink ( n i UL + DL ), for
each sector and for each service.
The activity status distribution between users is an average distribution. In fact, in each
simulation, the activity status of each user is randomly drawn. Therefore, if you compute
several simulations at once, average numbers of inactive, active on UL, active on DL and
active on UL and DL users correspond to the calculated distribution. But if you check each
simulation, the activity status distribution between users is different in each of them.
4.3.2 Power Control Simulation

The power control algorithm simulates the way a UMTS network regulates itself by using uplink and downlink power controls
in order to minimize interference and maximize capacity.
HSDPA users are linked to the A-DPCH radio bearer (an R99 radio bearer). Therefore, the network uses a A-DPCH power
control on UL and DL and then it performs fast link adaptation on DL in order to select an HSDPA radio bearer. For HSPA users,
the network first uses a E-DPCCH/A-DPCH power control on UL and DL, checks that there is an HSDPA connection on downlink
and then carries out noise rise scheduling in order to select an HSUPA radio bearer on uplink. Atoll simulates these network
regulation mechanisms with an iterative algorithm and calculates, for each user distribution, network parameters such as cell
power, mobile terminal power, active set and handoff status for each terminal. During each iteration of the algorithm, all the
users selected during the user distribution generation (1st step) attempt to connect one by one to network transmitters. The
process is repeated until the network is balanced, i.e., until the convergence criteria (on UL and DL) are satisfied.
229

Figure 4.2: UMTS HSPA Power Control Algorithm

As shown in Figure 4.2 on page 230, the simulation algorithm is divided in three parts. All users are evaluated by the R99 part
of the algorithm. HSDPA and HSPA users, unless they have been rejected during the R99 part of the algorithm, are then
evaluated by the HSDPA part of the algorithm. Finally, HSPA users, unless they have been rejected during the R99 or HSDPA
parts of the algorithm, are then evaluated by the HSUPA part of the algorithm.
The steps of this algorithm are detailed below.
4.3.2.1 Algorithm Initialization

The total power transmitted by the base station txi on the carrier ic m , P Tx txi ic m , is initialised to
P pilot txi ic m + P SCH txi ic m + P otherCCH txi ic m + P HSDPA txi ic m + P HSUPA txi ic .
UL intra
Uplink powers received by the base station txi on carrier ic m , I tot

are initialised to 0 W (i.e. no connected mobile).
230
UL extra
txi ic m , I tot
UL
txi ic m and I inter carrier txi ic m

AT332_TRR_E0
UL
I tot txi ic m
UL
Therefore, we have: X R99 txi ic m k = ------------------------------- = 0
UL
N tot txi ic m
4.3.2.2 R99 Part of the Algorithm

The algorithm is detailed for any iteration k. Xk is the value of the X (variable) at the iteration k. In the algorithm, the
UL
DL
thresholds, Q req and Q req , depend on the user mobility type and are defined in the R99 bearer selection table. All variables
are described in Definitions and formulas part. The bearer downgrading is not dealt with.
The algorithm applies to single frequency band networks and to multi-band networks.
For each mobile (Mb), Atoll only considers the cells (txi,ic) for which the pilot RSCP exceeds the minimum pilot RSCP:
P c txi M b ic RSCP min txi ic .
For each mobile Mb, we have the following steps:
Determination of Mbs Best Serving Cell
For each transmitter txi containing Mb in its calculation area and working on a frequency band supported by the Mbs terminal
).
BTS P c txi M b ic
Calculation of Q pilot txi ic Mb = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------Term
k
DL
DL
DL
DL
P tot txi ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0
If user selects without Pilot
BTS P c txi M b ic
Q pilot txi ic Mb = -----------------------------------------------------------------------------------------------------------------------------------------------------------k
DL
DL
DL
DL
I intra ic + I extra ic + I inter carrier ic + I inter techno log y ic
Term
+ N0
1 BTS P c txi M b ic
Determination of the candidate cells, (txBS,ic).

For each carrier ic, selection of the transmitter with the highest Q pilot txi M b ic , tx BS ic M i .
k
Analysis of candidate cells, (txBS,ic).

For each pair (txBS,ic), calculation of the uplink load factor:
UL
I tot tx BS ic
UL
UL
X R99 tx BS ic k = ------------------------------- + X
UL
N tot tx BS ic
UL
X corresponds to the load rise due to the mobile. For information on how this parameter is calculated, see "Admission
Control in the R99 Part" on page 274.
Rejection of bad candidate cells if the pilot is not received or if the uplink load factor is exceeded during the admission load
control (if simulation respects a loading factor constraint and Mb was not connected in previous iteration)
pilot
If Q pilot tx BS M b ic Q req Mobility M b then (txBS,ic) is rejected by Mb

k
UL
UL
If X R99 tx BS ic k X max , then (txBS,ic) is rejected by Mb

Else
Keep (txBS,ic) as good candidate cell
If no good candidate cell has been selected, Mb has failed to be connected to the network and is rejected.
For each NodeB having candidate cells, determination of the best carrier, icBS, within the set of candidate cells of the
NodeB.
For MC-HSDPA and DB-MC-HSDPA users, this carrier is referred to as the "anchor" carrier.
If a given carrier is specified for the service requested by Mb
ic BS M b is the carrier specified for the service
Else the carrier selection mode defined for the site equipment is considered.
231

If carrier selection mode is Min. UL Load Factor

UL
ic BS M b is the carrier where we obtain the lowest X R99 tx BS ic k

Else if carrier selection mode is Min. DL Total Power
ic BS M b is the carrier where we obtain the lowest P tx tx BS ic k
Else if carrier selection mode is Random
ic BS M b is randomly selected
Else if carrier selection mode is "Sequential"
UL
UL
ic BS M b is the first carrier where X R99 tx BS ic k X max

Endif
Determination of the best serving cell, (txBS,icBS)
max
(tx BS,ic BS) k M b is the best serving cell ( BestCell k M b ) and its pilot quality is Q pilot M b
k
In the following lines, we will consider ic as the carrier used by the best serving cell
Selection of the second serving cell for DC-HSDPA users
MC-HSDPA and DB-MC-HSDPA users are processed as DC-HSDPA users.
If txBS supports multi-cell HSDPA and if it has several carriers, selection of the second carrier, ic2, among the adjacent carriers.
For each carrier adjacent to the best serving carrier, icp, calculation of Q pilot tx BS ic p M b
k
Selection of the carrier, ic2, with the highest Q pilot tx BS ic p M b

k
pilot
If Q pilot tx BS ic 2 M b Q req Mobility M b then (txBS,ic2) is rejected by Mb

k
Else
Keep (txBS,ic2) as second serving cell
Active Set Determination
For each station txi containing Mb in its calculation area, using ic , and, if neighbours are used, neighbour of BestCell k M b
BTS P c txi M b ic
Calculation of Q pilot txi M b ic = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------DL
DL
DL
DL
Term
k
P tot txi ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0
BTS P c txi M b ic
Q pilot txi M b ic = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------k
DL
DL
DL
DL
txi
ic
+
I
ic
+
I
ic
+
I
ic
extra
inter carrier
inter techno log y
intra
Term
+ N0
1 BTS P c txi M b ic
Rejection of txi from the active set if difference with the best server is too high
max
If Q pilot M b Q pilot txi M b ic AS_Th BestCell k M b then txi is rejected

k
Else txi is included in the Mb active set

Rejection of a station if the mobile active set is full
Station with the lowest Q pilot in the active set is rejected
k
EndFor
Uplink Power Control
R99 req
Calculation of the terminal power required by Mb to obtain the R99 radio bearer: P term
For each cell (txi,ic) of the Mb active set
232
M b ic k

AT332_TRR_E0
Calculation of quality level on Mb traffic channel at (txi,ic), with the minimum power allowed on traffic channel for the Mb
service
req
P term R99 M b ic k 1
UL
P b R99 txi M b ic = ---------------------------------------------------L T txi M b
UL
UL
UL
UL
UL
P b DPDCH txi M b ic = P b R99 txi M b ic 1 r c

UL
P b DPCCH txi M b ic = P b R99 txi M b ic r c

UL
UL
UL
UL
UL
P b R99 txi M b ic = P b DPCCH txi M b ic + P b DPDCH txi M b ic if the user is active,

P b R99 txi M b ic = P b DPCCH txi M b ic if the user is inactive,
UL
term P b DPDCH txi M b ic k

UL
UL
- G UL
Q tch txi M b ic k = --------------------------------------------------------------------------------------------------------------------------------------------p Service M b G div
UL
Tx
UL
N tot txi ic 1 F MUD term P b R99 txi M b ic k 1
If user selects "Total noise",
UL
term P b DPDCH txi M b ic k

UL
UL
UL
Q tch txi M b ic k = ------------------------------------------------------------------------- G p Service Mb G div
UL
N tot txi ic
End For
If (Mb is in not in handoff)
UL
UL
Q k M b = Q tch txi M b ic k
Else if (Mi is in softer handoff)
UL
UL
Q k M b = f rake efficiency
UL
Q tch txi M b ic k
txi ActiveSet
Else if (Mb is in soft, or softer/soft without MRC)

UL
Qk Mb =
UL
UL
Max Q tch txi M b ic k G macro diversity 2 links

txi ActiveSet
Else if (Mb is in soft/soft)

UL
Qk Mb =
UL
UL
Max Q tch txi M b ic k G macro diversity 3 links

txi ActiveSet
Else if (Mb is in softer/soft with MRC)

UL
Qk Mb
UL
UL
UL
UL
= Max f rake efficiency
Q tch ic Q tch ic G macro diversity 2 links
other site
txi ActiveSet
samesite
End If
UL
Q req Service M b Mobility M b

req
- P req
P term R99 M b ic k = ------------------------------------------------------------------------------------term R99 M b ic k 1
UL
Qk Mb
If compressed mode is operated,
Compressed mode is operated if Mi and Sj support compressed mode, and
Resulting
CM activation
txi M b ic Q pilot
Either Q pilot
Or P c txi M b ic RSCP pilot
CM activation
if the Ec/I0 Active option is selected,
if the RSCP Active option is selected.
233

UL
UL
Q req Service M b Mobility M b Q req Service M b Mobility M b

req
- P req
P term R99 M b ic k = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------term R99 M b ic k 1
UL
Qk Mb
req
min
req
max
req
min
If P term R99 M b ic k P term M b then P term R99 M b ic k = P term txi M b

If P term R99 M b ic k P term M b then Mb cannot select any cell and its active set is cleared
R99
If TP P UL M b TP Max UL txi ic then Mb cannot be connected

Endif
Downlink Power Control
If (mobile does not use a packet switched service that is inactive on the downlink)
For each cell (txi,ic) in Mb active set
Calculation of quality level on (txi,ic) traffic channel at Mb with the minimum power allowed on traffic channel for the Mb
service
min
P tch Service M b
DL
P b txi M b ic = ----------------------------------------------L T txi M b
DL
BTS P b txi M b ic k
DL
DL
- G DL
Q tch txi M b ic k = -------------------------------------------------------------------------------------------------------------------------p Service M b G div
DL
DL
N tot ic 1 F ortho BTS P b txi M b ic k 1
If the user selects the option "Total noise"
DL
DL
Q tch txi
BTS P b txi M b ic k
DL
- G DL
M b ic k = -------------------------------------------------------p Service M b G div
DL
N tot ic
End For
DL
DL
DL
Q tch txi M b ic k
txi ActiveSet
Do
For each cell (txi,ic) in Mb active set
Calculation of the required power for DL traffic channel between (txi,ic) and Mb:
DL
Q req Service M b Mobility M b

req
- P min
P tch txi M b ic k = ------------------------------------------------------------------------------------tch Service M b
DL
Qk Mb
If compressed mode is operated.
DL
DL
Q req Service M b Mobility M b Q req Service Mb Mobility M b

req
- P min
P tch txi M b ic k = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------tch Service M b
DL
Qk Mb
Compressed mode is operated if Mi and Sj support compressed mode, and
req
Resulting
CM activation
txi M b ic Q pilot
Either Q pilot
Or P c txi M b ic RSCP pilot
CM activation
max
if the Ec/I0 Active option is selected,
if the RSCP Active option is selected.
max
If P tch txi M b ic k P tch Service M b then txi ic is set to P tch

DL
max
Recalculation of a decreased Q req (a part of the required quality is managed by the cells set to P tch )
req
P tch Service M b
DL
P b txi M b ic = ---------------------------------------------L T txi M b
234

AT332_TRR_E0
DL
BTS P b txi M b ic
DL
DL
- G DL
Q tch txi M b ic k = ----------------------------------------------------------------------------------------------------------------p Service M b G div
DL
DL
N tot ic 1 F ortho BTS P b txi M b ic
DL
DL
DL
If the user is inactive, then his contribution to interference in the calculation of N tot ic is P b txi M b ic r c .
EndFor
DL
DL
DL
Q tch txi M b ic k
txi ActiveSet
While
DL
Qk Mb
DL
Q req Service M b
Mobility M b and Mb active set is not empty
R99
If TP P DL M b TP Max DL txi ic then Mb cannot be connected

Endif
Uplink and Downlink Interference Update
Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones).
For each cell (txi,ic)
UL
Update of N tot txi ic

EndFor
For each mobile Mi
DL
Update of N tot ic
EndFor
EndFor
Control of Radio Resource Limits (OVSF Codes, Cell Power, Channel Elements, Iub Backhaul Throughput)
P tx txi ic
DL
While ----------------------------k %Powermax
P max
Rejection of the mobile with the lowest service priority starting from the last admitted
EndFor
While N
Codes
Codes
txi ic k N max txi ic
EndFor
For each NodeB, Ni
While N
CE DL
CE DL
N i k N max
Ni
While N
CE UL
CE UL
N i k N max
Ni
EndFor
For each NodeB, Ni
Max
While TP Iub DL N I k TP Iub DL N I

Max
While TP Iub UL N I k TPIub UL N I

235

EndFor
Uplink Load Factor Control
UL
UL
For each cell (txi,ic) with X R99 txi ic X max

EndFor
UL
UL
While at least one cell with X R99 txi ic X max exists.
4.3.2.3 HSDPA Part of the Algorithm

HSDPA BE, HSDPA VBR, HSPA BE and HSPA VBR service users active on DL as well as all HSPA CBR service users (i.e., active and
inactive), unless they have been rejected during the R99 part of the algorithm, are then evaluated by the HSDPA part of the
algorithm.
4.3.2.3.1
HSDPA Power Allocation

The total transmitted power of the cell ( P tx ic ) is the sum of the transmitted R99 power, the HSUPA power and the
transmitted HSDPA power.
P tx ic = P tx R99 ic + P tx H SDPA ic + P HSUPA ic
In case of a static HSDPA power allocation strategy, Atoll checks in the simulation that:
DL
P tx ic P max ic %Power max

where:
DL
%Power max is the maximum DL load allowed.

Therefore, if the maximum DL load is set to 100%, we have:
P tx ic P max ic
In case of dynamic HSDPA power allocation strategy, Atoll checks in the simulation that:
DL
P tx R99 ic + P HSUPA ic P max ic %Power max

And it calculates the available HSDPA power as follows:
P HSDPA ic = P max ic P Headroom ic P tx R99 ic P HSUPA ic
4.3.2.3.2
Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users

The number of HS-SCCH channels ( n HS SCCH ) is the maximum number of HS-SCCH channels that the cell can manage. This
parameter is used to manage the number of BE and VBR service users simultaneously connected to an HSDPA bearer. This
parameter is not taken into account for CBR service users as HS-SCCH-less operation (i.e., HS-DSCH transmissions without any
accompanying HS-SCCH) is performed.
Each HSDPA BE, HSDPA VBR, HSPA BE and HSPA VBR service user consumes one HS-SCCH channel. Therefore, at a time (over
a transmission time interval), the number of these users connected to an HSDPA bearer cannot exceed the number of HSSCCH channels per cell.
The maximum number of HSDPA bearer users ( n max ) corresponds to the maximum number of HSDPA bearer users that the
cell can support. Here, all HSDPA bearer users, i.e., HSDPA BE, HSDPA VBR, HSPA BE and HSPA VBR and HSPA CBR service users,
are taken into consideration.
Let us assume there are 30 users in the cell:
10 HSPA CBR service users with any activity status.

2 HSDPA VBR service users active on DL.
18 HSDPA BE and HSPA BE service users active on DL.
All users are connected to the A-DCH R99 bearer. Finally, the number of HS-SCCH channels and the maximum number of
HSDPA bearer users respectively equal 4 and 25.
The scheduler manages the maximum number of users within each cell. CBR service users have the highest priority and are
processed first, in the order established during the generation of the user distribution. After processing the CBR service users,
Atoll processes the remaining HSDPA bearer users (i.e., HSDPA VBR, HSPA VBR, HSDPA BE and HSPA BE service users). VBR
236

AT332_TRR_E0
service users have the highest priority and are managed before BE service users. For each type of service, the scheduler ranks
the users according to the selected scheduling technique. Users are treated as described in the figure below.
Figure 4.3: Connection status of HSDPA bearer users
All CBR service users may be served if there are enough HSDPA power, Iub backhaul throughput and OVSF codes
available in order for them to obtain the lowest HSDPA bearer that provides a peak RLC throughput higher or equal to
the minimum throughput demand defined for the service. In this case, they will be connected. Else, they will be
rejected.
The two VBR service users may be simultaneously served if there are enough HSDPA power, Iub backhaul throughput
and OVSF codes available in order for them to obtain an HSDPA bearer that provides a peak RLC throughput higher or
equal to the minimum throughput demand defined for the service. In this case, they will be connected. Else, they will
be rejected.
Then, among the BE service users:
4.3.2.3.3
The first two users may be simultaneously served if there are enough HSDPA power, Iub backhaul throughput and
OVSF codes available in order for them to obtain an HSDPA bearer. In this case, they will be connected. Else, they
will be delayed.
The next eleven ones will be delayed since there are no longer HS-SCCH channels available. Their connection
status will be "HS-SCCH Channels Saturation".
Finally, the last five users will be rejected because the maximum number of HSDPA bearer users has been fixed to
25. Their connection status will be "HSDPA Scheduler Saturation".
HSDPA Bearer Allocation Process

The HSDPA bearer allocation process depends on the type of service requested by the user. As explained before, CBR service
users have the highest priority and are processed first, in the order established during the generation of the user distribution.
After processing the CBR service users, the scheduler ranks the remaining users (i.e., VBR and BE service users) and shares the
cell radio resources between them. VBR service users have the highest priority and are managed before BE service users.
CBR Service Users
Let us focus on the ten CBR service users mentioned in the example of the previous paragraph "Number of HS-SCCH Channels
and Maximum Number of HSDPA Bearer Users" on page 236. Fast link adaptation is carried out on these users in order to
determine if they can obtain an HSDPA bearer that provides a peak RLC throughput higher or equal to the service minimum
throughput demand. As HS-SCCH less operation is performed, only HSDPA bearers using the QPSK modulation and two HSPDSCH channels at the maximum can be selected and allocated to the users. The users are processed in the order established
during the generation of the user distribution and the cells available HSDPA power is shared between them as explained
below. Several CBR service users can share the same HSDPA bearer. Then, Atoll calculates the HSDPA bearer consumption ( C
in %) for each user and takes into account this parameter when it determines the resources consumed by the user (i.e., the
HSDPA power used, the number of OVSF codes and the Iub backhaul throughput).
In the bearer allocation process shown below, the 10 CBR service users are represented by Mj, with j = 1 to 10. And, the initial
values of their respective HSDPA powers is 0, i.e. PHSDPA(B(MX)) = 0, where X = 0 to 10. These power values are assigned one
by one by the scheduler, so that with their allocated values, looped back to the starting point, are used in successive steps.
237

For the user, Mj, with j varying from 1 to 10:

PHSDPA
j 1
(PHSDPA
X 0
(M X )) served
Sufficient HS-SCCH
power to reach the
minimum quality
threshold?
No
Mj is rejected
Yes
Enough
16-bit OVSF codes
available to support the
lowest HSDPA bearer
allocated?
No
Mj is rejected
Yes
Sufficient Iub
backhaul throughput
to support the lowest
HSDPA bearer
allocated?
No
Mj is rejected
Yes
Sufficient HSDPA
power to obtain
the lowest HSDPA
bearer allocated?
No
Mj is rejected
Yes
Determination of the Best HSDPA Bearer BB(Mj)
Cell and UE both capable

of supporting BB(Mj)?
Yes
BB(Mj) selected
B(Mj) = BB(Mj)
No
Bearer Downgrading
B(Mj)
RLC Peak Rate of

B(Mj) > Mj Min Throughput
Demand?
No
Mj is rejected
Yes
Allocation of Min Throughput Demand to Mj
Mj connected with B(Mj)
(PHSDPA(Mj))served=PHS-PDSCH(B(Mj)) x C(B(Mj))
Update of Available Radio Resources
No
Mj = M10?
Yes
Resource allocation for Variable Bit Rate and

Best Effort service users
Figure 4.4: HSDPA Bearer Allocation Process for CBR Service Users
VBR Service Users
After processing the CBR service users, the scheduler shares the cells remaining resources between HSDPA and HSPA VBR
service users. Let us focus on the two HSDPA - VBR service users mentioned in the example of the previous paragraph,
"Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users" on page 236. A new fast link adaptation is
carried out on these users in order to determine if they can obtain an HSDPA bearer that provides a peak RLC throughput
higher or equal to the service minimum throughput demand. They are processed in the order defined by the scheduler and
the cells HSDPA power available after all CBR service users have been served is shared between them as explained below.
In the bearer allocation process shown below, the 2 VBR service users are represented by Mj, with j = 1 to 2. And, the initial
238

AT332_TRR_E0
PHSDPA
j 1
(PHSDPA
X 0
(M X )) served
Sufficient HS-SCCH
power to reach the
minimum quality
threshold?
No
Mj is rejected
Yes
Enough
16-bit OVSF codes
lowest HSDPA bearer
allocated?
No
Mj is rejected
Yes
Sufficient Iub
backhaul throughput
to support the lowest
HSDPA bearer
allocated?
No
Mj is rejected
Yes
Sufficient HSDPA
power to obtain
the lowest HSDPA
bearer allocated?
No
Mj is rejected
Yes
Determination of the Best HSDPA Bearer BB(Mj)
Bearer Downgrading B(Mj) until:

1. Cell and UE both capable of supporting B(Mj)
And
2. RLC Peak Rate of B(Mj) > Mj Min Throughput
Demand
RLC Peak Rate of

B(Mj) > Mj Min Throughput
Demand?
No
Mj is rejected
Yes
Mj connected with B(Mj)
(PHSDPA(Mj))served=PHS-PDSCH(B(Mj)) +nHS-SCCHxPHS-SCCH(Mj)
No
Mj = M2?
Yes
Resource allocation for Best Effort service users
Figure 4.5: HSDPA Bearer Allocation Process for VBR Service Users
BE Service Users
After processing the VBR service users, the scheduler shares the cells remaining resources between BE service users. Let us
focus on the HSDPA and HSPA BE service users, especially on the first four users mentioned in the example of the previous
paragraph, "Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users" on page 236. A new fast link
adaptation is carried out on these users in order to determine if they can obtain an HSDPA bearer. They are processed in the
order defined by the scheduler and the cells HSDPA power available after all CBR and VBR service users have been served is
shared between them as explained below.
In the bearer allocation process shown below, the 4 BE service users are represented by Mj, with j = 1 to 4. And, the initial
239

Figure 4.6: HSDPA Bearer Allocation Process for BE Service Users
4.3.2.3.4
Fast Link Adaptation Modelling

Fast link adaptation (or Adaptive Modulation and Coding) is used in HSDPA. The power on the HS-DSCH channel is transmitted
at a constant power while the modulation, the coding and the number of codes are changed to adapt to the radio conditions
variations. Based on the reported channel quality indicator (CQI), the node-B may change every 2ms the modulation (QPSK,
16QAM, 64QAM), the coding and the number of codes during a communication.
Atoll calculates for each user either the best pilot quality (CPICH Ec/Nt) or the best HS-PDSCH quality (HS-PDSCH Ec/Nt); this
depends on the option selected in Global parameters (HSDPA part): CQI based on CPICH quality or CQI based on HS-PDSCH
quality (CQI means channel quality indicator). Then, it determines the HS-PDSCH CQI, calculates the best bearer that can be
used and selects the suitable bearer so as to comply with cell and terminal user equipment HSDPA capabilities. Once the
bearer selected, Atoll finds the highest downlink throughput that can be provided to the user and may calculate the
application throughput.
240

AT332_TRR_E0
CQI Based on CPICH Quality

When the option CQI based on CPICH quality is selected, Atoll proceeds as follows.
1. CPICH Quality Calculation
Ec
Let us assume the following notation: ------ ic
corresponds to the CPICH quality.
Nt
pilot
Two options, available in Global parameters, may be used to calculate Nt: option Without useful signal or option Total noise.
Therefore, we have:
BTS P c ic
Eci
----
- for the total noise option,
Nt ic pilot = ----------------------------------------DL
N tot ic
And
BTS P c ic
Eci
----for the without useful signal option.
ic
= -------------------------------------------------------------------------------- Nt
pilot
DL
N tot ic 1 BTS P c ic
i
With
DL
DL
DL
DL
DL
term
N tot ic = I intra ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0
DL
DL
DL
P SCH ic
P SCH ic
DL
term
I intra ic = P tot ic + BTS 1 F MUD 1 P tot ic ------------------- BTS P tot ic ------------------
LT
LT
txi
txi
txi
DL
I extra ic =
DL
P tot ic
txj j i
Ptot icadj
DL
DL
j
-----------------------------------I inter carrier ic = txj
RF ic ic adj
icadj is a carrier adjacent to ic.

RF ic ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
DL
I inter techno log y ic is the inter-technology interference at the receiver on ic.

DL
I inter techno log y ic =
ni
ic i is the i
th
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
interfering carrier of an external transmitter
Tx m
ICPic ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the
i
frequency gap between ic i (external network) and ic .

P pilot ic
P c ic = --------------------i
LT
i
L path L Tx L term L body L indoor E Shadowing 3

- ( )
L T = ------------------------------------------------------------------------------------------------------------------G Tx G term
term
BTS , and N 0
3.
are defined in "Inputs" on page 213.
In the HSDPA coverage prediction, L T is calculated as follows:

-)
L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx G term
241

Atoll performs intra-cell interference computations based on the total power. You can
instruct Atoll to use maximum power by adding the following lines in the Atoll.ini file:
[CDMA]
PmaxInIntraItf = 1
In this case, Atoll considers the following formula:

P max ic P SCH ic
P max ic
DL
term
-
I intra ic = -------------------+ BTS 1 F MUD 1 ---------------------------------------------
LT
LT
P max ic P SCH ic
-
BTS ---------------------------------------------
LT
2. CPICH CQI Determination

Let us assume the following notation: CQI pilot corresponds to the CPICH CQI. CQI pilot is read in the table
Ec
. This table is defined for the terminal reception equipment and the selected mobility.
CQI pilot = f ------ ic
Nt
pilot
3. HS-PDSCH Quality Calculation
Atoll proceeds as follows:
1st step: Atoll calculates the HS-SCCH power ( P HS SCCH ).
P HS SCCH ic is the HS-SCCH power on carrier ic. It is either fixed by the user (when the option HS-SCCH Power Dynamic
Allocationin the cell property dialog box is unchecked) or dynamically calculated (when the option HS-SCCH Power Dynamic
Allocation is selected).
req
Ec
In this case, the HS-SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (noted ------ ic
). It is
Nt
HS SCCH
specified in mobility properties.
We have:
BTS P c ic
Eci
----= ------------------------------- for the total noise option,
ic
Nt
HS SCCH
DL
N tot ic
And
BTS P c ic
Eci
----
ic
=
---------------------------------------------------------------------------------------------------------------------------- for the without useful signal option.
Nt
HS SCCH
DL
term
N tot ic 1 F ortho 1 F MUD BTS P c ic
i
With
DL
DL
DL
DL
DL
term
DL
DL
DL
P SCH ic
P SCH ic
DL
term
I intra ic = P tot ic + BTS 1 F MUD 1 F ortho P tot ic ------------------- BTS P tot ic ------------------
LT
LT
txi
txi
txi
DL
I extra ic =
DL
P tot ic
txj j i
Ptot icadj
DL
DL
j
RF ic ic adj

DL
242

AT332_TRR_E0
DL
ic i is the i
th
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
Tx m
i

P HS SCCH ic
P c ic = ------------------------------i
LT
i
and
- ( )
L T = ------------------------------------------------------------------------------------------------------------------G Tx G term
term
term
BTS , F ortho , F MUD and N 0
Therefore,
req
EcDL
---- ic
N tot ic
Nt
HS SCCH
- L T for the total noise option,
P HS SCCH ic = ----------------------------------------------------------------i
BTS
And
req
EcDL
----
ic
Nt
HS SCCH N tot ic
------------------------------------------------------------------------------------------------------------------------------------------P HS SCCH ic =
L T for the without useful signal option.
req
i
Ecterm
1 + 1 F
----
ic
BTS
ortho
MUD
Nt
HS SCCH
2nd step: Atoll calculates the HS-PDSCH power ( P HS PDSCH ).

P HSDPA ic is the power available for HSDPA on the carrier ic. This parameter is either a simulation output, or a user-defined
cell input.
P HSDPA ic = P HS PDSCH ic + n HS SCCH P HS SCCH ic
Therefore, we have:
P HS PDSCH ic = P HSDPA ic n HS SCCH P HS SCCH ic
n HS SCCH is the number of HS-SCCH channels.
3rd step: Then, Atoll evaluates the HS-PDSCH quality
Ec
corresponds to the HS-PDSCH quality.
Nt
HS PDSCH
We have:
BTS P c ic
Eci
---- ic
= ------------------------------- for the total noise option,
Nt
HS PDSCH
DL
N tot ic
And
BTS P c ic
Ec
i
----- ic
= ---------------------------------------------------------------------------------------------------------------------------- for the without useful signal option.
Nt
HS PDSCH
P c ic
DL
term
i
N tot ic 1 F ortho 1 F MUD BTS --------------n
4.

-)
L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx G term
243

Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5).
With
DL
DL
DL
DL
DL
term
DL
DL
DL
P SCH ic
P SCH ic
DL
term
I intra ic = P tot ic + BTS 1 F MUD 1 F ortho P tot ic ------------------- BTS P tot ic -------------------
LT
LT
txi
txi
txi
DL
I extra ic =
DL
P tot ic
txj j i
Ptot icadj
DL
DL
j
RF ic ic adj

DL

DL
ni
ic i is the i
th
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
Tx m
ICP ic ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the
i

P HS PDSCH ic
P c ic = ---------------------------------i
LT
i
And
L T = ------------------------------------------------------------------------------------------------------------------- ( )
G Tx G term
term
term
[CDMA]
PmaxInIntraItf = 1

P max ic P SCH ic
P max ic P SCH ic
P max ic
DL
term
- BTS ----------------------------------------------
L
LT
LT
T
4. HS-PDSCH CQI Determination
The best bearer that can be used depends on the HS-PDSCH CQI. Let us assume the following notation: CQI HS PDSCH
corresponds to the HS-PDSCH CQI. Atoll calculates CQI HS PDSCH as follows:
CQI HS PDSCH = CQI pilot P pilot + P HS PDSCH
5. HSDPA Bearer Selection
5.

-)
L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx G term
244

AT332_TRR_E0
Atoll selects the HSDPA bearer associated to this CQI (in the table Best Bearer=f(HS-PDSCH CQI) defined for the terminal
reception equipment and the user mobility) and compatible with the user equipment and cell capabilities.
HSDPA bearers can be classified into two categories:
HSDPA bearers using QPSK and 16QAM modulations: They can be selected for all users connected to HSPA and HSPA+
capable cells. The number of HS-PDSCH channels required by the bearer must not exceed the maximum number of
HS-PDSCH codes available for the cell.
For VBR service users, the selected HSDPA bearer must provide a peak RLC throughput between the minimum and the
maximum throughput demands defined for the service.
For CBR service users, HS-SCCH-less operation (i.e., HS-DSCH transmissions without any accompanying HS-SCCH) is
performed. In this case, the UE is not informed about the transmission format and has to revert to blind decoding of
the transport format used on the HS-DSCH. Complexity of blind detections in the UE is decreased by limiting the
transmission formats that can be used (i.e., the HSDPA bearers available). Therefore, only HSDPA bearers using the
QPSK modulation and two HS-PDSCH channels at the maximum can be selected and allocated to these users.
Additionally, the selected HSDPA bearer must provide a peak RLC throughput higher or equal to the minimum
throughput demand defined for the service.
HSDPA bearers using 64QAM modulation (improvement introduced by the release 7 of the 3GPP UTRA specifications,
referred to as HSPA+): These HSDPA bearers can be allocated to VBR and BE service users connected to cells with
HSPA+ capabilities only. The number of HS-PDSCH channels required by the bearer must not exceed the maximum
number of HS-PDSCH codes available for the cell. For VBR service users, the selected HSDPA must provide a peak RLC
throughput between the minimum and the maximum throughput demands defined for the service. These HSDPA
bearers cannot be allocated to CBR service users.
Atoll considers an HSDPA bearer as compatible with the user equipment if:
The transport block size does not exceed the maximum transport block size supported by the user equipment.
The number of HS-PDSCH channels required by the bearer does not exceed the maximum number of HS-PDSCH
channels that the terminal can use.
The modulation is supported by the user equipment.
When there are several HSDPA bearers compatible, Atoll selects the HSDPA bearer that provides the highest RLC peak
throughput. When several HSDPA bearers can supply the same RLC peak throughput, Atoll chooses the HSDPA bearer with
the highest modulation scheme. Finally, if no HSDPA bearer is compatible, Atoll allocates a lower HSDPA bearer compatible
with the user equipment and cell capabilities which needs fewer resources.
Lets consider the following examples.
Example1: One HSDPA BE service user with category 13 user equipment and a 50km/h mobility.
The user equipment capabilities are:
Maximum transport block size: 35280 bits

Maximum number of HS-PDSCH channels: 15
Highest modulation supported: 64QAM
MIMO Support: No
Figure 4.7: HSDPA UE Categories Table

The cell to which the user is connected supports HSPA+ functionalities (i.e. 64QAM modulation in the DL and MIMO systems)
and the maximum number of HS-PDSCH channels is 15.
245

1st case: The CQI experienced by the user equals 26. Therefore, Atoll can choose between two HSDPA bearers, the bearer
indexes 26 and 31.
Characteristics of the bearer index 26 are:
Transport block size: 17237 bits

Number of HS-PDSCH channels used: 12
16QAM modulation is used
Peak RLC Throughput: 8.32 Mb/s

Both HSDPA bearers are compatible with the user equipment and cell capabilities. Atoll selects the HSDPA bearer that
provides the highest RLC peak throughput, i.e. the bearer index 26.
Figure 4.8: HSDPA Radio Bearers Table

2nd case: The CQI experienced by the user equals 27. Therefore, Atoll can choose between two HSDPA bearers, the bearer
indexes 27 and 32.


Both HSDPA bearers are compatible with the user equipment and cell capabilities and the peak RLC throughput they provide
is the same. Atoll selects the HSDPA bearer using the highest modulation scheme, i.e. the bearer index 32.
Example 2: One HSDPA BE user experiencing a CQI of 26.
Therefore, Atoll can choose between two HSDPA bearers, the bearer indexes 26 and 31.
246


AT332_TRR_E0

1st case: The user equipment category is 9. The cell to which the user is connected supports HSPA+ functionalities (i.e. 64QAM
modulation in the DL and MIMO systems) and the maximum number of HS-PDSCH channels is 15.
The user equipment characteristics are the following:

MIMO Support: No
The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the terminal. Only the
bearer index 26 is compatible with the user equipment capabilities. Atoll selects it.
2nd case: The user equipment category is 8. The cell to which the user is connected supports HSPA+ functionalities (i.e. 64QAM
modulation in the DL and MIMO systems) and the maximum number of HS-PDSCH channels is 15.
The user equipment characteristics are the following:

MIMO Support: No
Here, none of HSDPA bearers are compatible with the user equipment capabilities.
The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the terminal. With the
bearer index 26, the number of HS-PDSCH channels (12) exceeds the maximum number of HS-PDSCH channels the terminal
can use (10), and the transport block size (17237 bits) exceeds the maximum transport block size (14411 bits) the terminal can
carried.
In the HSDPA Radio Bearer table, Atoll selects a lower HSDPA bearer compatible with cell and UE category capabilities. It
selects the bearer index 25.
The number of HS-PDSCH channels (10) does not exceed the maximum number of HS-PDSCH channels the terminal
can use (10) and the maximum number of HS-PDSCH channels available at the cell level (15),
The transport block size (14411 bits) does not exceed the maximum transport block size (14411 bits) the terminal can
carried.
16QAM modulation is supported by the terminal and the cell.
3rd case: The user equipment category is 13. The cell to which the user is connected supports HSPA functionalities and the
maximum number of HS-PDSCH channels is 15.
The user equipment capabilities are:

Maximum number of HS-PDSCH channels:15
MIMO Support: No
The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the cell. On the other hand,
the bearer index 26 is compatible with cell and UE category capabilities. Therefore, it is allocated.
6. HS-PDSCH Quality Update
Once the bearer selected, Atoll exactly knows the number of HS-PDSCH channels. Therefore, when the method Without
useful signal is used, it may recalculate the HS-PDSCH quality with the real number of HS-PDSCH channels (A default value
(5) was taken into account in the first HS-PDSCH quality calculation).
CQI Based on HS-PDSCH Quality
When the option CQI based on HS-PDSCH quality is selected, Atoll proceeds as follows.
1. HS-PDSCH Quality Calculation
1st step: Atoll calculates the HS-SCCH power ( P HS SCCH ).
247

P HS SCCH ic is the HS-SCCH power on carrier ic. It is either fixed by the user (when the option HS-SCCH Power Dynamic
Allocationin the cell property dialog box is unchecked) or dynamically calculated (when the option HS-SCCH Power Dynamic
Allocation is selected).
req
Ec
In this case, the HS-SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (noted ------ ic
). It is
Nt
HS SCCH
specified in mobility properties.
We have:
BTS P c ic
Eci
---- ic
Nt
HS SCCH
DL
N tot ic
And
BTS P c ic
Eci
----
ic
=
---------------------------------------------------------------------------------------------------------------------------- for the without useful signal option.
Nt
HS SCCH
DL
term
i
With
DL
DL
DL
DL
DL
term
DL
DL
DL
P SCH ic
P SCH ic
DL
term
I intra ic = P tot ic + BTS 1 F MUD 1 F ortho P tot ic ------------------- BTS P tot ic -------------------
LT
LT
txi
txi
txi
DL
I extra ic =
DL
P tot ic
txj j i
Ptot icadj
DL
DL
j
RF ic ic adj

DL

DL
ni
ic i is the i
th
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
Tx m
i

P HS SCCH ic
P c ic = ------------------------------i
LT
i
And
L T = ------------------------------------------------------------------------------------------------------------------- ( )
G Tx G term
term
term
Therefore,
6.

-)
L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx G term
248

AT332_TRR_E0
req
EcDL
----
ic
HS SCCH N tot ic
Nt
----------------------------------------------------------------P HS SCCH ic =
L T for the total noise option,
i
BTS
And
req
EcDL
----
Nt ic HS SCCH N tot ic
- L T for the without useful signal option.

P HS SCCH ic = ------------------------------------------------------------------------------------------------------------------------------------------req
i
Ec
term
1 + 1 F
ortho 1 F MUD ------ ic

BTS
Nt
HS SCCH
2nd step: Atoll calculates the HS-PDSCH power ( P HS PDSCH )
P HSDPA ic is the power available for HSDPA on the carrier ic. This parameter is either a simulation output, or a user-defined
cell input.
Therefore, we have:
n HS SCCH is the number of HS-SCCH channels.
3rd step: Then, Atoll evaluates the HS-PDSCH quality
Ec
Nt
HS PDSCH
We have:
BTS P c ic
Ec
i
----
Nt ic HS PDSCH
DL
N tot ic
And
BTS P c ic
Eci
----
- for the without useful signal option.
Nt ic HS PDSCH = ---------------------------------------------------------------------------------------------------------------------------P c ic
DL
term
i
N tot ic 1 F ortho 1 F MUD BTS --------------n
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5). Then, it calculates the HS-PDSCH CQI and
the bearer to be used. Once the bearer selected, Atoll exactly knows the number of HS-PDSCH channels and recalculates the
HS-PDSCH quality with the real number of HS-PDSCH channels.
With
DL
DL
DL
DL
DL
term
DL
DL
DL
P SCH ic
P SCH ic
DL
term
I intra ic = P tot ic + BTS 1 F MUD 1 F ortho P tot ic ------------------- BTS P tot ic ------------------LT
LT
txi
txi
txi
DL
I extra ic =
DL
P tot ic
txj j i
Ptot icadj
DL
DL
j
RF ic ic adj

DL
249

DL
ic i is the i
th
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
Tx m
i

P HS PDSCH ic
P c ic = ---------------------------------i
LT
i
And
- ( )
L T = ------------------------------------------------------------------------------------------------------------------G Tx G term
term
term
[CDMA]
PmaxInIntraItf = 1

P max ic P SCH ic
P max ic P SCH ic
P max ic
DL
term
- BTS ----------------------------------------------
L
LT
LT
T
2. HS-PDSCH CQI Determination
Let us assume the following notation: CQI HS PDSCH corresponds to the HS-PDSCH CQI. CQI HS PDSCH is read in the table
Ec
. This table is defined for the terminal reception equipment and the specified
CQI HS PDSCH = f ------ ic
Nt
HS PDSCH
mobility.
3. HSDPA Bearer Selection
The bearer is selected as described in "HSDPA Bearer Selection" on page 244.
4.3.2.3.5
MIMO Modelling
MIMO - Transmit Diversity
If the user is connected to a cell that supports HSPA+ with transmit diversity and if he has a MIMO-capable terminal (i.e., a
terminal with an HSDPA UE category supporting MIMO), he will benefit from downlink diversity gain on the HS-PDSCH Ec/Nt.
EcEc
DL
DL
----= ------ ic
+ G TD + G TD in dB
ic
Nt
HS PDSCH
Nt
HS PDSCH
Where
DL
G TD is the downlink transmit diversity gain (in dB) corresponding to the numbers of transmission and reception antenna ports
(respectively defined in the transmitter and terminal properties).
DL
G TD is the additional diversity gain in downlink (in dB). It is defined for the clutter class of the user.
7.

-)
L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx G term
250

AT332_TRR_E0
MIMO - Spatial Multiplexing

If the user is connected to a cell that supports HSPA+ with spatial multiplexing and if he has a MIMO-capable terminal (i.e., a
terminal with an HSDPA UE category supporting MIMO), he will benefit from the spatial multiplexing gain in its peak RLC
throughput.
In this case, the peak RLC throughput obtained by the user is the following:
DL
DL
Max
TP P R LC = TP P R LC Index HSDPABearer 1 + f SM Gain G SM 1

Where
DL
TP P R LC Index HSDPABearer is the peak RLC throughput that the selected HSDPA bearer ( Index HSDPABearer ) can provide in the
cell (Txi, ic). It is read in the HSDPA Radio Bearer table.
Max
G SM
is the maximum spatial multiplexing gain (in dB) for a given number of transmission and reception antennas
(respectively defined in the transmitter and terminal properties).

f SM Gain is the spatial multiplexing gain factor defined for the clutter
4.3.2.3.6
Scheduling Algorithms
The scheduler manages the maximum number of users within each cell. CBR service users have the highest priority and are
processed first, in the order established during the generation of the user distribution. After processing the CBR service users,
the scheduler processes the remaining users (i.e., VBR and BE service users). VBR service users have the highest priority and
are managed before BE service users. For each type of service, the scheduler ranks the users according the scheduling
technique. Three scheduling algorithms are available, Max C/I, Round Robin and Proportional Fair. Impact they have on the
simulation result is described in the tables below.
Let us consider a cell with 16 HSDPA and HSPA BE service users. All of them are active on DL and connected to the A-DCH R99
bearer. There is neither CBR service user, nor VBR service user in the cell and the number of HS-SCCH channels and the
maximum number of HSDPA bearer users have been respectively set to 4 and 15.
Max C/I
15 users (where 15 corresponds to the maximum number of HSDPA bearer users defined) enters the scheduler in the same
order as in the simulation. Then, they are sorted in descending order by the channel quality indicator (CQI), i.e. in a best bearer
descending order.
Mobiles
Simulation Rank
Best Bearer (kbps)
DL Obtained
Throughput (kbps)
Connection Status
M1
2400
2400+3.4
Connected
M2
15
2400
1440+3.4
Connected
M3
2080
160+3.4
Connected
M4
2080
3.4
Delayed
M5
10
2080
3.4
Delayed
M6
12
2080
3.4
Delayed
M7
13
2080
3.4
Delayed
M8
14
2080
3.4
Delayed
M9
1920
3.4
Delayed
M10
1600
3.4
Delayed
M11
1600
3.4
Delayed
M12
1600
3.4
Delayed
M13
1600
3.4
Delayed
M14
1600
3.4
Delayed
M15
11
1440
3.4
Delayed
M16
16
2080
Scheduler Saturation
Round Robin
Users are taken into account in the same order than the one in the simulation (random order).
251

Mobiles
Simulation Rank
Best Bearer (kbps)
DL Obtained
Throughput (kbps)
Connection Status
M1
1600
1600+3.4
Connected
M2
2400
960+3.4
Connected
M3
1600
3.4
Delayed
M4
1600
3.4
Delayed
M5
1600
3.4
Delayed
M6
1600
3.4
Delayed
M7
1920
3.4
Delayed
M8
2080
3.4
Delayed
M9
2080
3.4
Delayed
M10
10
2080
3.4
Delayed
M11
11
1440
3.4
Delayed
M12
12
2080
3.4
Delayed
M13
13
2080
3.4
Delayed
M14
14
2080
3.4
Delayed
M15
15
2400
3.4
Delayed
M16
16
2080
Scheduler Saturation
Proportional Fair
15 users (where 15 corresponds to the maximum number of HSDPA bearer users defined) enters the scheduler in the same
order as in the simulation. Then, they are sorted in an ascending order according to a new random parameter which
corresponds to a combination of the user rank in the simulation and the channel quality indicator (CQI).
For a user i, the random parameter RP i is calculated as follows:
Simu
RP i = 50 R i
CQI
+ 50 R i
Where,
Simu
Ri
CQI
Ri
is the user rank in the simulation.
is the user rank according to the CQI.

You can change the default weights by editing the Atoll.ini file. For more information, see
the Administrator Manual.
252
CQI Rank
RP
Best Bearer
(kbps)
DL Obtained
Throughput
(kbps)
150
2400
2400
Connected
10
550
1600
960
Connected
M3
550
2080
160
Connected
M4
650
2080
3.4
Delayed
M5
11
700
1600
3.4
Delayed
M6
10
750
2080
3.4
Delayed
M7
12
800
1600
3.4
Delayed
M8
800
1920
3.4
Delayed
Mobiles
Simulation
Rank
M1
M2
Connection
Status
M9
15
850
2400
3.4
Delayed
M10
13
900
1600
3.4
Delayed
M11
12
900
2080
3.4
Delayed
M12
14
1000
1600
3.4
Delayed
M13
13
1000
2080
3.4
Delayed

AT332_TRR_E0
4.3.2.3.7
Mobiles
Simulation
Rank
CQI Rank
RP
Best Bearer
(kbps)
DL Obtained
Throughput
(kbps)
Connection
Status
M14
14
1100
2080
3.4
Delayed
M15
11
15
1300
1440
3.4
Delayed
M16
16
2080
Scheduler
Saturation
Dual-Cell HSDPA
For transmitters that support multi-cell HSDPA mode, the scheduler manages a single queue of users at the Node B.
MC-HSDPA and DB-MC-HSDPA users are processed as DC-HSDPA users if they are connected to two carriers. Otherwise, they
are considered as single-cell HSDPA users. All users belonging to the transmitter, i.e., DC-HSDPA and single-carrier HSDPA
users, are ranked together in a unique list. DC-HSDPA users are considered twice in the list as they may be assigned two
different HSDPA bearers in the two cells.
CBR service users have the highest priority and are processed first, in the order established during the generation of the user
distribution. After processing the CBR service users, the scheduler processes the remaining users (i.e., VBR and BE service
users). VBR service users have the highest priority and are managed before BE service users. For each type of service, the
scheduler ranks the users according the scheduling technique (Max C/I, Round Robin and Proportional Fair). After the users
have been ranked, the scheduler allocates HSDPA resources to each user following the calculated order as long as there are
resources available. Even if there is a unique list of users at the transmitter level, the resources of each cell are not shared and
each carrier has its own pool of resources (number of HS-SCCH channels, maximum number of HSDPA bearer users, HSDPA
power, number of OVSF codes). Only site-level resources (such as the Iub throughput and the channel elements) are shared
between the users of the two cells.
Let us consider a transmitter with 16 BE service users. The transmitter supports the multi-cell HSDPA mode. There is neither
CBR service user, nor VBR service users. All users are active in DL and connected to the A-DCH R99 bearer. Among the users,
there are 6 DC-HSDPA users (i.e., terminal with UE categories 21 to 24).
Simulation Rank
DC-HSDPA Support
Carriers
Comments
Yes
1 and 2
Anchor carrier: 2
No
No
Yes
1 and 2
No
No
No
No
Yes
1 and 2
10
No
11
No
12
Yes
1 and 2
13
No
14
Yes
1 and 2
15
No
16
Yes
1 and 2
Anchor carrier: 2
Anchor carrier: 1
Anchor carrier: 1
Anchor carrier: 1
Anchor carrier: 2
In each cell, the number of HS-SCCH channels and the maximum number of HSDPA bearer users have been respectively set
to 4 and 7.
The scheduling algorithms defined for the two cells are the same as the one selected for the transmitter.
Each DC-HSDPA user is counted twice, once in each cell, as he may be assigned two different HSDPA bearers in the two cells.
Therefore, the scheduler manages the users ranked 1st to 11th (i.e. 4 single-carrier users connected to the first carrier, 4 singlecarrier users connected to the second carrier and 3 DC-HSDPA users). Users ranked 12th to 16th are rejected because the
maximum number of HSDPA bearer users that the scheduler can manage in a cell is exceeded.
Impact the scheduling algorithms have on the simulation results is described in the tables below.
253

Max C/I
7 users from each cell (where 7 corresponds to the maximum number of HSDPA bearer users defined for each cell), i.e., a total
of 14 users enter the scheduler in the same order as in the simulation. Then, they are sorted in the order of decreasing channel
quality indicator (CQI), i.e. in a best bearer descending order.
Mobiles
Carrier
Simulation
Rank
CQI
Best Bearer
(kbps)
DL Obtained
Throughput
(kbps)
Connection
Status
M1
21
3040
3040+3.4
Connected
M2
(DC-HSDPA)
19
2400
2400+3.4
Connected
M3
18
2080
1440+3.4
Connected
M2
(DC-HSDPA)
17
1920
1920
Connected
M4
(DC-HSDPA)
17
1920
960+3.4
Connected
M5
16
1600
3.4
Delayed
M4
(DC-HSDPA)
16
1600
1120
Connected
M6
15
1440
3.4
Delayed
M7
14
1120
3.4
Delayed
M8
10
14
1120
3.4
Delayed
M9
(DC-HSDPA)
13
960
3.4
Delayed
M10
13
960
3.4
Delayed
M9
(DC-HSDPA)
12
800
Delayed
M11
11
12
800
3.4
Delayed
M12
(DC-HSDPA)
1
2
12
14
15
1120
1440
Scheduler
Saturation
M13
13
17
1920
Scheduler
Saturation
M14
(DC-HSDPA)
1
2
14
13
15
960
1440
Scheduler
Saturation
M15
15
17
1920
Scheduler
Saturation
M16
(DC-HSDPA)
1
2
16
12
14
800
1120
Scheduler
Saturation
The scheduled DC-HSDPA users have the following status:
The user ranked 4th (here M2) is connected to an HSDPA bearer in each cell. He obtains a total DL throughput of 4323.4
kbps (2403.4+1920).
kbps (963.4+1120).
The first user (here M9) is delayed in the two cells. He obtains a total DL throughput of 3.4 kbps.
Round Robin
of 14 users enter the scheduler in the same order as in the simulation.
254
Mobiles
Carrier
Simulation
Rank
CQI
Best Bearer
(kbps)
DL Obtained
Throughput
(kbps)
Connection
Status
M1
(DC-HSDPA)
12
800
800
Connected

AT332_TRR_E0
Mobiles
Carrier
Simulation
Rank
CQI
Best Bearer
(kbps)
DL Obtained
Throughput
(kbps)
Connection
Status
M1
(DC-HSDPA)
13
960
960+3.4
Connected
M2
15
1440
1440+3.4
Connected
M3
16
1600
1600+3.4
Connected
M4
(DC-HSDPA)
19
2400
1600+3.4
Connected
M4
(DC-HSDPA)
17
1920
960
Connected
M5
21
3040
480+3.4
Connected
M6
13
960
160+3.4
Connected
M7
14
1120
3.4
Delayed
M8
18
2080
3.4
Delayed
M9
(DC-HSDPA)
16
1600
Delayed
M9
(DC-HSDPA)
17
1920
3.4
Delayed
M10
10
14
1120
3.4
Delayed
M11
11
12
800
3.4
Delayed
M12
(DC-HSDPA)
1
2
12
14
15
1120
1440
Scheduler
Saturation
M13
13
17
1920
Scheduler
Saturation
M14
(DC-HSDPA)
1
2
14
13
15
960
1440
Scheduler
Saturation
M15
15
17
1920
Scheduler
Saturation
M16
(DC-HSDPA)
1
2
16
12
14
800
1120
Scheduler
Saturation
The first user (here M1) is connected to an HSDPA bearer in each cell. He obtains a total DL throughput of 1763.4 kbps
(800+963.4).
kbps (1603.4+960).
The user ranked 9th (here M9) is delayed in the two cells. He obtains a total DL throughput of 3.4 kbps.
Proportional Fair
of 14 users enter the scheduler in the same order as in the simulation. Then, they are sorted in an ascending order according
to a new random parameter which corresponds to a combination of the user rank in the simulation and the channel quality
indicator (CQI).
For a user i, the random parameter RPi is calculated as follows:
Simu
RPi = 50 R i
CQI
+ 50 R i
Where,
Simu
Ri
CQI
Ri
is the user rank in the simulation.
is the user rank according to the CQI.
255

You can change the default weights by editing the Atoll.ini file. For more information, see
DL
Best Bearer Obtained Connection
Throughput
(kbps)
Status
(kbps)
Mobiles
Carrier
Simulation
Rank
M1
DC-HSDPA
19
300
2400
2400+3.4
Connected
M2
21
300
3040
3040+3.4
Connected
M1
DC-HSDPA
17
400
1920
1440
Connected
M3
16
450
1600
800+3.4
Connected
M4
15
500
1440
1120+3.4
Connected
M5
18
550
2080
800+3.4
Connected
M6
DC-HSDPA
13
11
600
960
480+3.4
Connected
M6
DC-HSDPA
12
13
700
800
Delayed
M7
DC-HSDPA
17
700
1920
3.4
Delayed
M8
14
800
1120
3.4
Delayed
M7
DC-HSDPA
16
800
1600
Delayed
M9
13
12
900
960
3.4
Delayed
M10
10
14
10
1000
1120
3.4
Delayed
M11
11
12
14
1250
800
3.4
Delayed
M12
(DC-HSDPA)
1
2
12
14
15
1120
1440
Scheduler
Saturation
Scheduler
Saturation
M13
13
17
1920
Scheduler
Saturation
Scheduler
Saturation
M14
(DC-HSDPA)
1
2
14
13
15
960
1440
Scheduler
Saturation
Scheduler
Saturation
M15
15
17
1920
Scheduler
Saturation
Scheduler
Saturation
M16
(DC-HSDPA)
1
2
16
12
14
800
1120
Scheduler
Saturation
Scheduler
Saturation
CQI
CQI Rank
RP
kbps (2403.4+1440).
The first user (here M6) is connected to an HSDPA bearer in his anchor cell and delayed in the other cell. He obtains a
total DL throughput of 483.4 kbps (483.4+0).
The user ranked 9th (here M7) is delayed in the two cells. He obtains a total DL throughput of 3.4 kbps.
4.3.2.4 HSUPA Part of the Algorithm

HSPA VBR and BE service users active in the UL as well as all HSPA CBR service users (i.e., active and inactive), unless they have
been rejected during the R99 or HSDPA parts of the algorithm, are then evaluated by the HSUPA part of the algorithm. Atoll
manages the maximum number of users within each cell. CBR service users have the highest priority and are processed first,
in the order established during the generation of the user distribution. Then, Atoll considers VBR service users in the order
established during the generation of the user distribution and lastly, it processes BE service users in the order established
during the generation of the user distribution.
Let us assume there are 12 HSPA users in the cell:
256

AT332_TRR_E0
3 CBR service users with any activity status. All of them have been connected to an HSDPA bearer.
2 packet VBR service users. They have been connected to an HSDPA bearer.
7 packet BE service users active on UL. The first two users have been connected to an HSDPA bearer, the last one has
been rejected and the remaining four have been delayed in the HSDPA part.
Finally, the maximum number of HSUPA bearer users equals 10.

In this case, Atoll will consider the first ten HSPA users only and will reject the last two users in order not to exceed the
maximum number of HSUPA bearer users allowed in the cell (their connection status is "HSUPA scheduler saturation").
4.3.2.4.1
Evaluation by the
HSDPA
HSUPA part of the
Connection Status
algorithm
Mobiles
Service
Simulation Rank
M1
CBR
Connected
Yes
M2
CBR
Connected
Yes
M3
CBR
Connected
Yes
M4
VBR
Connected
Yes
M5
VBR
Connected
Yes
M6
BE
Connected
Yes
M7
BE
Connected
Yes
M8
BE
Delayed
Yes
M9
BE
Delayed
Yes
M10
BE
10
Delayed
Yes
M11
BE
11
Delayed
No
M12
BE
12
Rejected
No
Admission Control
During admission control, Atoll selects a list of HSUPA bearers for each user. The selected HSUPA bearers have to be
compatible with the user equipment and capabilities of each HSUPA cell of the active set.
For CBR service users, the list is restricted to HSUPA bearers that provide a peak RLC throughput higher than the minimum
throughput demand.
For VBR service users, the list of compatible bearers is restricted to HSUPA bearers that provide a peak RLC throughput
between the maximum and the minimum throughput demands.
Let us focus on one HSPA-BE service user with category 3 user equipment and a 50km/h mobility. This user is connected to
one cell only. The cell supports HSPA+ functionalities, i.e the cell supports QPSK and 16QAM modulations in the UL.
HSUPA user equipment categories are provided in the HSUPA User Equipment Categories table. The capabilities of the
category 3 user equipment are:
Maximum Number of E-DPDCH codes: 2

TTI 2 ms: No so it supports 10 ms TTI
Minimum Spreading Factor: 4
Maximum Block Size for a 2ms TTI: no value
Maximum Block Size for a 10ms TTI: 14484 bits
Highest Modulation Supported: QPSK
Figure 4.9: HSUPA UE Categories Table

HSUPA bearer characteristics are provided in the HSUPA Bearer table. An HSUPA bearer is described with following
characteristics:
Radio Bearer Index: The bearer index number.
257

TTI Duration (ms): The TTI duration in ms. The TTI can be 2 or 10 ms.
Transport Block Size (Bits): The transport block size in bits.
Number of E-DPDCH Codes: The number of E-DPDCH channels used.
Minimum Spreading Factor: The smallest spreading factor used.
Modulation: the modulation used (QPSK or 16QAM)
Peak RLC Throughput (bps): The RLC peak throughput represents the peak throughput without coding (redundancy,
overhead, addressing, etc.).
HSUPA bearers can be classified into two categories:
HSUPA bearers using QPSK modulation: They can be selected for users connected to HSPA and HSPA+ capable cells.
HSUPA bearers using 16QAM modulation (improvement introduced by the release 7 of the 3GPP UTRA specifications,
referred to as HSPA+). These HSUPA bearers can be allocated to users connected to cells with HSPA+ capabilities only.
Atoll considers an HSUPA bearer as compatible with the category 3 user equipment if:
The TTI duration used by the bearer is supported by the user equipment (10 ms).
The transport block size does not exceed the maximum transport block size supported by the user equipment (14484
bits):
The number of E-DPDCH channels required by the bearer does not exceed the maximum number of E-DPDCH channels
that the terminal can use (2).
The minimum spreading factor used by the bearer is not less than the smallest spreading factor supported by the
terminal (4).
The modulation required by the bearer is supported by the terminal.
The HSUPA bearers compatible with category 3 user equipment are framed in red:
Figure 4.10: HSUPA Radio Bearers Table

Then, during admission control, Atoll checks that the lowest compatible bearer in terms of the required E-DPDCH EcNt does
not require a terminal power higher than the maximum terminal power allowed.
Atoll uses the HSUPA Bearer Selection table. Among the compatible HSUPA bearers, Atoll chooses the one with the lowest
required Ec/Nt threshold.
Here, this is the index 1 HSUPA bearer; the required Ec/Nt threshold to obtain this bearer is -21.7dB.
Ec req
req
Then, from the required Ec/Nt threshold, ------
, Atoll calculates the required terminal power, P term HSUPA .
Nt E DPDCH
Ec req
req
UL
P term HSUPA = ------
L T N tot
Nt E DPDCH
With
UL
tx
UL intra
N tot ic = 1 F MUD term I tot
UL extra
ic + I tot

- ( )
L T = ------------------------------------------------------------------------------------------------------------------G Tx G term
258
UL
tx
ic + I inter carrier ic + N 0

AT332_TRR_E0
tx
UL
intra
term , F MUD , I tot
UL
extra
, I tot
UL
tx
, I inter carrier and N 0 are defined in "Inputs" on page 213.
Figure 4.11: HSUPA Bearer Selection Table

req
Atoll rejects the user if the terminal power required to obtain the lowest compatible HSUPA bearer ( P term HSUPA ) exceeds
the maximum terminal power (his connection status is "HSUPA Admission Rejection").
At the end of this step, the number of non-rejected HSUPA bearer users is n HSUPA . All of them will be connected to an HSUPA
bearer at the end.
4.3.2.4.2
HSUPA Bearer Allocation Process

The HSUPA bearer allocation process depends on the type of service requested by the user. As explained before, CBR service
users have the highest priority and are processed first, in the order established during the generation of the user distribution.
After the admission control on CBR service users, Atoll performs a noise rise scheduling, followed by a radio resource control.
Then, it repeats the same steps on VBR service users first, and lastly on BE service users, in the order established during the
generation of the user distribution.
CBR Service Users
Let us focus on the three CBR service users mentioned in the example of the previous paragraph "HSUPA Part of the
Algorithm" on page 256. We assume that all of them have been admitted. Noise rise scheduling and radio resource control
are carried out on each user in order to determine the best HSUPA bearer that the user can obtain. Several CBR service users
can share the same HSUPA bearer. Then, Atoll calculates the HSUPA bearer consumption ( C in %) for each user and takes into
account this parameter when it determines the resources consumed by the user (i.e., the terminal power used, the number
of channel elements and the Iub backhaul throughput).
In the bearer allocation process shown below, the 3 CBR service users are represented by Mj, with j = 1 to 3.
8.
In the HSUPA coverage prediction, L T is calculated as follows:

UL
-)
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
259


Determination of the best HSUPA bearer B(Mj)
Allocation of the minimum throughput demand to

Mj
Calculation of C(B(Mj))
Sufficient Iub
backhaul throughput
to support the HSUPA
bearer?
No
Is there a lower HSUPA

bearer available?
Yes
No
Yes
Enough
channel elements
HSUPA bearer?
Downgrading to lower
HSUPA bearer
Mj is rejected
Yes
No

bearer available?
No
Yes
Mj is rejected
Pterm-HSUPA recalculation and interference update
No
Mj = M3?
Yes
Resource allocation for packet (HSPA

Variable Bit Rate) service users
Figure 4.12: HSUPA Bearer Allocation Process for CBR Service Users
VBR Service Users
Let us focus on the two VBR service users mentioned in the example of the previous paragraph "HSUPA Part of the Algorithm"
on page 256. We assume that all of them have been admitted. Noise rise scheduling and radio resource control are carried
out on each user in order to determine the best HSUPA bearer that the user can obtain.
In the bearer allocation process shown below, the 2 VBR service users are represented by Mj, with j = 1 to 2.
Determination of the best HSUPA bearer
Sufficient Iub
backhaul throughput
bearer?
No

bearer available?
Yes
No
Yes
Enough
channel elements
HSUPA bearer?
Downgrading to lower HSUPA

bearer
Mj is rejected
Yes
No

bearer available?
No
Yes
Mj is rejected
No
Mj = M2?
Yes
Resource allocation for packet (HSPA Best

Effort) service users
Figure 4.13: HSUPA Bearer Allocation Process for VBR Service Users
BE Service Users
Let us focus on the five BE service users mentioned in the example of the previous paragraph "HSUPA Part of the Algorithm"
on page 256. We assume that all of them have been admitted. Noise rise scheduling and radio resource control are carried
out on each user in order to determine the best HSUPA bearer that the user can obtain.
In the bearer allocation process shown below, the 5 BE service users are represented by Mj, with j = 1 to 5.
260

AT332_TRR_E0

Determination of the best HSUPA bearer
Sufficient Iub
backhaul throughput
bearer?
No

bearer available?
Yes
No
Yes
Downgrading to lower HSUPA

bearer
Mj is rejected
Enough
channel elements
HSUPA bearer?
Yes
No

bearer available?
No
Yes
Mj is rejected
No
Mj = M5?
Figure 4.14: HSUPA Bearer Allocation Process for BE Service Users
4.3.2.4.3
Noise Rise Scheduling

Determination of the Obtained HSUPA Bearer
The obtained HSUPA radio bearer is the bearer that the user obtains after noise rise scheduling and radio resource control.
CBR service users have the highest priority and are processed first. Therefore, after the admission control, the noise rise
scheduling algorithm attempts to evenly share the remaining cell load between the CBR service users admitted in admission
control; in terms of HSUPA, each user is allocated a right to produce interference. The remaining cell load factor on uplink
UL
( X HSPA CBR txi ic ) depends on the maximum load factor allowed on uplink and how much uplink load is produced by the
served R99 traffic. It can be expressed as follows:
UL
UL
UL
X HSPA CBR txi ic = X max txi ic X R99 txi ic

Then, Atoll evenly shares the remaining cell load factor between the CBR service users admitted during the previous step
( n HSPA CBR ).
UL
X HSPA CBR txi ic

UL
X user txi ic = -----------------------------------------------n HSPA CBR
Ec max
From this value, Atoll calculates the maximum E-DPDCH EcNt allowed ( ------
) for each CBR service user. For further
Nt E DPDCH
information on the calculation, see "Uplink Load Factor Due to One User" on page 279.
Ec- max
1
----- for the Without useful signal option
Nt E DPDCH = ------------------------------------------UL
F txi ic 1
---------------------------------UL
X user txi ic
UL
X user
Ec- max
----for the Total noise option
Nt E DPDCH = --------------UL
F
Then, it selects an HSUPA bearer. The allocation depends on the maximum E-DPDCH EcNt allowed and on UE and cell capabilities. Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. This is the HSUPA bearer ( Index HSUPABearer )
UL
TP P R LC Index HSUPABearer
with the highest potential throughput ( ----------------------------------------------------------------- ) where:
N Rtx Index HSUPABearer
Ec- req
Ec- max
---- ---- Nt E DPDCH Nt E DPDCH
And P term HSUPA P term
req
max
Ec req
When several HSUPA bearers are available, Atoll selects the one with the lowest ------
.
Nt E DPDCH
261

After the noise rise scheduling, Atoll carries out radio resource control, verifying if enough channel elements and Iub backhaul
throughput are available for the HSUPA bearer assigned to the user. For information on radio resource control, see "Radio
Resource Control" on page 265.
After processing all CBR service users, Atoll carries out noise rise scheduling and radio resource control on VBR service users.
During the noise rise scheduling, Atoll distributes the remaining cell load factor available after all CBR service users have been
served. It can be expressed as follows:
UL
UL
UL
UL
X HSPA VBR txi ic = X max txi ic X R99 txi ic X HSPA CBR txi ic
The remaining cell load factor is shared equally between the admitted VBR service users ( n HSPA VBR ).
UL
X HSPA VBR txi ic

UL
X user txi ic = -----------------------------------------------n HSPA VBR
max
Ec
) as explained above and selects an
Nt E DPDCH
HSUPA bearer for each VBR service user. After the noise rise scheduling, Atoll carries out radio resource control on VBR service
users. For information on radio resource control, see "Radio Resource Control" on page 265.
After processing VBR service users, Atoll carries out noise rise scheduling and radio resource control on BE service users.
During the noise rise scheduling, Atoll distributes the remaining cell load factor available after all CBR and VBR service users
have been served. It can be expressed as follows:
UL
UL
UL
UL
UL
X HSPA txi ic = X max txi ic X R99 txi ic X HSPA CBR txi ic X HSPA VBR txi ic
The remaining cell load factor is shared equally between the admitted BE service users ( n HSPA ).
UL
X HSPA txi ic
UL
X user txi ic = -----------------------------------n HSPA
Ec max
) as explained above and selects an
Nt E DPDCH
HSUPA bearer for each BE service user. After the noise rise scheduling, Atoll carries out radio resource control on BE service
users. For information on radio resource control, see "Radio Resource Control" on page 265.
Example: We have a cell with six BE service users, and neither CBR user nor VBR user. All BE service users have been admitted.
The remaining cell load factor equal to 0.6 is shared between the BE service users. Therefore, the UL load factor allotted to
each user is 0.1. Lets take the cell UL reuse factor equal to 1.5. Atoll calculates the maximum E-DPDCH EcNt allowed (the
Without useful signal option is selected).
max
Ec
We have: ------
= -11.5 dB
Nt E DPDCH
Here, the obtained HSUPA bearer is the index 5 HSUPA bearer. It provides a potential throughput of 128 kbps and requires
E-DPDCH EcNt of -13 dB (lower than -11.5 dB) and a terminal power lower than the maximum terminal power allowed.
.
HSUPA Bearers Index
Required Ec/Nt
Threshold (dB)
Nb of Retransmissions
Peak RLC Throughput

(kbps)
Potential Throughput
(kbps)
-21.7
32
16
-19
64
32
-16.1
128
64
-13.9
192
96
-13
256
128
-10.1
512
256
-8
768
384
-7
1024
512
AtollAtollNoise Rise Scheduling in Soft Handover

With HSUPA, uplink soft handover impacts the scheduling operation. While HSDPA sends data from one cell only, with HSUPA
all cells in the active set receive the transmission from the terminal. Therefore, all the cells are impacted by the transmission
in terms of noise rise.
262

AT332_TRR_E0
For each HSPA-capable cell of the active set tx k ic , Atoll calculates the maximum E-DPDCH EcNt allowed
max
Ec
( ------
tx ic ) as explained in "HSUPA Bearer Allocation Process" on page 259.
Nt E DPDCH k
For each cell of the active set tx k ic , Atoll calculates the maximum terminal power allowed to obtain an HSUPA radio bearer
max
( P term HSUPA tx k ic ).
max
Ec
max
UL
max
P term HSUPA tx k ic = min ------
tx ic L T N tot P term
Nt E DPDCH k
With
UL
UL
tx
intra
UL extra
ic + I tot
tx
UL

L T = ------------------------------------------------------------------------------------------------------------------- ( )
G Tx G term
tx
UL intra
UL extra
, I tot
UL
tx
As HSUPA bearer users in soft handover use the lowest granted noise rise, Atoll chooses the lowest of maximum terminal
power allowed for each cell of the active set tx k ic .
max
P term HSUPA = min
tx AS
k
max
P term HSUPA tx k ic
max
Once Atoll knows the selected maximum terminal power ( P term HSUPA ), it recalculates the maximum E-DPDCH EcNt allowed
Ec max
tx ic ) for each HSUPA-capable cell of the active set.
( ------
Nt E DPDCH k
max
P term HSUPA
Ec- max
---- tx ic = ---------------------------- Nt E DPDCH k
UL
L T N tot
max
Ec
Then, Atoll calculates the maximum E-DPDCH EcNt allowed ( ------
) after signal recombination of all HSUPA capable
Nt E DPDCH
cells of the active set 10.
For softer (1/2) and softer-softer (1/3) handovers, we have:
max
Ec-
UL
----= f rake efficiency
Nt E DPDCH
max
Ec-
---- tx ic
Nt E DPDCH k
txk ActiveSet
samesite
Ec max
For soft (2/2) and soft-soft (3/3) handovers, we have: ------
Nt E DPDCH =
txk
Ec- max
Max ---- Nt E DPDCH tx k ic

ActiveSet
For softer-soft handover (2/3), it depends on if the MRC option is selected (option available in Global parameters). If selected,
we have:
9.
In the HSUPA coverage prediction, L T is calculated as follows:

UL
-)
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
263

max
Ec-
---- Nt E DPDCH =
UL
Ec- max
Ec- max
-------- tx k ic
tx l ic
tx ,tx ActiveSet f rake efficiency
Nt
Nt
k l
E DPDCH
E DPDCH
tx samesite
tx
k
k
Max
tx othersite
l
max
Ec
Else, we have: ------
=
Nt E DPDCH
Ec- max
Max ---- tx ic
Nt E DPDCH k
txk ActiveSet
Then, Atoll selects an HSUPA bearer as previously explained in "HSUPA Bearer Allocation Process" on page 259. The allocation
depends on the maximum E-DPDCH EcNt allowed and on UE and cell capabilities. Atoll selects the best HSUPA bearer from
the HSUPA compatible bearers. This is the HSUPA bearer ( Index HSUPABearer ) with the highest potential throughput
UL
TP P R LC Index HSUPABearer
( ----------------------------------------------------------------- ) where:
N Rtx Index HSUPABearer
Ec- req
Ec max
---- ------
Nt E DPDCH Nt E DPDCH
Ec req
When several HSUPA bearers are available, Atoll selects the one with the lowest ------
.
Nt E DPDCH
10.
In HSUPA coverage predictions, we have the following:

max
Ec
UL
= f rake efficiency
For softer (1/2) and softer-softer (1/3) handovers: ------
Nt E DPDCH
max
Ec-
---- Nt E DPDCH tx k ic
txk ActiveSet
samesite
Ec max
=
For soft handover (2/2): ------
Nt E DPDCH
Ec- max
UL
Max ---- tx ic G macro diversity 2links
Nt E DPDCH k
txk ActiveSet
Ec max
=
For soft-soft handover (3/3): ------
Nt E DPDCH
tx k
Ec- max
UL
Max ---- tx ic G macro diversity 3links
Nt E DPDCH k
ActiveSet
For softer-soft handover (2/3), it depends on if the MRC option is selected (option available in Global parameters). If selected,
we have:
max
Ec
----=
Nt E DPDCH
UL
txk ,txl ActiveSet f rake efficiency
tx k samesite
tx k
Max
Ec max
Ec max
---- tx k ic ------
tx l ic
Nt E DPDCH
Nt E DPDCH
tx l othersite
UL
G macro diversity 2links
max
Ec
=
Else, we have: ------
Nt E DPDCH
264
Ec- max
UL
Max ---- Nt E DPDCH tx k ic G macro diversity 2links
txk ActiveSet

AT332_TRR_E0
Determination of the Requested HSUPA Bearer

The requested HSUPA radio bearer is selected from the HSUPA bearers compatible with the user equipment. Atoll determines
the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. The user is treated as if he is the
only user in the cell. Therefore, if we go on with the previous example, the maximum E-DPDCH EcNt allowed is equal to -1.8
dB and the requested HSUPA bearer is the index 7 HSUPA bearer. It requires E-DPDCH EcNt of -8 dB (lower than -1.8 dB) and
a terminal power lower than the maximum terminal power allowed.
4.3.2.4.4
Radio Resource Control

Atoll checks to see if enough channel elements are available and if the Iub backhaul throughput is sufficient for the HSUPA
bearer assigned to the user (taking into account the maximum number of channel elements defined for the site and the
maximum Iub backhaul throughput allowed on the site in the uplink). If not, Atoll allocates a lower HSUPA bearer
("downgrading") which needs fewer channel elements and consumes lower Iub backhaul throughput. If no channel elements
are available, the user is rejected. On the same hand, if the maximum Iub backhaul throughput allowed on the site in the
uplink is still exceeded even by using the lowest HSDPA bearer, the user is rejected.
4.3.2.5 Convergence Criteria

The convergence criteria are evaluated for each iteration, and can be written as follow:
DL
DL
DL
max
max
P tx ic k P tx ic k 1
N user ic k N user ic k 1
Stations
Stations
- 100
= max int ----------------------------------------------------------------------------------- 100 int --------------------------------------------------------------------------------------------DL

P tx ic k
N user ic k
UL
UL
UL
UL
max
max
I tot ic k I tot ic k 1
Stations
Stations
- 100 int --------------------------------------------------------------------------------------------- 100
UL = max int ----------------------------------------------------------------------------------UL
UL
I tot ic k
N user ic k

Atoll stops the algorithm if:

1st case: Between two successive iterations, UL and DL are lower than their respective thresholds (defined when creating
a simulation).
The simulation has reached convergence.
Example: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5. If
UL 5 and DL 5 between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th iteration. Convergence has
been reached.
2nd case: After 30 iterations, UL and/or DL are still higher than their respective thresholds and from the 30th iteration, UL
and/or DL do not decrease during the next 15 successive iterations.
The simulation has not reached convergence (specific divergence symbol).
Examples: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5.
1. After the 30th iteration, UL and/or DL equal 100 and do not decrease during the next 15 successive iterations: Atoll
stops the algorithm at the 46th iteration. Convergence has not been reached.
2. After the 30th iteration, UL and/or DL equal 80, they start decreasing slowly until the 40th iteration (without going
under the thresholds) and then, do not change during 15 successive iterations: Atoll stops the algorithm at the 56th
iteration without reaching convergence.
3rd case: After the last iteration.
If UL and/or DL are still strictly higher than their respective thresholds, the simulation has not reached convergence
(specific divergence symbol).
If UL and DL are lower than their respective thresholds, the simulation has reached convergence.
4.3.3 Results
4.3.3.1 R99 Related Results
This table contains some R99 specific simulation results provided in the Cells and Mobiles tabs of the simulation property
dialog box.
265

Name
Value
Unit
Description
Nb E1 T1 Ethernet
E1 T1 Ethernet
TPIub DL N I TP

RoundUp Max
E1 T1 Ethernet
TP Iub UL N I TP
None
Number of E1/T1/Ethernet links

required by the site
None



DL
P tot txi
DL
I intra txi ic
DL
SCH txi ic
P tot txi ic P
----------------------------F
ortho
BTS
ic
LT
txi
DL
1 F ortho BTS P b txi ic
DL
I extra ic
DL
P tot txj ic
txj j i
DL
P tot txj
DL
I inter carrier ic
RF ic ic adj
DL
ni
DL
DL
I tot ic
DL
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP n ic
Total effective interference at

terminal on carrier ic (after
unscrambling)
I tot ic + N 0

carrier ic
Pb

ic

carrier ic

DL
DL
Term
UL
UL
extra
I tot
txi
ic
ic
term
txi
ic
UL
P b ic
term
txj j i
UL
P b ic adj
UL
term
txj j
-----------------------------------
RF ic ic adj
UL
I tot txi ic
UL
N tot txi ic
UL
extra
I tot
a
DL
DL
intra
I tot
txi
N tot ic
UL
ic adj
txj
j
---------------------------------------------
UL
Tx
intra

UL
UL
txi ic +I inter carrier txi icW
tx
I tot txi ic + N 0


ic
None
Cell uplink load factor on carrier ic
UL
UL
I tot txi ic
---------------------------UL
N tot txi ic
UL
I tot txi ic
--------------------------------------------------------------------------------------UL intra
Tx
I tot
txi ic 1 F MUD term
UL
1
-------------------------UL
F txi ic
X txi ic
UL
F txi ic
E txi ic
266
None Cell uplink reuse factor on carrier ic
None
Cell uplink reuse efficiency factor on

carrier ic

AT332_TRR_E0
Name
Value
Unit
Description
None
Downlink load factor on carrier ic
Simulation result available per cell

DL
I extra ic
tch
DL
+ I inter carrier ic L T
--------------------------------------------------------------------------------- + 1 F ortho BTS
DL
P Tx txi ic
--------------------------------------------------------------------------------------------------------------------------------1 - + 1 F
---------
ortho
DL
CI req
DL
X txi ic
BTS
DL
Q req
DL
with CI req = --------DL
Gp
Simulation result available per mobile
DL
I tot ic
-----------------DL
N tot ic
DL
I tot ic
----------------------------DL
I intra txi ic
DL
F txi ic
dB
Noise rise on downlink
UL
dB
Noise rise on uplink
10 log 1 X txi ic
UL
10 log 1 X txi ic
NR txi ic
a.
DL
DL
NR txi ic
None Downlink reuse factor on a carrier ic
4.3.3.2 HSPA Related Results

At the end of the R99 part, the users can be:
Either connected and in this case, they obtain the requested R99 bearer,
Or rejected exactly for the same reasons as R99 users.
Only connected HSDPA and HSPA users are considered in the HSDPA part. At the end of the HSDPA part, BE service users can
be:
Either connected if they obtain an HSDPA bearer,

Or rejected if the maximum number of HSDPA bearer users per cell is exceeded,
Or delayed in case of lack of resources (HSDPA power, HS-SCCH power, HS-SCCH channels, OVSF codes).
VBR service users can be:

Or rejected for the following reasons: the maximum number of HSDPA bearer users per cell is exceeded, the lowest
HSDPA bearer the user can obtain does not provide a peak RLC throughput higher than the minimum throughput
demand, the HS-SCCH signal quality is not sufficient, there are no more OVSF codes available, the maximum Iub
backhaul throughput allowed on the site in the downlink is exceeded.
CBR service users can be:

Or rejected for the following reasons: the maximum number of HSDPA bearer users per cell is exceeded, the lowest
HSDPA bearer the user can obtain does not provide a peak RLC throughput higher than the minimum throughput
demand, the HS-SCCH signal quality is not sufficient, there are no more OVSF codes available, the maximum Iub
backhaul throughput allowed on the site in the downlink is exceeded.
In the HSUPA part, Atoll processes HSPA service users who are connected to an HSDPA bearer or were delayed in the previous
step. At the end, they can be:
4.3.3.2.1
Either connected if they obtain an HSUPA bearer,

Or rejected for the following reasons: the maximum number of HSUPA bearer users per cell is exceeded, the terminal
power required to obtain the lowest compatible HSUPA bearer exceeds the maximum terminal power, there are no
more channel elements available, the maximum Iub backhaul throughput allowed on the site in the uplink is
exceeded, the lowest compatible HSUPA bearer they can obtain does not provide a peak RLC throughput higher than
the minimum throughput demand (only for CBR and VBR service users).
Statistics Tab
In the Statistics tab, Atoll displays as results:
The number of rejected users.
267

The number of delayed users.

The number of R99 bearer users connected to a cell (result of the R99 part). This figure includes R99 users as well as
HSDPA and HSPA users since all of them request an R99 bearer.
The number of R99 bearer users per frequency band.

The number of R99 bearer users per activity status.
The downlink and uplink peak throughputs ( TP P D L and TP P U L ) generated by their connection to R99 bearers.
R99
R99
Only active users are considered.
R99
TP P D L =
R99
R99
TP P DL R99 Bearer and TP P U L =
Active
users
R99
TP P UL R99 Bearer
Active
users
R99
R99
TP P DL R99 Bearer is the downlink peak throughput of the user R99 radio bearer and TP P UL R99 Bearer is the uplink
peak throughput of the user R99 radio bearer.
The number of connected users with an HSDPA bearer (result of the HSDPA part) and the downlink peak RLC
throughput they generate. HSDPA and HSPA service users are considered since they all request an HSDPA bearer. On
DL
the other hand, only active users are taken into consideration in the downlink throughput calculation ( TP HSDPA ).
DL
TP HSDPA =
DL
TP P RLC
Active
users
DL
TP P RLC is the peak RLC throughput provided in the downlink.
The number of connected users with an HSUPA bearer (result of the HSUPA part). Only HSPA service users are
considered.
In addition, Atoll indicates the uplink peak RLC throughput generated by active users connected with an HSUPA bearer
UL
( TP HSUPA ):
UL
TP HSUPA =
UL
TP P RLC
Active
users
UL
TP P RLC is the peak RLC throughput provided in the uplink.
4.3.3.2.2Mobiles Tab
In the Mobiles tab, Atoll indicates for each user:
UL
DL
The uplink and downlink total requested throughputs in kbps (respectively, TP requested M b and TP requested M b )
For R99 users, the DL and UL total requested throughputs correspond to the DL and UL peak throughputs of the R99 bearer
associated to the service.
DL
R99
UL
R99
TP requested M b = TP P DL R99 Bearer

TP requested M b = TP P UL R99 Bearer
For HSDPA users, the uplink requested throughput corresponds to the peak throughput of ADPCH R99 radio bearer and the
downlink requested throughput is the sum of the ADPCH radio bearer peak throughput and the peak RLC throughput(s) that
the selected HSDPA radio bearer(s) can provide. Here, the user is treated as if he is the only user in the cell and then, Atoll
determines the HSDPA bearer the user would obtain by considering the entire HSDPA power available of the cell.
DL
R99
DL
R99
DL
TP requested M b = TP P DL ADPCH R99 Bearer + TP P RLC for single-carrier users

TP requested M b = TP P DL ADPCH R99 Bearer AnchorCell +
DL
TP P RLC c for dual-carrier users
c Serving Cells
UL
R99
TP requested M b = TP P UL ADPCH R99 Bearer

For HSPA users, the uplink requested throughput is equal to the sum of the ADPCH-EDPCCH radio bearer peak throughput and
the peak RLC throughput of the requested HSUPA radio bearer. The requested HSUPA radio bearer is selected from the HSUPA
bearers compatible with the user equipment. Here, the user is treated as if he is the only user in the cell and then, Atoll
determines the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. The downlink
268

AT332_TRR_E0
requested throughput is the sum of the ADPCH-EDPCCH radio bearer peak throughput and the peak RLC throughput(s) that
the requested HSDPA radio bearer(s) can provide. The requested HSDPA radio bearer is determined as explained in the
previous paragraph.
DL
R99
DL
R99
DL
TP requested M b = TP P DL ADPCH EDPCCH R99 Bearer + TP P RLC for single-carrier users

TP requested M b = TP P DL ADPCH EDPCCH R99 Bearer AnchorCell +
DL
TP P RLC c for dual-carrier users
c Serving cells
UL
TP requested M b
R99
TP P UL ADPCH
EDPCCH R99 Bearer +
UL
TP P RLC
UL
DL
The uplink and downlink total obtained throughputs in kbps (respectively, TP obtained M b and TP obtained M b )
For R99 service users, the obtained throughput is the same as the requested throughput if he is connected without being
downgraded. Otherwise, the obtained throughput is lower (it corresponds to the peak throughput of the selected R99 bearer).
If the user is rejected, the obtained throughput is zero.
In the downlink, HSDPA bearer users can be connected to a single cell or to two adjacent cells of the same transmitter when
the user has a DC-HSDPA-capable terminal and when the transmitter supports the multi-cell HSDPA mode.
For a single-carrier HSDPA service user connected to an HSDPA bearer, the downlink obtained throughput corresponds to the
instantaneous throughput; this is the sum of the A-DPCH radio bearer peak throughput and the peak RLC throughput provided
by the selected HSDPA radio bearer after scheduling and radio resource control. If the user is delayed (he is only connected
to an R99 radio bearer), downlink obtained throughput corresponds to the downlink peak throughput of the ADPCH radio
bearer. Finally, if the user is rejected either in the R99 part or in the HSDPA part (i.e., because the HSDPA scheduler is
saturated), the downlink obtained throughput is zero.
For a dual-carrier HSDPA service user connected to two HSDPA bearers, the downlink obtained throughput corresponds to
the instantaneous throughput; this is the sum of the peak throughput provided by the A-DPCH radio bearer in the anchor cell
and the peak RLC throughputs provided by the selected HSDPA radio bearers after scheduling and radio resource control. If
the user is connected to one cell and delayed in the other cell, the downlink obtained throughput is the sum of the peak
throughput provided by the A-DPCH radio bearer in the anchor cell and the peak RLC throughput provided by the selected
HSDPA radio bearer after scheduling and radio resource control. If the user is delayed in the two cells (he is only connected
to an R99 radio bearer in the anchor cell), the downlink obtained throughput corresponds to the downlink peak throughput
of the ADPCH radio bearer in the anchor cell. Finally, if the user is rejected either in the R99 part or in the HSDPA part (i.e.,
because the HSDPA scheduler is saturated), the downlink obtained throughput is zero.
In the uplink, HSDPA service users can only have a single-carrier connection. When the user is either connected or delayed,
the uplink obtained throughput corresponds to the uplink peak throughput of the ADPCH radio bearer. If the user is rejected
either in the R99 part or in the HSDPA part (i.e., because the HSDPA scheduler is saturated), the uplink obtained throughput
is zero.
For single-carrier HSPA VBR and BE service users, on downlink, if the user is connected to an HSDPA bearer, the downlink
obtained throughput corresponds to the instantaneous throughput. The instantaneous throughput is the sum of the ADPCHEDPCCH radio bearer peak throughput and the peak RLC throughput provided by the selected HSDPA radio bearer after
scheduling and radio resource control. If the user is delayed, the downlink obtained throughput corresponds to the downlink
peak throughput of ADPCH-EDPCCH radio bearer. If the user is rejected, the downlink obtained throughput is "0".
For dual-carrier HSPA VBR and BE service users connected to two HSDPA bearers, the downlink obtained throughput
corresponds to the instantaneous throughput; this is the sum of the peak throughput provided by the ADPCH-EDPCCH radio
bearer in the anchor cell and the peak RLC throughputs provided by the selected HSDPA radio bearers after scheduling and
radio resource control. If the user is connected to one cell and delayed in the other cell, the downlink obtained throughput is
the sum of the peak throughput provided by the ADPCH-EDPCCH radio bearer in the anchor cell and the peak RLC throughput
provided by the selected HSDPA radio bearer after scheduling and radio resource control. If the user is delayed in the two cells
(he is only connected to an R99 radio bearer in the anchor cell), the downlink obtained throughput corresponds to the
downlink peak throughput of the ADPCH-EDPCCH radio bearer in the anchor cell. Finally, if the user is rejected, the downlink
obtained throughput is zero.
In uplink, HSPA VBR and BE service users can only have a single-carrier connection. When the user is connected to an HSUPA
bearer, the uplink obtained throughput is the sum of the ADPCH-EDPCCH radio bearer peak throughput and the peak RLC
throughput provided by the selected HSUPA radio bearer after noise rise scheduling. If the user is rejected, the uplink obtained
throughput is zero.
For a connected HSPA CBR service user, the uplink and downlink total obtained throughputs are the sum of the ADPCHEDPCCH radio bearer peak throughput and the minimum throughput demand defined for the service. If the user is rejected,
the uplink and downlink total obtained throughputs are "0".
The mobile total power ( P term )

UL
P term = P term R99 f act EDPCCH + P term HSUPA for HSPA VBR and BE service users.
269

UL
P term = P term R99 f act EDPCCH + P term HSUPA C HSDPABearer for HSPA CBR service users.
UL
For HSPA CBR service users, f act EDPCCH = 0.1 .
And
P term = P term R99 for R99 and HSDPA users.
DL
The HSDPA application throughput in kbps ( TP A M b )
This is the net HSDPA throughput without coding (redundancy, overhead, addressing, etc.).
DL
TP P RLC c 1 BLER HSDPA
DL
c Serving cells
TP A M b = ------------------------------------------------------------------------------------------------------- f TP Scaling TP Offset
TTI
Where:
DL
TP P RLC is the peak RLC throughput provided to the user by the selected HSDPA radio bearer after scheduling and radio
resource control.
BLER HSDPA is read in the quality graph defined for the triplet reception equipment-selected bearer-mobility (HSDPA Quality
Graphs tab in the Reception equipment properties). This graph describes the variation of BLER as a function of the measured
quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll calculates the corresponding BLER.
TP Offset and f TP Scaling represent the scaling factor between the application throughput and the RLC (Radio Link Control)
throughput and the throughput offset respectively. These two parameters model the header information and other
supplementary data that does not appear at the application level. They are defined in the service properties.
TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal user
equipment category properties.
The number of OVSF codes
This is the number of 512-bit length OVSF codes consumed by the user.
The required HSDPA power in dBm ( P HSDPA required )
It corresponds to the HSDPA power required to provide the HSDPA bearer user with the downlink requested throughput. The
downlink requested throughput is the throughput the user would obtain if he was the only user in the cell. In this case, Atoll
determines the HSDPA bearer the user would obtain by considering the entire HSDPA power available of the cell.
P HSDPA required = P HS PDSCH used + n HS SCCH P HS SCCH
P HS PDSCH used is the HS-PDSCH power required to obtain the selected HSDPA bearer (in dBm). If the HSDPA bearer
allocated to the user is the best one, P HS PDSCH used corresponds to the available HS-PDSCH power of the cell. On the other
hand, if the HSDPA bearer has been downgraded in order to be compliant with cell and UE capabilities or for another reason,
P HS PDSCH used will be lower than the available HS-PDSCH power of the cell.
The served HSDPA power in dBm ( P HSDPA served )
This is the HSDPA power required to provide the HSDPA bearer user with the downlink obtained throughput. The downlink
obtained rate is the throughput experienced by the user after scheduling and radio resource control.
P HSDPA served = P HS PDSCH used + n HS SCCH P HS SCCH for HSDPA users, HSPA BE and VBR service users.
And
P HSDPA served = P HS PDSCH used C HSDPABearer for HSPA CBR service users
Where
P HS PDSCH used is the HS-PDSCH power required to obtain the selected HSDPA bearer.
The No. of HSUPA Retransmissions (Required)
The maximum number of retransmissions in order to have the requested HSUPA radio bearer with a given BLER.
270

AT332_TRR_E0
The No. of HSUPA Retransmissions (Obtained)
The maximum number of retransmissions in order to have the obtained HSUPA radio bearer with a given BLER.
UL
The HSUPA application throughput in kbps ( TP A M b )
This is the net HSUPA throughput without coding (redundancy, overhead, addressing, etc.).
UL
TP P RLC M b 1 BLER HSUPA f TP Scaling TP Offset

UL
TP A M b = ---------------------------------------------------------------------------------------------------------------------------------------N Rtx
Where:
UL
TP P RLC is the peak RLC throughput provided by the selected HSUPA radio bearer after noise rise scheduling.
BLER HSUPA is the residual BLER after N Rtx retransmissions. It is read in the quality graph defined for the quartet reception
equipment-selected bearer-number of retransmissions-mobility (HSUPA Quality Graphs tab in the Reception equipment
properties). This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt). Knowing the EDPDCH Ec/Nt, Atoll calculates the corresponding BLER.
TP Offset and f TP Scaling respectively represent the scaling factor between the application throughput and the RLC (Radio Link
Control) throughput and the throughput offset. These two parameters model the header information and other
N Rtx is the maximum number of retransmissions for the obtained HSUPA bearer. This figure is read in the HSUPA Bearer
Selection table.
The following columns appear if, when creating the simulation, you select "Detailed information about mobiles":
The uplink and downlink requested peak RLC throughputs (kbps)
Downlink and uplink requested peak RLC throughputs are not calculated for R99 users.
For HSDPA users, the uplink peak RLC throughput is not calculated and the downlink requested peak RLC throughput is the
throughput that the selected HSDPA radio bearer(s) can provide. Here, the user is treated as if he is the only user in the cell
and then, Atoll determines the HSDPA bearer he would obtain by considering the entire HSDPA power available of the cell.
For HSPA users, the requested uplink peak RLC throughput is the throughput of the requested HSUPA radio bearer. The
requested HSUPA radio bearer is selected from the HSUPA bearers compatible with the user equipment. Here, the user is
treated as if he is the only user in the cell and then, Atoll determines the HSUPA bearer the user would obtain by considering
the entire remaining load of the cell. If the user is connected to one or two HSDPA bearers in the downlink, the downlink
requested peak RLC throughput is the throughput that the requested HSDPA radio bearer(s) can provide. The requested
HSDPA radio bearer is determined as explained in the previous paragraph.
The uplink and downlink obtained peak RLC throughput (kbps)
Downlink and uplink obtained peak RLC throughputs are not calculated for R99 users.
For HSDPA users connected to one or two HSDPA bearers, the uplink obtained peak RLC throughput is not calculated, and the
downlink obtained peak RLC throughput is the throughput provided by the selected HSDPA radio bearer(s) after scheduling
and radio resource control.
For connected HSPA BE and VBR service users, on uplink, if the user is connected to an HSUPA bearer, the obtained uplink
peak RLC throughput is the throughput provided by the selected HSUPA radio bearer after noise rise scheduling. On downlink,
if the user is connected to one or two HSDPA bearers, the downlink obtained peak RLC throughput is the throughput provided
by the selected HSDPA radio bearer(s) after scheduling and radio resource control.
For a connected HSPA CBR service user, the uplink and downlink obtained peak RLC throughputs are the uplink and downlink
minimum throughput demands defined for the service.
4.3.3.2.3
Cells Tab
In the Cells tab, Atoll gives:
The available HSDPA power in the cell, c, in dBm ( P HSDPA c ):
This is:
Either a fixed value in case of a static HSDPA power allocation strategy,

Or a simulation result when the option "HSDPA Power Dynamic Allocation" is selected. We have:
P HSDPA c = P max c P Headroom c P tx R99 c P HSUPA c
271

with P tx R99 c = P pilot c + P SCH c + P OtherCCH c +
tch used for

R99 users
P tch c +
DL
P tch c f act ADPCH
tch used for

HSPA users
The transmitted HSDPA power in the cell, c, in dBm ( P tx H SDPA c ):
It corresponds to the HSDPA power used to serve HSDPA bearer users.

P tx H SDPA c =
P HSDPA M b served
Mb c
The number of HSDPA users in the cell
They are the connected and delayed HSDPA bearer users. HSDPA and HSPA users are considered since they all request an
HSDPA bearer. DC-HSDPA users are accounted for once in each serving cell.
The number of simultaneous HSDPA users in the cell ( n M )

b
It corresponds to the number of connected HSDPA bearer users that the cell supports at a time, i.e. within one transmission
time interval. All these users are connected to the cell at the end of the HSDPA part of the simulation; they have a connection
with the R99 bearer and an HSDPA bearer. DC-HSDPA users are accounted for once in each serving cell.
DL
The instantaneous HSDPA throughput in the cell, c, in kbps ( TP Inst c )
This is the number of kilobits per second that the cell supports on downlink to provide simultaneous connected HSDPA bearer
users with an HSDPA bearer. We will differentiate single-carrier users (Ms) from DC-HSDPA users (Md-HSDPA stands for HSDPA
BE and VBR users, and Md-HSPA refers to HSPA BE and VBR service users).
DL
R99
DL
TP requested M b = TP P DL ADPCH R99 Bearer + TP P RLC
DL
TP obtained M s +
M c
s
R99
DL
TP P DL R99 Bearer + TP P RLC M d HSDPA
M d HSDPA c
c is the anchor cell
DL
TP P RLC M d HSDPA +
M d HSDPA c
c is the secondary cell
DL
TP Inst cell =
R99
DL
TP P DL R99 Bearer + TP P RLC M d HSPA +
M d HSPA c
c is the anchor cell
DL
M d HSPA c
TP P RLC M d HSPA
c is the secondary cell

DL
TP P RLC is the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource
control.
R99
TP P DL R99 Bearer is the peak throughput of the ADPCH radio bearer if the user is an HSDPA user. For HSPA users, it
corresponds to the ADPCH-EDPCCH radio bearer.
DL
DL
The instantaneous HSDPA Effective MAC Throughput in the cell, c, in kbps ( TPE M AC c )
TP E M AC c =
M c
b
S block M b
--------------------------------------T TTI TTI M b
Where,
S block M b is the transport block size (in kbits) of the HSDPA bearer selected by the user; it is defined for each HSDPA bearer
in the HSDPA Radio Bearers table.
TTI M b is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal user
272

AT332_TRR_E0
3
T TTI is the TTI duration, i.e. 2 10 s (2000 TTI in one second). This value is specified by the 3GPP.
DL
The average instantaneous HSDPA throughput in the cell, c, in kbps ( TP Av Inst c )

DL
TP Inst c
DL
TP Av Inst c = -------------------nM
b
DL
The HSDPA application throughput in the cell, c, in kbps ( TP A c )

DL
Either TP A c =
DL
Mb c
TP P RLC M b 1 BLER HSDPA f TP Scaling TP Offset

-------------------------------------------------------------------------------------------------------------------------------------- if the scheduling algorithm is Round Robin or
TTI
Proportional Fair,
DL
TP P RLC M b maxC I 1 BLER HSDPA f TP Scaling TP Offset

DL
- if the scheduling algorithm is Max C/I.
Or TPA c = -------------------------------------------------------------------------------------------------------------------------------------------------------------TTI
M b maxC I is the user with the highest C I in the cell.
DL
TP P RLC is the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource
control.
quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll calculates the corresponding BLER.
f TP Scaling and TP Offset respectively represent the scaling factor between the application throughput and the RLC (Radio Link
The minimum HSDPA RLC peak throughput in kbps (
DL
min TP P RLC M b )
M b cell
It corresponds to the lowest of RLC peak throughputs obtained by HSDPA bearer users connected to the cell.
The maximum HSDPA RLC peak throughput in kbps (
DL
max TP P RLC M b )
M cell
b
It corresponds to the highest of RLC peak throughputs obtained by HSDPA bearer users connected to the cell.
The number of HSUPA users in the cell ( n M ):

c
They are the HSUPA bearer users connected to the cell.
UL
The HSUPA application throughput in the cell, c, in kbps ( TP A c )
UL
TP A c =
UL
TP A M b
Mb c
UL
The uplink cell load factor due to HSUPA traffic ( X HSUPA c ):

UL
I tot c HSUPA
UL
X HSUPA c = --------------------------------UL
N tot c
Where
UL
I tot c HSUPA is the total interference at transmitter received from HSUPA bearer users.
4.3.3.2.4
Sites Tab
In the Sites tab, Atoll displays:
DL
The instantaneous HSDPA throughput carried by the site in kbps ( TP Inst site )
273

DL
TP Inst site =
DL
TP Inst c
c site
DL
The instantaneous HSDPA Effective MAC Throughput carried by the site in kbps ( T MAC site in kbps)
DL
TP E M AC site =
DL
TP E M AC c
c site
UL
UL
The HSUPA throughput carried by the site in kbps ( TP site )
TP site =
UL
TP obtained M c
M c site
4.3.4 Appendices
4.3.4.1 Admission Control in the R99 Part
During admission control in the R99 part of the simulation, Atoll calculates the uplink load factor of a considered cell assuming
the mobile concerned is connected to it. Here, activity status assigned to users is not taken into account. So even if the mobile
is not active on UL, it can be rejected due to cell load saturation. To calculate the cell UL load factor, either Atoll takes into
account the mobile power determined during power control if mobile was connected in previous iteration, or it estimates a
load rise due to the mobile and adds it to the current load. The load rise ( X
X
UL
UL
) is calculated as follows:
1
= --------------------------------------------W
1 + -----------------------------------UL
UL
Q req R nominal
4.3.4.2 Resources Management

4.3.4.2.1
OVSF Codes Management

OVSF codes are managed in the downlink during the simulation since this resource is downlink limited only. Atoll checks the
availability of this resource during the simulation, first in the R99 part and then in the HSDPA part. It determines the number
of codes that will be consumed by each cell.
OVSF codes form a binary tree. Codes of longer lengths are generated from codes of a shorter length. Length-k OVSF codes
are generated from length-k/2 OVSF codes. Therefore, if one channel needs 1 length-k/2 OVSF code, it is equivalent to use 2
length-k OVSF codes, or 4 length-2k OVSF codes and so on.
512 512-bit-length codes per cell are available in UMTS HSPA projects.
In the R99 part, during the resource control, Atoll determines the number of 512 bit-length codes that will be consumed for
each cell.
If the cell supports HSPA, Atoll allocates codes for the DL channels used for HSUPA:
A 128 bit-length code for the E-HICH and E-RGCH channels (i.e. four 512 bit-length OVSF codes), for each cell.
Therefore, Atoll will take four 512-bit-length codes,
A 256 bit-length code for the E-AGCH channel (i.e. two 512 bit-length OVSF codes), for each cell. Therefore, Atoll will
take two 512-bit-length codes,
If the cell supports HSDPA, Atoll reserves for potential HSDPA bearer users:
HS PDSCH Min
The minimum number of HS-PDSCH codes defined for the cell, N Codes
. They are 16-bit length OVSF codes
HS PDSCH Min
(i.e. thirty-two 512 bit-length OVSF codes). Therefore, Atoll will take 32 N Codes
512-bit-length codes,
A 128 bit-length code per HS-SCCH channel (i.e. four 512 bit-length OVSF codes), for each cell. Therefore, Atoll will
take 4 n HS SCCH 512-bit-length codes,
Then, it allocates to the cell OVSF codes to support R99 bearers required by users:
A 256 bit-length code per common channel (i.e. two 512 bit-length OVSF codes), for each cell. Therefore, Atoll will
Overhead
take 2 N Codes
512-bit-length codes,
A code per cell-receiver link, for TCH (traffic channels). The length of code to be allocated, Code_Length, depends on
the user activity. We have:
DL
Either Code_Length = F spreading Active user when the user is active,
274

AT332_TRR_E0
DL
Or Code_Length = F spreading Inactive user if the user is inactive.

TCH
The number of 512 bit-length OVSF codes needed N Codes is calculated from the length of the code to be allocated as follows:
TCH
512
N Codes = ------------------------------Code_Length
Figure 4.15: OVSF Code Tree Indices (Not OVSF Code Numbers)
The OVSF code allocation follows the Buddy algorithm, which guarantees that:
If a k-length OVSF code is used, all of its children with lengths 2k, 4k, , cannot be used as they will not be orthogonal.
If a k-length OVSF code is used, all of its ancestors with lengths k/2, k/4, , cannot be used as they will not be
orthogonal.
Example: We consider a user with a service requiring the UDD64 R99 radio bearer. This user is active on DL while connected
to a cell (which does not support HSDPA). The spreading factor for active users has been set to 64 and site equipment requires
four overhead downlink channel elements per cell. Atoll will consume four 256 bit-length OVSF codes for common channels
(i.e. eight 512 bit-length OVSF codes) and a 64 bit-length OVSF code for traffic channels (i.e. eight additional 512 bit-length
OVSF codes).
In the R99 part, the OVSF code allocation follows the mobile connection order (mobile
order in the Mobiles tab).
In DC-HSDPA, A-DPCH is only transmitted in the anchor carrier. Therefore, a DCHSDPA user requires R99 resources in the best serving cell only and consumes the
same amount of R99 resources as a single-cell HSDPA user.
The OVSF code and channel element management is differently dealt with in case of
softer handover. Atoll allocates OVSF codes for each cell-mobile link while it
globally assigns channel elements to a site.
In the HSDPA part, HSDPA and HSPA users are assigned an HSDPA bearer (Fast link adaptation).
Therefore, Atoll allocates to the cell:
16-bit length OVSF codes per cell-receiver, for HS-PDSCH. This figure depends on the HSDPA bearer assigned to the
user and on the type of service.
HS PDSCH
For HSDPA users, HSPA VBR and BE service users, Atoll needs 32 N Codes
HS PDSCH
connected to the cell. N Codes
is the number of HS-PDSCH channels required by the HSDPA bearer.

HS PDSCH
For HSPA CBR service users, Atoll needs 32 N Codes

HS PDSCH
to the cell. N Codes
512-bit-length codes for each user
C HSDPABearer 512-bit-length codes for each user connected
is the number of HS-PDSCH channels required by the HSDPA bearer.
DC-HSDPA users have two HSDPA bearers, one for each serving cell. Therefore, one DC-HSDPA user consumes OVSF
codes in both cells.
275

When HSDPA bearer users (at least one) are connected to the cell, Atoll gives the cell
HS PDSCH Min
back the minimum number of OVSF codes reserved for HS-PDSCH ( N Codes
). On
the other hand, if no HSDPA bearer user is connected, Atoll still keeps these codes and
the codes for HS-SCCH too. This is the same with HSUPA bearer users. Even if no HSUPA
bearer user is connected to the cell, Atoll still keeps the codes for E-HICH, E-RGCH and EAGCH channels.
4.3.4.2.2
Channel Elements Management

Channel elements are controlled in the R99 and the HSUPA parts of the simulation. Atoll checks the availability of this resource
in the uplink and downlink.
In the R99 part, during the resource control, Atoll determines the number of channel elements required by each site for R99
bearers in the uplink and downlink. Then, in the HSUPA part, Atoll carries out another resource control after allocating HSUPA
bearers. It takes into account the channel elements consumed by HSUPA bearer users in the uplink and recalculates the
number of channel elements required by each site in the uplink.
In the uplink, Atoll consumes N CE UL j channel elements for each cell j on a site NI. This figure includes:
Channel elements for R99 bearers:

Overhead
N CE UL
R99 T CH
N CE UL
channel elements for control channels,

per cell-receiver link, for R99 TCH (traffic channels).
Channel elements for HSUPA bearers:

HSUPA
per cell-receiver link, for packet (HSPA - BE) and packet (HSPA - VBR) service users.
HSUPA
C HSUPABearer per cell-receiver link, for CBR service users.
N CE
N CE
Therefore, the number of channel elements required in the uplink at the site level, N CE UL N I , is:
N CE UL N I =
NCE UL j
j NI
In the downlink, Atoll consumes N CE DL j channel elements for each cell j on a site NI. This figure includes:
Channel elements for R99 bearers

Overhead
N CE DL
N CE DL
R99 T CH
channel elements for control channels (Pilot channel, Synchronisation channel, common channels),
Therefore, the number of channel elements required in the downlink at the site level, N CE DL N I , is:
N CE DL N I =
NCE DL j
j NI
4.3.4.2.3
In DC-HSDPA, A-DPCH is only transmitted on the anchor carrier. Therefore, a DCHSDPA user requires R99 resources in the best serving cell only and consumes the
same amount of R99 resources as a single-cell HSDPA user.
In case of softer handover (the mobile has several links with co-site cells), Atoll
allocates channel elements for the best serving cell-mobile link only.
Iub Backhaul Throughput

The Iub backhaul throughput is controlled in the R99, the HSDPA and the HSUPA parts of the simulation. Atoll checks the
availability of this resource in the uplink and downlink.
In the R99 part, during the resource control, Atoll determines the Iub throughput required by each site for R99 bearers in the
uplink and downlink. Then, in the HSDPA part, Atoll performs a resource control in the downlink after allocating HSDPA
bearers. It takes into account the Iub backhaul throughput consumed by HSDPA bearer users in the downlink and recalculates
the Iub backhaul throughput required by each site in the downlink. Finally, in the HSUPA part, Atoll carries out a resource
control in the uplink after allocating HSUPA bearers. It takes into account the Iub backhaul throughput consumed by HSUPA
bearer users in the uplink and updates the Iub backhaul throughput required by each site in the uplink.
In the uplink, the Iub backhaul throughput consumed by each cell j on a site NI, TP Iub UL j , includes:
276

AT332_TRR_E0
The Iub backhaul throughput required for R99 bearers:
R99 T CH
TP Iub UL per cell-receiver link, for R99 TCH (traffic channels).
The Iub backhaul throughput required for HSUPA bearers:

HSUPA
TP Iub
per cell-receiver link, for HSPA BE and VBR service users.
HSUPA
TP Iub
C HSUPABearer per cell-receiver link, for HSPA CBR service users.
Therefore, the Iub backhaul throughput required on uplink at the site level, TP Iub UL N I , is:
TP Iub UL N I =
TPIub UL j
j NI
In the downlink, the Iub backhaul throughput consumed by each cell j on a site NI, TP Iub DL j , includes:
The Iub backhaul throughput required for R99 bearers:

Overhead
TP Iub DL
R99 T CH
TP Iub DL
for R99 control channels (Pilot channel, Synchronisation channel, common channels).
The Iub backhaul throughput required for HSDPA bearers:
TP Iub
HSDPA
per cell-receiver link, for HSDPA, HSPA BE and VBR service users.
HSDPA
TP Iub
C HSDPABearer per cell-receiver link, for HSPA CBR service users.
HSDPA
With TP Iub
DL
HSDPA
= TP P RLC + Overhead Iub
DL
TP P RLC
Therefore, the Iub backhaul throughput required on downlink at the site level, TP Iub DL N I , is:
TP Iub DL N I =
TPIub DL j
j NI
In DC-HSDPA, A-DPCH is only transmitted on the anchor carrier. Therefore, a DCHSDPA user requires R99 resources in the best serving cell only and consumes the
same amount of R99 resources as a single-cell HSDPA user. On the other hand, the
DC-HSDPA user has two HSDPA bearers (one for each serving cell) and consumes
HSDPA resources in both cells.
In case of softer handover (the mobile has several links with co-site cells), Iub
backhaul throughput is consumed by the best serving cell-mobile link only.
4.3.4.3 Downlink Load Factor Calculation

Atoll calculates a downlink load factor for each cell (available in the Cells tab of any simulation result) and each connected
mobile (available in the Mobiles tab of any given simulation result).
4.3.4.3.1
Downlink Load Factor per Cell

Approach for downlink load factor evaluation is highly inspired by the downlink load factor defined in the book WCDMA for
UMTS by Harry Holma and Antti Toskala.
DL
Q req
- be the required quality.
Let CI req = --------DL
Gp
DL
DL
G p and Q req are the processing gain on downlink and the Eb/Nt target on downlink respectively.
In case of soft-handoff, required quality is limited to the effective contribution of the transmitter.
DL
P tx c = P pilot c + P SCH c + P otherCCH c +
Ptch c
tch
DL
P tx c
ortho
P CCH c
nonOrtho
P CCH
c
Ptch c
tch
where
ortho
P CCH c = P pilot c + P otherCCH c
277

nonOrtho
P CCH
c = P SCH c
At mobile level, we have a required power, Ptch:

term
P tch c = CI req I extra c + I inter carrier c + I inter techno log y c + I intra c + N 0
LT r
DL
With r = 1 when the user is active on the downlink and r = r c when the user is inactive. In case of an HSDPA bearer user,
DL
r = f act ADPCH .
P tch c = CI req
I extra c + I inter carrier c + I inter techno log y c

DL
nonOrtho
nonOrtho
P tx c P CCH
c P tch c P CCH
c
- + ------------------------------ + N term
+ 1 F ortho BTS ------------------------------------------------------------------------------0
LT
LT
L r
T
DL
I extra c + I inter carrier c + I inter techno log y c L T r + 1 F ortho BTS P tx c r
+
nonOrtho
term
F ortho BTS P CCH
c r + N0 LT r
P tch ic = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------1 - 1 F
------------------+ ortho BTS
CIreq r
I intra c is the total power received at the receiver from the cell with which it is connected.
I extra c is the total power received at the receiver from other cells.
I inter carrier c is the inter-carrier interference received at the terminal.
I inter techno log y c is the inter-technology interference received at the terminal from an external transmitter.
We have:
ortho
nonOrtho
P CCH c + P CCH
DL
P tx c
I extra c + I inter carrier c + I inter techno log y c L T r
DL
nonOrtho
term
1
F
+
r
+
F
r
+
N
ortho
BTS
tx
ortho
BTS
CCH
0
T
+
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 - + 1 F
------------------ortho BTS
tch
CI req r

------------------------------------------------------------------------------------------------------------------------------------------- P DL
tx c
DL
DL
P tx c
P tx c = P ortho c + P nonOrtho c + ---------------------------------------------------------------------------------------------------------------------------------------------------------------- +
CCH
CCH
1
-------------------- + 1 F ortho BTS
tch
CI req r
DL
1 F ortho BTS P tx c r
--------------------------------------------------------------------------- +
1
tch -------------------- + 1 F ortho BTS
CI req r
nonOrtho
term
F ortho BTS P CCH

c r + N0 LT r
----------------------------------------------------------------------------------------------------------------1
-------------------- + 1 F ortho BTS
tch
CI req r

------------------------------------------------------------------------------------------------------------------------------------------ + 1 F ortho BTS r
DL
P tx c
DL
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- P DL
P tx c
tx ic
1 - + 1 F
------------------
ortho
BTS
tch
CI req r
ortho
nonOrtho
= P CCH c + P CCH
278
c +
nonOrtho
term
F ortho BTS P CCH

c r + N0 LT r
----------------------------------------------------------------------------------------------------------------1
-------------------- + 1 F ortho BTS
tch
CI req r

AT332_TRR_E0
nonOrtho
term
F ortho BTS P CCH

c r + N0 LT r
----------------------------------------------------------------------------------------------------------------1
-------------------- + 1 F ortho BTS
tch
CIreq r
DL
P tx c = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------I extra c + I inter carrier c + I inter techno log y c L T r
------------------------------------------------------------------------------------------------------------------------------------------ + 1 F ortho BTS r
DL
tx
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1
1 - + 1 F
------------------
ortho BTS
tch
r
CI
req
ortho
nonOrtho
P CCH c + P CCH
c +
Therefore, the downlink load factor can be expressed as:
DL

------------------------------------------------------------------------------------------------------------------------------------------- + 1 F ortho BTS r
DL
P tx c
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------=
1 - + 1 F
------------------
tch
ortho
BTS
CI req r
The downlink load factor represents the signal degradation in relation to the reference interference (thermal noise plus
synchronisation channel power).
4.3.4.3.2
Downlink Load Factor per Mobile

Atoll evaluates the downlink load factor for any connected mobile as follows:
X
DL
DL
I tot c
= ---------------DL
N tot c
4.3.4.4 Uplink Load Factor Due to One User

UL
This part details how Atoll calculates the contribution of one user to the UL load factor ( X k ).
UL
In this calculation, we assume that the cell UL reuse factor ( F txi ic ) is constant.
The result depends on the option used to calculate Nt (Without useful signal or Total noise that you may select in Global
parameters).
Without Useful Signal Option
UL
P b k req
W - ---------------------------------------------------------------------------------------------------------------UL
Q req k = ------------------------
R99
UL
tx
TP P UL k I intra P b k req + I extra + I inter carrier + N 0
UL
P b k req
W
UL
- -----------------------------------------------------------------------Q req k = ------------------------UL
R99
UL
tx
TP P UL k I intra F P b k req + N 0
R99
R99
TP P UL k
TP P UL k
UL
UL
UL
tx
- = Q UL
P b k req 1 + Qreq k ------------------------+ N0
req k -------------------------- I intra F
W
W
R99
UL
P b k req
R99
TP P UL k
TP P UL k
UL
tx
- I intra F UL Q UL
Q req k ------------------------req k -------------------------- N 0
W
W
= ----------------------------------------------------------------------------------- + ----------------------------------------------------------------R99
R99
TP P UL k
TP P UL k
UL
UL
1 + Q req k -------------------------1 + Q req k -------------------------W
W
R99
req
TP P UL k
Ec
UL
We note ------ k
= Q req k ------------------------ Nt E DPDCH
W
UL
tx
I intra F
N0
UL
P b k req = ------------------------------------------------------ + -----------------------------------------------------

1
1
- + 1 --------------------------------------- + 1
--------------------------------------req
req
Ec
Ec
---- ----
Nt- k
Nt- k
E DPDCH
E DPDCH
As I intra =
Pb
UL
k req , we have:
279

I intra = I intra F
UL
tx
- + N 0 ------------------------------------------------------ -----------------------------------------------------
N0

tx
1
- + 1
--------------------------------------Ec- req
----
Nt k
E DPDCH
1
- + 1
--------------------------------------Ec- req
----
Nt k
E DPDCH
-----------------------------------------------------
1
- + 1
--------------------------------------req
Ec
----
Nt- k
E DPDCH
= -------------------------------------------------------------------------------------UL
1
1F
------------------------------------------------------
K
1
--------------------------------------+
1
req
Ec
------
Nt k
E DPDCH
K
I intra
UL
tx
N0 F
I intra = -------------------------------------------------------------------------------------1
----------------------------------------------------------------------------1
UL
1
F
-----------------------------------------------------
K
1
- + 1
--------------------------------------req
Ec
----
Nt- k
E DPDCH
UL
UL
I intra + I extra + I inter carrier
I intra F
1
= ------------------------------------------------------------------------------= --------------------------------------= ---------------------------------UL
tx
tx
tx
I intra + I extra + I inter carrier + N 0
I intra F + N 0
N0
1 + ------------------------UL
I intra F
Therefore, we have:
X
UL
= F
UL
-----------------------------------------------------
1
- + 1
--------------------------------------req
Ec
------
Nt k
E DPDCH
So, we can conclude that the contribution of one user to the UL load is defined as:
UL
X k = F
UL
1
------------------------------------------------------
1
- + 1
--------------------------------------req
Ec
----
Nt- k
E DPDCH
Total Noise Option

UL
P b k req
W - ------------------------------------------------------------------------------UL
Q req k = ------------------------
R99
tx
TP P UL k I intra + I extra + I inter carrier + N 0
UL
P b k req
W - --------------------------------------UL
Q req k = ------------------------
UL
R99
tx
TP P UL k I intra F + N 0
R99
TP P UL k
UL
UL
- I intra F UL + N tx
P b k req = Q req k ------------------------0
W
R99
req
TP P UL k
Ec
UL
We note ------ k
= Q req k ------------------------ Nt E DPDCH
W
req
Ec
UL
UL
tx
P b k req = ------ k
I intra F + N 0
Nt E DPDCH
280

AT332_TRR_E0
As I intra =
Pb
UL
k req , we have:
I intra = I intra F
UL
tx
+ N0
req
- k
----Nt E DPDCH
Ec
K
tx
N0
Ec- req
----k
Nt E DPDCH
K
I intra = ------------------------------------------------------------UL
1F
UL
UL
I intra + I extra + I inter carrier
I intra F
1
= ------------------------------------------------------------------------------= --------------------------------------= ---------------------------------UL
tx
tx
tx
I intra + I extra + I inter carrier + N 0
I intra F + N 0
N0
1 + ------------------------UL
I intra F
Therefore, we have:
X
UL
= F
UL
req
Ec-
----k
Nt E DPDCH
So, we can conclude that the contribution of one user to the UL load is defined as:
UL
X k = F
UL
req
Ec
------ k
Nt E DPDCH
4.3.4.5 Inter-carrier Power Sharing Modelling

Inter-carrier power sharing enables the network to dynamically allocate available power from R99-only and HSDPA carriers
among HSDPA carriers.
In this part, we will consider the most common scenario, a network consisting of an R99-only carrier (c1) and an HSDPA carrier
with dynamic power allocation (c2) (c2 does not support HSUPA).
As explained in The User Manual, the maximum power of the HSDPA cell must be set to the same value as the maximum
shared power in order to use power sharing efficiently. In this case, the HSDPA cell can use 100% of the available power, i.e,
all of the R99-only cells unused power can be allocated to the HSDPA cell.
Lets take the following example to measure the impact of the inter-carrier power sharing.
1st case: Inter-carrier power sharing is not activated
On c1, we have: P max Tx c 1 = 43dBm and P tx R99 Tx c 1 = 39,1dBm .

On c2, we have: P max Tx c 2 = 43dBm , P tx R99 Tx c 2 = 36,1dBm and P Headroom Tx c 2 = 0dB .
Therefore, P HSDPA Tx c 2 = P max Tx c 2 P tx R99 Tx c 2 P Headroom Tx c 2 = 42dBm
2nd case: Inter-carrier power sharing is activated and P max Tx = 46dBm
On c1, we have: P max Tx c 1 = 43dBm and P tx R99 Tx c 1 = 39,1dBm .

On c2, we have: P max Tx c 2 = 46dBm , P tx R99 Tx c 2 = 36,1dBm and P Headroom Tx c 2 = 0dB .
Therefore, P HSDPA Tx c 2 = P max Tx P tx R99 Tx c 1 P tx R99 Tx c 2 P Headroom Tx c 2 = 44,4dBm
4.3.4.6 Best Serving Cell Determination in Monte Carlo Simulations - Old Method
Before Atoll 2.8.0, best serving cell determination used to be performed by selecting the best carrier within transmitters
according to the selected method (site equipment) and then the best transmitter using the best carrier. To switch back to this
method, add the following lines in the Atoll.ini file:
[CDMA]
MultiBandSimu = 0
The method is described below:
For each station txi containing Mb in its calculation area and using a frequency band supported by the Mbs terminal.
281

Determination of BestCarrier k txi M b .

If a given carrier is specified for the service requested by Mb and if it is used by txi
BestCarrier k txi M b is the carrier specified for the service.
Else the carrier selection mode defined for txi is considered.
For each carrier ic used by txi, we calculate current loading factor:
UL
I tot txi ic
UL
UL
X k txi ic = ---------------------------- + X
UL
N tot txi ic
EndFor
UL
BestCarrier k txi M b is the carrier with the lowest X k txi ic

BestCarrier k txi M b is the carrier with the lowest P tx txi ic k
BestCarrier k txi M b is randomly selected
UL
UL
BestCarrier k txi M b is the first carrier so that X k txi ic X max

Calculation of
BTS P c txi M b BestCarrier
Q pilot txi BestCarrier = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------k
DL
DL
P tot txi BestCarrier k txi M b + I extra BestCarrier k txi M b +
DL
DL
Term
I inter carrier BestCarrier k txi M b + I inter techno log y BestCarrier k txi M b + N 0

BTS P c txi M b BestCarrier
Q pilot txi BestCarrier = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------k
DL
DL
P tot txi BestCarrier k txi M b + I extra BestCarrier k txi M b
DL
DL
+I
BestCarrier k txi M b + I inter techno log y BestCarrier k txi M b
inter
carrier
Term
N
+
txi
BestCarrier
0
BTS
c
b
Rejection of station txi if the pilot is not received

pilot
If Q pilot txi M b BestCarrier Q req Mobility M b then txi is rejected by Mb

k
max
If Q pilot txi M b BestCarrier Q pilot M b

k
Admission control (If simulation respects a loading factor constraint and Mb was not connected in previous iteration).
UL
UL
If X k txi BestCarrier txi M b X max , then txi is rejected by Mb

Else
max
Q pilot M b = Q pilot txi M b BestCarrier

k
Tx BS M b = txi
Endif
EndFor
If no TxBS has been selected, Mb has failed to be connected to the network and is rejected.
282

AT332_TRR_E0
4.4 UMTS HSPA Prediction Studies

4.4.1 Best Serving Cell and Active Set Determination
The mobile active set is the list of the cells to which the mobile is connected. The active set may consist of one or more cells
depending on whether the service supports soft handover and on the terminal active set size. The best serving cell and other
cells of the active set must fulfil a set of conditions:
They must use a frequency band with which the terminal is compatible.
They must also belong to layers supported by the service and the terminal, and these layers must support a speed
higher than the user mobility.
In addition, the pilot signal level received from these cells must exceed the defined minimum RSCP threshold.
These cells are referred to as potential serving cells.

The layer priority, the quality of the pilot ( Q pilot ), the handover margin ( M HO ) and the cell individual offset ( CIO ) are considered to select the best serving cell.
Among the potential serving cells, Atoll selects a list of candidate cells whose pilot quality exceeds the Ec/I0 threshold defined
in the properties of the mobility type. The cell of the highest priority layer with the highest RSCP is considered as the best
serving cell candidate ( c BC ).
Atoll calculates the best server indicator ( I BS ) for the best serving cell candidate and the other candidate cells ( c OC ):
I BS c BC = Q pilot c BC + M HO c BC + C IO c BC for the best serving cell candidate,
I BS c OC = Q pilot c OC + C IO c OC for the other candidate cells.
The candidate cells are ranked according to the best server indicator ( I BS ). The cell with the highest I BS is selected as the best
serving cell.
Each other cell of the active set is selected among the potential serving cells as follows:
It must use the same carrier as the best serving cell.

The pilot quality difference between the cell and the best serving cell must not exceed the AS-threshold set per cell.
It must belong to the neighbour list of the best serving cell if it is located on a site where the equipment imposes this
restriction (the restricted to neighbours option selected in the equipment properties).
You can return to the old best serving cell selection mechanism as in Atoll 3.2.1, by setting
an option in the Atoll.ini file. For more information about setting options in the Atoll.ini file,
see the Administrator Manual.
4.4.2 Point Analysis - AS Analysis Tab

Let us suppose a receiver with a terminal, a service and a mobility type. This receiver does not create any interference. You
can make the prediction for a specific carrier, for all carriers, for a specific layer, or for all layers. For DC-HSDPA, MC-HSDPA
and DB-MC-HSDPA users, selecting one specific carrier or one layer associated with one unique carrier is not suitable. If you
have selected a DC-HSDPA user or a MC-HSDPA user, select "Best (All/Specific band)" as the carrier or layers associated with
several carriers. For a DB-MC-HSDPA user, select "Best (All bands)" as the carrier or layers associated with several carriers on
different frequency bands.
The analysis is based on the following parameters:
The uplink load factor and the downlink total power of cells,
The available HSDPA power of the cell in case of an HSDPA bearer user,
The cell UL reuse factor, the cell UL load factor due to HSUPA and the maximum cell UL load factor for HSUPA bearer
users.
These parameters can be results of a given simulation, average values calculated from a group of simulations, or user-defined
cell inputs. In the last case, when no value is defined in the Cells table, Atoll uses the following default values:
Total transmitted power = 50% of the maximum power (i.e, 40 dBm if the maximum power is set to 43 dBm)
Uplink load factor = 50%.
Uplink reuse factor = 1
Uplink load factor due to HSUPA = 0%
Maximum uplink load factor = 75%
On the other hand, no default value is used for the HSDPA power; this parameter must be defined by the user.
283

Results are displayed for any point of the map where the pilot signal level exceeds the defined minimum RSCP.
4.4.2.1 Bar Graph and Pilot Sub-Menu

Atoll performs a first selection of potential serving cells depending on if you have chosen "Carrier" or "Layer".
We can consider the following cases:
1st case: Analysis based on a specific carrier
The carrier that can be used is fixed. The potential serving cells are selected among all transmitters i that contain the receiver
in their calculation areas and that use the selected carrier ic.
2nd case: Analysis based on all carriers of all frequency bands/a specific frequency band
If you have selected "Best (All bands)", the potential serving cells are selected among all transmitters i that contain the
receiver in their calculation areas.
If the frequency band is fixed ("Best (Specific band)"), the potential serving cells are selected among all transmitters i that
contain the receiver in their calculation areas and that use a carrier of the selected frequency band.
3rd case: Analysis based on the best layer
The layer that can be used is fixed. The potential serving cells are selected among all transmitters i that contain the receiver
in their calculation areas and that have cells using the selected layer.
4th case: Analysis based on all layers
The potential serving cells are selected among all transmitters i that contain the receiver in their calculation areas.
In addition, potential serving cells must satisfy the following conditions:
They must use a frequency band with which the terminal is compatible.
They must also belong to layers supported by the service and the terminal, and these layers must support a speed
higher than the user mobility.
The pilot signal level received from these cells must exceed the defined minimum RSCP threshold.
Atoll calculates the pilot quality for all potential serving cells (i, ic).
ic is the studied carrier and icadj is another carrier adjacent to ic. The interference reduction factor, RF ic ic adj , is defined
between ic and icadj and set to a value different from 0.
Two ways may be used to calculate I0.
Option Total noise: Atoll considers the noise generated by all the transmitters and the thermal noise.
Option Without pilot: Atoll considers the total noise deducting the pilot signal.
Calculation option may be selected in Global parameters.
Therefore, we have:
BTS P c i ic
Q pilot i ic = --------------------------------------------DL
I 0 ic
With,
DL
DL
DL
DL
DL
term
I 0 ic = P tot i ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0
for the total noise option,
And
DL
DL
DL
DL
DL
term
I 0 ic = I intra ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0

pilot option.
1st step: P c i ic calculation for each potential serving cell (i, ic)
P c i ic is the pilot power of a transmitter i on carrier ic at the receiver.
P pilot i ic
P c i ic = ------------------------LT
I
L T is the total loss between transmitter i and receiver.

I
284
1 BTS P c i ic for the without

AT332_TRR_E0
L Tx L path L term L body L Indoor M Shadowing Ec Io

L T = -----------------------------------------------------------------------------------------------------------------------------------I
G Tx G term
DL
DL
DL
2nd step: P tot j ic , P tot i ic and P tot j ic adj calculations

We have:
DL
I extra ic =
DL
P tot j ic
txj j i
P SCH ic
DL
DL
DL
I intra ic = P tot i ic BTS P tot i ic ------------------LT
Ptot j icadj
DL
DL
txj j
I inter carrier ic = ---------------------------------------RF ic ic adj
and
DL
ni
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
DL
For each transmitter of the network, P tot ic is the total power received at the receiver from the transmitter on the carrier ic.
P Tx ic
DL
P tot ic = ---------------LT
P Tx ic is the total power transmitted by the transmitter on the carrier ic.Total power transmitted by each cell is either a
simulation result (provided in Simulation properties (Cells tab)) or a value user-defined in Cell properties.
DL
For each transmitter of the network, P tot ic adj is the total power received at the receiver from the transmitter on the carrier
icadj.
P Tx ic adj
DL
P tot ic adj = ---------------------LT
P Tx ic adj is the total power transmitted by the transmitter on the carrier icadj. Total power transmitted by each cell is either
a simulation result (provided in Simulation properties (Cells tab)) or a value user-defined in Cell properties.
term
3rd step: N 0
term
N0
calculation
Tx DL
= NF Term K T W NR inter techno log y

DL
4th step: I 0 ic and Q pilot i ic evaluation using formulas described above

DL
5th step: G macro diversity calculation

DL
The macro-diversity gain, G macro diversity , models the decrease in shadowing margin due to the fact there are several
available pilot signals at the mobile.
DL
npaths

npaths
M Shadowing Ec Io is the shadowing margin when the mobile receives n pilot signals (not necessarily from transmitters
belonging to the mobile active set).
This parameter is determined from cell edge coverage probability and Ec/I0 standard
deviation. When the Ec/I0 standard deviation is set to 0, the macro-diversity gain equals
0.
6th step: Determination of the best serving cell
Among the potential serving cells, Atoll selects the cells whose pilot quality exceeds the Ec/I0 threshold defined in the properties of the mobility type.
285

req
If Q pilot i ic Q pilot , no cell among the potential serving cells can enter the active set. Pilot is unavailable.
req
If Q pilot i ic Q pilot , the cells are selected as candidate cells.

Among all candidate cells, Atoll first selects the cells which belong to the highest priority layer and then, the one with the highest RSCP. This cell is referred to as the best serving cell candidate ( c BC ).
Atoll calculates the best server indicator ( I BS ) for the best serving cell candidate ( c BC ) and the other candidate cells ( c OC ):
I BS c BS = Q pilot c BS + M HO c BS + C IO c BS for the best serving cell candidate,
I BS c OC = Q pilot c OC + C IO c OC for the other candidate cells.
The candidate cells are ranked according to the best server indicator ( I BS ).
Atoll takes the cell with the highest best server indicator ( c max I
Resulting
cell edge coverage probability, Q pilot

Resulting
Q pilot
BS
) and calculates the best pilot quality received with a fixed
DL
= G macro diversity Q pilot c max I
BS
This cell enters the active set as best serving cell, BS. Its carrier (icBS) will be used by other transmitters of the active set (when
active set size is greater than 1). Pilot is available.
7th step: Determination of the active-set
Then, pilot qualities received from all potential serving cells other than BS ( Q pilot i ic BS ) are recalculated to determine the
entire receiver active set (when active set size is greater than 1). Same formulas and calculation method are used to update
DL
I 0 ic BS value and determine Q pilot i ic BS .

We have:
BTS P c i ic
Q pilot i ic = --------------------------------------------DL
I 0 ic
With,
DL
DL
DL
DL
DL
term
I 0 ic = P tot i ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0
for the total noise option,
And
DL
DL
DL
DL
DL
term
I 0 ic = I intra ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0
1 BTS P c i ic for the without
pilot option.
Other cells (i,icBS) in the active set must satisfy the following criteria:
Q pilot i ic BS Q pilot BS AS_threshold BS
i ic BS neighbour list BS (optionally)
4.4.2.2 Downlink R99 Sub-Menu

The Downlink R99 sub-menu contains R99-related results.
Atoll calculates the traffic channel quality from each cell (k,icBS) of the receivers active set at the receiver. No power control
is performed as in simulations. Here, Atoll determines the downlink traffic channel quality at the receiver for the maximum
allowed traffic channel power per transmitter. Then, after combination, the total downlink traffic channel quality is evaluated
and compared with the specified target quality.
Eb/Nt Target
DL
Eb/Nt target ( Q req ) is defined for a given R99 bearer, a mobility type and a reception equipment. This parameter is available
in the R99 Bearer Selection table.
286

AT332_TRR_E0
Compressed mode is operated when amobile supporting compressed mode is connected

to a cell located on a site with a compressed-mode-capable equipment, and
Either the received Ec/I0 is lower than the Ec/I0 activation threshold (Global
Resulting
parameters): Q pilot
CM activation
Q pilot
Or the pilot RSCP is lower than the pilot RSCP activation threshold (Global
CM activation
parameters): P c RSCP pilot
When compressed mode is activated, the downlink Eb/Nt target is increased by the value
DL
user-defined for the DL Eb/Nt target increase field (Global parameters), Q req .
Required transmitter power on traffic channels
req
The calculation of the required transmitter power on traffic channels ( P tch ) may be divided into three steps.
DL
1st step: Q max k ic BS evaluation for each cell

DL
Let us assume the following notation: Eb/Nt max corresponds to Q max

Therefore, for each cell (k,icBS), we have:
DL
BTS P b max k ic BS
DL
DL
- G DL
Q max k ic BS = -----------------------------------------------------p G Div
DL
N tot ic BS
max
P tch
DL
With P b max k ic BS = ---------LT
k
DL
DL
DL
DL
DL
term
and N tot ic BS = I intra ic BS + I extra ic BS + I inter carrier ic BS + I inter techno log y ic BS + N 0

Where
max
P tch is the maximum power allowed on traffic channels. This parameter is user-defined in the R99 Radio Bearers table.
DL
N tot ic BS is the total noise at the receiver on the carrier of the best serving cell.
DL
I intra ic BS is the intra-cell interference at the receiver on the carrier of the best serving cell.
P SCH k ic BS
DL
I intra ic BS = P DL k ic BTS F ortho P DL k ic ----------------------------- tot
tot
BS
BS
L
T
DL
I extra ic BS is the extra-cell interference at the receiver on the carrier of the best serving cell.
DL
I extra ic BS =
Ptot j icBS
DL
j j k
DL
I inter carrier ic BS
is the inter-carrier interference at the receiver on the carrier of the best serving cell.
Ptot j icadj
DL
DL
j
---------------------------------------I inter carrier ic BS = txj
RF ic BS ic adj
icadj is a carrier adjacent to icBS.

RF ic BS ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
DL
I inter techno log y ic BS is the inter-technology interference at the receiver on the carrier of the best serving cell.
DL
I inter techno log y ic BS =
ni
ic i is the i
th
Tx
P Transmitted ic i
----------------------------------------Tx
Tx m
L total ICP ic ic
i
BS
287

Tx m
ICP ic ic
i
BS
is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the
frequency gap between ic i (external network) and ic BS .

2nd step: Calculation of the total traffic channel quality
DL
Q MAX is the traffic channel quality at the receiver on icBS after signal combination of all the transmitters k of the active set.
On downlink, if there is no handoff, we have:
DL
DL
Q MAX ic BS = Q max k ic BS
For any other handoff status, we have:
DL
DL
Q MAX ic BS = f rake efficiency
Qmax k icBS
DL
Where
DL
f rake efficiency is the downlink rake efficiency factor defined in Terminal properties.
req
3rd step: P tch calculation

DL
Q req
req
- P max
P tch = -------------------------tch
DL
Q MAX ic BS
Compressed mode is operated when a mobile supporting compressed mode is
connected to a cell located on a site with a compressed-mode-capable equipment, and
Either the received Ec/I0 is lower than the Ec/I0 activation threshold (Global
Resulting
parameters): Q pilot
CM activation
Q pilot
Or the pilot RSCP is lower than the pilot RSCP activation threshold (Global
CM activation
parameters): P c RSCP pilot
When compressed mode is activated, the downlink Eb/Nt target is increased by the value
DL
user-defined for the DL Eb/Nt target increase field (Global parameters), Q req . In this
DL
DL
Q req Q req
req
- P max
case, we have: P tch = -----------------------------tch
DL
Q MAX ic BS
Max Eb/Nt for Each Cell of Active Set
For each cell (k,icBS), we have:
DL
BTS P b max k ic BS
DL
DL
- G DL
Q max k ic BS = -----------------------------------------------------p G Div
DL
N tot ic BS
max
P tch
DL
k
DL
DL
DL
DL
DL
term
N tot ic BS = I intra ic BS + I extra ic BS + I inter carrier ic BS + I inter techno log y ic BS + N 0
max
req
P tch P tch
P SCH k ic BS
DL
I intra ic BS = P DL k ic BTS F ortho P DL k ic ----------------------------- 1 BTS max (--------------------------,0)
tot
tot
BS
BS
L
L
T
DL
I extra ic BS =
Ptot j icBS
DL
j j k
Ptot j icadj
DL
DL
txj j
I inter carrier ic BS = ---------------------------------------RF ic BS ic adj
288
Tk

AT332_TRR_E0
DL
Tx
P Transmitted ic i
----------------------------------------Tx
Tx m
L total ICP ic ic
i
BS
Where
req
P tch is the required transmitter power on traffic channels.

Max Eb/Nt
DL
Q MAX is the traffic channel quality at the receiver on icBS after signal combination of all the transmitters k of the active set.
DL
DL
Q MAX ic BS = Q max k ic BS
DL
DL
Qmax k icBS
DL
Where
DL
DL
DL
DL
DL
DL
Therefore, the service on the downlink traffic channel is available if Q MAX ic BS Q req (or Q MAX ic BS Q req Q req when
compressed mode is activated).
Effective Eb/Nt
DL
Q eff is the effective traffic channel quality at the receiver on icBS.

DL
DL
DL
DL
DL
DL
DL
Q eff = min Q MAX Q req (or Q eff = min Q MAX Q req Q req when compressed mode is activated).
Downlink Soft Handover Gain
DL
G SHO corresponds to the DL soft handover gain.

DL
Q MAX ic BS
DL
G SHO = -----------------------------------------------DL
max Qmax k ic BS
DL
DL
max Qmax k ic BS corresponds to the highest Q max k ic BS value.
4.4.2.3 Uplink R99 Sub-Menu

The Uplink R99 sub-menu contains R99-related results.
For each cell (k,icBS) in the receivers active set, Atoll calculates uplink traffic channel quality from receiver. No power control
is performed as in simulations. Here, Atoll determines the uplink traffic channel quality at the cell for the maximum terminal
power allowed. Then, the total uplink traffic channel quality is evaluated with respect to the receiver handover status. From
this value, Atoll calculates the terminal power required to obtain the R99 bearer and compares it to the maximum terminal
power allowed.
Max Terminal Power
max
Max terminal power ( P term ) is an input user-defined for each terminal. It corresponds to the terminals maximum power.
Required Terminal Power
req
The calculation of the terminal power required to obtain an R99 bearer ( P term R99 ) may be divided into three steps.
UL
1st step: Q max k ic BS evaluation for each cell

For each cell (k,icBS) in the receivers active set, we have:
289

UL
term P b max k ic BS
UL
UL
- G UL
Q max k ic BS = -------------------------------------------------------p G Div
UL
N tot k ic BS
max
UL
P term 1 r c
UL
With P b max k ic BS = --------------------------------------LT
k
UL
N tot k ic BS is the total noise at the transmitter on the carrier of the best serving cell. This value is calculated from the cell
UL
uplink load factor X k ic BS .

tx
N0
UL
N tot k ic BS = -----------------------------------UL
1 X k ic BS
tx
N 0 is the transmitter thermal noise.

2nd step: Calculation of the total traffic channel quality
UL
Q MAX ic BS is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters k of the
active set.
UL
UL
If there is no handoff (1/1): Q MAX ic BS = Q max k ic BS

For soft handoff (2/2):
UL
UL
UL
Q MAX ic BS = G macro diversity 2 links max Q max k ic BS

UL
G macro diversity 2 links is the uplink macro-diversity gain. This parameter is determined from the fixed cell edge coverage
probability and the uplink Eb/Nt standard deviation. When the option Shadowing taken into account is not selected
(Prediction properties), Atoll considers the uplink macro-diversity gain defined by the user in Global parameters.
UL
UL
max Q max k ic BS corresponds to the highest Q max k ic BS value.

For soft-soft handoffs (3/3):
UL
UL
UL

UL
For softer and softer-softer handoffs (1/2 and 1/3):
UL
UL
Qmax k icBS
UL
For softer-soft handoffs (2/3), there are two possibilities. If the MRC option is selected (option available in Global parameters),
we have:
UL
UL
UL
Q MAX ic BS = G macro diversity 2 links max f rake efficiency
UL
UL
Q max k ic BS Q max
k on the same site
k on the same site
k ic BS
Else,
UL
UL
UL

req
3rd step: P term R99 calculation

req
P term R99 is the required terminal power.

UL
Q req
req
- P max
P term R99 = -------------------------term
UL
Q MAX ic BS
UL
Q req is the uplink traffic quality target defined by the user for a given reception equipment, a given R99 bearer and a given
mobility type. This parameter is available in the R99 Bearer Selection table.
290

AT332_TRR_E0
Compressed mode is operated when a mobile supporting compressed mode is

connected to a cell located on a site with a compressed-mode-capable equipment, and
The received Ec/I0 is lower than the Ec/I0 activation threshold (Global parameters):
Resulting
Q pilot
CM activation
Q pilot
The pilot RSCP is lower than the pilot RSCP activation threshold (Global parameters):
CM activation
P c RSCP pilot
When compressed mode is activated, the uplink Eb/Nt target is increased by the value
UL
user-defined for the UL Eb/Nt target increase field (Global parameters), Q req . In this
UL
UL
Q req Q req
req
- P max
case, we have: P term R99 = -----------------------------term
UL
Q MAX ic BS
req
max
Therefore, the service on the uplink traffic channel is available if P term R99 P term .
Eb/Nt Max
For each cell (k,icBS) in the receivers active set, we have:
UL
term P b max k ic BS
UL
UL
- G UL
Q max k ic BS = -------------------------------------------------------p G Div
UL
N tot k ic BS
max
UL
P term 1 r c
UL
With P b max k ic BS = --------------------------------------LT
k
UL
N tot k ic BS is the total noise at the transmitter on the carrier of the best serving cell. This value is calculated from the cell
UL
uplink load factor X k ic BS .

tx
max
req
N0
P term P term R99
UL
N tot k ic BS = -----------------------------------,0)
- + 1 term max (-----------------------------------------UL
LT
1 X k ic BS
k
tx

UL
Q MAX ic BS is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters k of the
active set.
UL
UL
If there is no handoff (1/1): Q MAX ic BS = Q max k ic BS

For soft handoff (2/2):
UL
UL
UL

UL
UL
UL
max Qmax k ic BS corresponds to the highest Q max k ic BS value.

For soft-soft handoffs (3/3):
UL
UL
UL

UL
For softer and softer-softer handoffs (1/2 and 1/3):
UL
UL
Qmax k icBS
UL
291

For softer-soft handoffs (2/3), there are two possibilities. If the MRC option is selected (option available in Global parameters),
we have:
UL
UL
UL
UL
UL
Q max k ic BS Q max
k on the same site
k on the same site
k ic BS
Else,
UL
UL
UL

Effective Eb/Nt
UL
Q eff is the effective traffic channel quality at the transmitter on icBS.

UL
UL
UL
UL
UL
UL
UL
Q eff = min Q MAX Q req (or Q eff = min Q MAX Q req Q req when compressed mode is activated).
Uplink Soft Handover Gain
UL
G SHO corresponds to the uplink soft handover gain.

UL
Q MAX ic BS
UL
G SHO = -----------------------------------------------UL
max Q max k ic BS
UL
UL
4.4.2.4 HSDPA Sub-Menu

The HSDPA sub-menu contains HSDPA-related results for HSDPA and HSPA users when the HS-SCCH quality is sufficient and
if the user can obtain an HSDPA bearer.
Atoll determines the best HSDPA bearer that the user can obtain. The HSDPA bearer user is processed as if he is the only user
in the cell, i.e. he uses the entire HSDPA power available in the cell. For further information on the fast link adaptation
modelling, see "Fast Link Adaptation Modelling" on page 240.
When modelling MC-HSDPA users (including DC-HSDPA users) and DB-MC-HSDPA users, Atoll determines the serving cells and
the best HSDPA bearer obtained in each serving cell. In each cell, the user is processed as if he is the only user in the cell. Atoll
details the results for each cell to which the user is connected. For further information on MC-HSDPA user modelling, see "MCHSDPA Users" on page 294. For further information on DB-MC-HSDPA user modelling, see "DB-MC-HSDPA Users" on
page 294.
General Results
Atoll displays the name of the cell to which the user is connected, the frequency band used by the transmitter, the selected
carrier, and the maximum available HSDPA power of the cell.
HS-PDSCH Ec/Nt
Atoll calculates the best HS-PDSCH quality (HS-PDSCH Ec/Nt). The way of calculating it depends on the selected option in the
transmitters global parameters (HSDPA part): CQI based on CPICH quality or CQI based on HS-PDSCH quality.
For further details on the HS-PDSCH quality calculation, see either "HS-PDSCH Quality Calculation" on page 242 if the selected
option is "CQI based on CPICH quality" or "HS-PDSCH Quality Calculation" on page 247 if the selected option is "CQI based on
HS-PDSCH quality".
HS-PDSCH Power
Atoll calculates the available HS-PDSCH power. For further details on the HS-PDSCH power calculation, see either "HS-PDSCH
Quality Calculation" on page 242 if the selected option is "CQI based on CPICH quality" or "HS-PDSCH Quality Calculation" on
page 247 if the selected option is "CQI based on HS-PDSCH quality".
HS-SCCH Ec/Nt Threshold
Atoll displays the HS-SCCH Ec/Nt threshold set for the selected mobility type.
HS-SCCH Ec/Nt
Atoll displays the obtained HS-SCCH quality.
292

AT332_TRR_E0
When the HS-SCCH power allocation strategy is dynamic, this parameter corresponds to the HS-SCCH Ec/Nt threshold defined
for the selected mobility type.
When the HS-SCCH power allocation strategy is static, the HS-SCCH Ec/Nt is calculated from the fixed HS-SCCH power.
We have:
BTS P c ic
Eci
----
ic
=
------------------------------- for the total noise option,
Nt
HS SCCH
DL
N tot ic
And
BTS P c ic
Eci
---- ic
= ---------------------------------------------------------------------------------------------------------------------------- for the without useful signal option.
Nt
HS SCCH
DL
term
i
With
DL
DL
DL
DL
DL
term
DL
DL
DL
P SCH ic
P SCH ic
DL
term
I intra ic = P tot ic + BTS 1 F MUD 1 F ortho P tot ic ------------------- BTS P tot ic ------------------
LT
LT
txi
txi
txi
DL
I extra ic =
DL
P tot ic
txj j i
Ptot icadj
DL
DL
j
RF ic ic adj

DL

DL
ni
ic i is the i
th
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
Tx m
i

P HS SCCH ic
P c ic = ------------------------------i
LT
i
And
L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx G term
term
term
CQI
It corresponds to the HS-PDSCH CQI.
The way of calculating it depends on the selected option in the transmitters global parameters (HSDPA part): CQI based on
CPICH quality or CQI based on HS-PDSCH quality.
For further details on the HS-PDSCH quality calculation, see either "HS-PDSCH CQI Determination" on page 244 if the selected
option is "CQI based on CPICH quality" or "HS-PDSCH CQI Determination" on page 250 if the selected option is "CQI based on
HS-PDSCH quality".
293

HSDPA Bearer Parameters

Knowing the HS-PDSCH CQI, Atoll calculates the best HSDPA bearer that can be used and selects a bearer compatible with cell
and terminal user equipment HSDPA capabilities. For further details on the HSDPA bearer selection, see "HSDPA Bearer
Selection" on page 244.
Atoll displays the parameters of the selected HSDPA bearer:
The transport block size,

The modulation scheme used,
The number of HS-PDSCH channels used.
Peak RLC Throughput

DL
Once the bearer selected, Atoll determines the peak RLC throughput that can be provided to the user TP P R LC .
Effective RLC Throughput
DL
Atoll displays the Effective RLC throughput ( TP E RLC ) provided to the user. The Effective RLC throughput is calculated as
follows:
DL
TP P RLC
DL
TP E RLC = ----------------TTI
Where TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal
user equipment category properties.
BLER
Atoll reads the BLER in the quality graph BLER = f(HS-PDSCH Ec/Nt) that is defined for the selected bearer and mobility type.
Knowing the HS-PDSCH Ec/Nt, it finds the corresponding BLER.
Bearer Consumption
Atoll provides this result for HSPA CBR service users only. The minimum throughput demand required by the service is
allocated to these users. Therefore, they partly consume the HSDPA bearer. The bearer consumption expressed in %,
C HSDPABearer , is calculated as follows:
TPD Min DL
C HSDPABearer = --------------------------------------------------DL
TP P R LC I HSDPABearer
MC-HSDPA Users
When multi-cell HSDPA is active, MC-HSDPA users can simultaneously connect to several HSDPA cells of the transmitter for
data transfer. The maximum number of cells to which the user can simultaneously connect depends on the DL multi-cell mode
set for the HSDPA UE category of the terminal.
Atoll determines the best serving cell using the best serving cell selection algorithm. For information on how the best serving
cell is selected, see "Best Serving Cell and Active Set Determination" on page 282. If the best carrier belongs to a transmitter
that supports the multi-cell HSDPA mode and if the transmitter has several HSDPA carriers, Atoll selects the other serving
cells, i.e., the secondary cells. The secondary cells belong to the same transmitter and are chosen among the adjacent carriers
according to the CQI. When two adjacent carriers are available, Atoll takes the one with the highest CQI value. Atoll selects
secondary cells as long as HSDPA carriers are available in the transmitter and the maximum number of cells to which the user
can simultaneously connect is not exceeded. In each serving cell (i.e., the best cell and the secondary cells), Atoll determines
the best HSDPA bearer obtained. In each cell, the user is processed as if he is the only user in the cell. The user is connected
to a cell if he obtains an HSDPA bearer.
DB-MC-HSDPA Users
When multi-cell HSDPA and dual-band HSDPA modes are active, DB-MC-HSDPA users can simultaneously connect to HSDPA
cells of two co-site transmitters using different frequency bands. If the two co-site transmitters work on the same frequency
band, then the users can only connect to the HSDPA cells of one transmitter. The maximum number of cells to which the user
can simultaneously connect depends on the DL multi-cell mode set for the HSDPA UE category of the terminal.
Lets consider the following configuration:
294
A site with transmitters working on two different frequency bands,

The site equipment supports the dual-band HSDPA mode,
Each transmitter has several HSDPA carriers,

AT332_TRR_E0
The multi-cell HSDPA mode is active for each transmitter.
Atoll determines the best serving cell using the best serving cell selection algorithm. For information on how the best serving
cell is selected, see "Best Serving Cell and Active Set Determination" on page 282. The secondary cells are taken in the same
band as the best carrier (i.e., they belong to the same transmitter), as long as carriers are available. Then, if additional carriers
are required and if there are no more carriers available in this transmitter, Atoll selects the carriers in a transmitter using the
second frequency band. Within one frequency band, the secondary cells are first selected according to an adjacency criterion
and then, according to the CQI value. When two adjacent carriers are available, Atoll takes the one with the highest CQI value.
In each serving cell (i.e., the best cell and the secondary cells), Atoll determines the best HSDPA bearer obtained. In each cell,
the user is processed as if he is the only user in the cell. The user is connected to a cell if he obtains an HSDPA bearer.
Results for MC-HSDPA and DB-MC-HSDPA Users
When the user is simultaneously connected to several HSDPA cells, Atoll details the results for each cell. In addition, it displays
the following results under Total:
DL
The Peak RLC Throughput
TP P RLC =
DL
TP P RLC c
c Serving cell
The Effective RLC Throughput

DL
TP P RLC
DL
TP E RLC = ----------------TTI
The Application Throughput
DL
DL
Serving cells
- f TP Scaling TP Offset
TP A = c---------------------------------------------------------------------------------------------------------TTI
Where:
quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll finds the corresponding BLER.
4.4.2.5 HSUPA Sub-Menu

The HSUPA sub-menu contains HSUPA-related results for HSPA users if the user can obtain an HSUPA bearer.
Atoll determines the best HSUPA bearer that the user can obtain. The HSUPA bearer user is processed as if he is the only user
in the cell, i.e. he uses the entire remaining load of the cell.
For further information on the HSUPA bearer selection, see "HSUPA Bearer Allocation Process" on page 259.
Required E-DPDCH Ec/Nt
Ec req
It corresponds to the E-DPDCH Ec/Nt required to obtain the HSUPA bearer ( ------
). This value is defined for an HSUPA
Nt E DPDCH
bearer ( Index HSUPABearer ) and a number of retransmissions ( N Rtx ) in the HSUPA Bearer Selection table.
Required Terminal Power
Ec req
req
From ------
, Atoll calculates the terminal power required to obtain the HSUPA bearer, P term HSUPA .
Nt E DPDCH
Ec req
req
UL
P term HSUPA = ------
L N tot
Nt E DPDCH T
295

With
UL
UL
tx
intra
UL
extra
ic + I tot
tx
UL

UL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
tx
UL intra
UL extra
, I tot
UL
tx
Max E-DPDCH Ec/Nt

max
Ec
Atoll calculates the maximum E-DPDCH EcNt allowed ( ------
). For further details on the calculation, see "Max ENt E DPDCH
DPDCH Ec/Nt" on page 296.
HSUPA Bearer Parameters
Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. This is the HSUPA bearer with the highest potential
UL
TP P RLC Index HSUPABearer

- ) where:
throughput ( -----------------------------------------------------------------N Rtx Index HSUPABearer
Ec- req
Ec- max
---- ----
Nt E DPDCH Nt E DPDCH
And P term HSUPA P term
max
req
With
max
P term : the maximum terminal power allowed.

Atoll displays the parameters of the selected HSUPA bearer:
The radio bearer index,

The TTI duration,
The modulation scheme used,
The number of E-DPDCH codes used.
Peak RLC Throughput

UL
After selecting the HSUPA bearer, Atoll determines the corresponding RLC peak throughput, TP P RLC .
Peak RLC Throughput/No. of RTX
UL
TP P RLC Index HSUPABearer

- ). Atoll considers the
Atoll displays the peak RLC throughput to number of retransmissions ratio ( -----------------------------------------------------------------N Rtx Index HSUPABearer
ratio to select the HSUPA bearer when several HSUPA bearers meet the selection criteria.
Min Effective RLC Throughput
UL
From the RLC peak throughput, Atoll calculates the minimum effective RLC throughput, TP Min E RLC .
UL
TP P RLC 1 BLER HSUPA

UL
TP Min E RLC M b = -----------------------------------------------------------------N Rtx
Where:
BLER HSUPA is the residual BLER after N Rtx retransmissions.
Application Throughput
UL
Atoll displays the provided application throughput ( TP A ). The application throughput represents the net throughput after
deduction of coding (redundancy, overhead, addressing, etc.). This one is calculated as follows:
296

AT332_TRR_E0
UL
TP P RLC 1 BLER HSUPA f TP Scaling TP Offset

UL
TP A M b = --------------------------------------------------------------------------------------------------------------------------N Rtx
BLER
Knowing the E-DPDCH Ec/Nt, Atoll finds the corresponding BLER. It is read in the quality graph defined for the quartet
reception equipment-selected bearer-number of retransmissions-mobility (HSUPA Quality Graphs tab in the Reception
equipment properties). This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt).
Bearer Consumption
Atoll provides this result for CBR service users only. The minimum bit rate required by the service is allocated to these users.
Therefore, they parly consume the HSUPA bearer. The bearer consumption expressed in %, C HSUPABearer , is calculated as
follows:
TPD min UL
C HSUPABearer = ----------------------------------------------------UL

Atoll calculates UMTS-specific coverage studies on each pixel where the pilot signal level exceeds the minimum RSCP
threshold. Let us assume each pixel on the map corresponds to a probe receiver with a terminal, a mobility type and a service.
This receiver does not create any interference. You can make the coverage prediction for a specific carrier, for all carriers, for
a specific layer, or for all layers. Coverage predictions are based on parameters that can be either simulation results, or userdefined cell inputs.
4.4.3.1 Pilot Quality Analysis

Atoll determines the best serving cell (BS) for each pixel and calculates the pilot quality received with a fixed cell edge
Resulting
coverage probability, Q pilot
BS .
Potential serving cells are filtered depending on the prediction definition (selected layers or carriers, layers supported by the
service and the terminal, mobility type) and the pilot signal level which must exceed the defined minimum RSCP threshold.
For further information on formulas, see "Definitions" on page 212. For information on the best serving cell selection and
pilot quality calculation, see "Bar Graph and Pilot Sub-Menu" on page 283.
4.4.3.1.1
Prediction Study Inputs

The Pilot Quality Analysis depends on the downlink total transmitted power of cells. This parameter can be either a simulation
output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the total transmitted power,
Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power is set to 43 dBm).
4.4.3.1.2
Study Display Options

Atoll displays the best pilot quality received with a fixed cell edge coverage probability.
Single colour
Resulting
Atoll displays a coverage if Q pilot
req
req
BS Q pilot . Coverage consists of a single layer with a unique colour. Q pilot is a target
value defined in the Mobility table by the user.

Colour per transmitter
Resulting
req
BS Q pilot . Coverage consists of several independent layers that can be displayed and
hidden on the map. There is a layer per transmitter. Layer colour is the colour assigned to the transmitter of the best serving
cell (BS).
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned.
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userResulting
defined mobility defined in the Mobility Types sub-folder. For each layer, area is covered if Q pilot
req
BS Q pilot .
297

Colour per probability

This display option is available only if analysis is based on all simulations in a group (i.e. if you select a group of simulations
and the All option in the Condition tab of prediction properties). Coverage consists of several independent layers that can
be displayed and hidden on the map. There is a layer per user-defined probability level defined in the Display tab (Prediction
Resulting
properties). For each layer, area is covered if Q pilot
req
BS Q pilot in the required number of simulations.
Colour per cell edge coverage probability

Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined cell edge coverage probability, p, defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting
Q pilot
req
BS p Q pilot .
Colour per quality level (Ec/I0)

Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined quality threshold defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting
Q pilot
BS Q pilot threshold .
Colour per quality margin (Ec/I0 margin)

Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined quality margin defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting
Q pilot
req
BS Q pilot Q pilot m arg in .
Colour per pilot signal level (Ec)

Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined pilot signal level defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting
Q pilot
BS Q pilot threshold .
4.4.3.2 Downlink Service Area Analysis

As in point predictions, Atoll calculates traffic channel quality at the receiver for each cell (k,icBS) in the receivers active set.
No power control is performed as in simulations. Here, Atoll determines downlink traffic channel quality at the receiver for a
DL
maximum allowed traffic channel power for transmitters. Then, the total downlink traffic channel quality ( Q MAX ic BS ) is
evaluated after recombination.
Atoll displays the total traffic channel quality in the downlink.
For information on best serving cell selection and active set determination, see "Best Serving Cell and Active Set
Determination" on page 282. For further details on calculations, see "Downlink R99 Sub-Menu" on page 286.
4.4.3.2.1

The Downlink Service Area Analysis depends on the downlink total transmitted power of cells. This parameter can be either
a simulation output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the total
transmitted power, Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power is set to
43 dBm).
4.4.3.2.2

Single colour
DL
DL
DL
DL
DL
Atoll displays a coverage with a unique colour if Q MAX ic BS Q req (or Q MAX ic BS Q req Q req if compressed mode is
activated).
DL
Q req is the downlink traffic quality target defined by the user for a given reception equipment, a given R99 bearer and a given
mobility type. This parameter is available in the R99 Bearer Selection table.
DL
Q req is the DL Eb/Nt target increase; this parameter is user-defined in the Global parameters.
298

AT332_TRR_E0

DL
DL
DL
DL
DL
Atoll displays a coverage if Q MAX ic BS Q req (or Q MAX ic BS Q req Q req if compressed mode is activated). Coverage
consists of several independent layers that can be displayed and hidden on the map. There is a layer per transmitter. Layer
colour is the colour assigned to best serving transmitter.
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned. Coverage consists of several independent layers
that can be displayed and hidden on the map. There is a layer per user-defined mobility defined in Mobility sub-folder. For
DL
DL
DL
DL
DL
each layer, area is covered if Q MAX ic BS Q req (or Q MAX ic BS Q req Q req if compressed mode is activated).
Colour per service
In this case, receiver is not completely defined and no service is assigned. Coverage consists of several independent layers that
can be displayed and hidden on the map. There is a layer per user-defined service defined in Services sub-folder. For each
DL
DL
DL
DL
DL
layer, area is covered if Q MAX ic BS Q req (or Q MAX ic BS Q req Q req if compressed mode is activated).
DL
DL
properties). For each layer, area is covered if Q MAX ic BS Q req in the required number of simulations.
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined cell edge coverage probability, p, defined in the Display tab (Prediction properties). For each layer, area is covered if
DL
DL
DL
DL
DL
Q MAX ic BS p Q req (or Q MAX ic BS Q req Q req if compressed mode is activated).

Colour per maximum quality level (max Eb/Nt)
DL
Q MAX ic BS Threshold .
Colour per effective quality level (Effective Eb/Nt)
DL
DL
DL
DL
DL
DL
DL
DL
Q eff ic BS Threshold . Q eff ic BS = min Q MAX ic BS Q req (or Q eff ic BS = min Q MAX ic BS Q req Q req when
compressed mode is activated).
Colour per quality margin (Eb/Nt margin)
DL
DL
DL
DL
DL
Q MAX ic BS Q req M arg in (or Q MAX ic BS Q req Q req M arg in when compressed mode is activated).
Colour per required power
req
Atoll calculates the downlink required power, P tch ic BS , as follows:

DL
Q req
req
- P max
P tch ic BS = -------------------------tch
DL
Q MAX ic BS
Where
DL
Q req is the Eb/Nt target on downlink. This parameter, available in the R99 Bearer Selection table, is user-defined for a given
R99 bearer, a given reception equipment and a mobility type.
299

max
P tch is a user-defined input for each bearer related to a service. It corresponds to the maximum allowable traffic channel
power for a transmitter.
DL
DL
Q req Q req
req
- P max
When compressed mode is activated, we have: P tch ic BS = -----------------------------tch .
DL
Q MAX ic BS
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined required power threshold defined in the Display tab (Prediction properties). For each layer, area is covered if
req
P tch ic BS Threshold .
Colour per required power margin
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined power margin defined in the Display tab (Prediction properties). For each layer, area is covered if
req
max
P tch ic BS P tch M arg in .
4.4.3.3 Uplink Service Area Analysis

As in point prediction, Atoll calculates uplink traffic channel quality from the receiver for each cell (k,icBS) in receiver active
set. No power control simulation is performed. Atoll determines uplink traffic channel quality at the transmitter for the
UL
maximum terminal power allowed. Then, the total uplink traffic channel quality ( Q MAX ic BS ) is evaluated with respect to
receiver handover status.
Atoll displays the total traffic channel quality in the uplink.
For information on best serving cell selection and active set determination, see "Best Serving Cell and Active Set
Determination" on page 282. For further details on calculations, see "Uplink R99 Sub-Menu" on page 289.
4.4.3.3.1

The Uplink Service Area Analysis depends on the UL load factor of cells. This parameter can be either a simulation output, or
a user-defined cell input. In the last case, when no value is defined in the Cells table for the uplink load factor, Atoll uses 50%
as default value.
4.4.3.3.2

Single colour
UL
UL
UL
UL
UL
colour is unique.
UL
Q req is defined for a reception equipment, a R99 bearer and a mobility type. This parameter is available in the R99 Bearer
Selection table.
UL
Q req is the UL Eb/Nt target increase; this parameter is user-defined in the Global parameters.
UL
UL
UL
UL
UL
consists of several independent layers that can be displayed and hidden on the map. There is a layer per transmitter. Layer
colour is the colour assigned to best server transmitter.
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned. Coverage consists of several independent layers
that can be displayed and hidden on the map. There is a layer per user-defined mobility defined in Mobility sub-folder. For
UL
UL
UL
UL
UL
each layer, area is covered if Q MAX ic BS Q req (or Q MAX ic BS Q req Q req if compressed mode is activated).
Colour per service
In this case, receiver is not completely defined and no service is assigned. Coverage consists of several independent layers that
can be displayed and hidden on the map. There is a layer per user-defined service defined in Services sub-folder. For each
UL
UL
UL
UL
UL
layer, area is covered if Q MAX ic BS Q req (or Q MAX ic BS Q req Q req if compressed mode is activated).
300

AT332_TRR_E0

UL
UL
UL
UL
UL
properties). For each layer, area is covered if Q MAX ic BS Q req (or Q MAX ic BS Q req Q req if compressed mode is
activated) in the required number of simulations.
Colour per maximum quality level (Max Eb/Nt)
UL
Q MAX ic BS Threshold .
UL
Q effective ic BS Threshold .
UL
UL
UL
Q eff ic BS = min Q MAX ic BS Q req
UL
UL
UL
UL
(or Q eff ic BS = min Q MAX ic BS Q req Q req
when compressed mode is
activated).
UL
UL
UL
UL
UL
Q MAX ic BS Q req M arg in (or Q MAX ic BS Q req Q req M arg in if compressed mode is activated).
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined power threshold defined in the Display tab (Prediction properties). For each layer, area is covered if
req
P term R99 ic BS Threshold .

Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined power margin defined in the Display tab (Prediction properties). For each layer, area is covered if
req
max
P term R99 ic BS P term M arg in .

Colour per soft handover gain
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per soft
UL
handover gain value defined in the Display tab (Prediction properties). For each layer, area is covered if G SHO Threshold .
4.4.3.4 Downlink Total Noise Analysis

Atoll determines the downlink total noise level generated by cells. You can make the coverage prediction for a specific carrier,
for all carriers, for a specific layer, or for all layers. We assume that each pixel on the map corresponds to a probe receiver
with a terminal, a mobility type and a service.
When you select "Best (All/Specific band)" as the carrier or layers associated with several carriers, Atoll determines the DL
total noise level on each carrier supported by the user service. When only one carrier is analysed, Atoll determines DL total
noise or DL noise rise on this carrier.
Ptot icadj
DL
DL
N tot ic =
txj j
DL
txj j
P tot ic + -----------------------------------+
RF ic ic adj
ni
Tx
P Transmitted ic i
term
------------------------------------- + N0
Tx
Tx m
L total ICP ic ic
i
DL
Downlink noise rise, NR DL ic , is calculated from the downlink total noise, N tot , as follows:
301

term
N0
-
NR DL ic = 10 log ----------- N DL
tot
4.4.3.4.1
Study Inputs
The Downlink Total Noise Analysis depends on the downlink total transmitted power of cells. This parameter can be either a
simulation output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the total
transmitted power, Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power is set to
43 dBm).
4.4.3.4.2
Display Options
The following display options are available for the prediction:
Colour per minimum noise level
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined noise level defined in the Display tab (Prediction properties). For each layer, area is covered if
DL
minN tot ic Threshold .

ic
Colour per maximum noise level

DL
maxN tot ic Threshold .

ic
Colour per average noise level

DL
averageN tot ic Threshold .

ic
Colour per minimum noise rise

Atoll displays bins where minNR DL ic Threshold . Coverage consists of several independent layers that can be displayed
ic
and hidden on the map. There is a layer per user-defined noise rise threshold defined in the Display tab.
Colour per maximum noise rise
Atoll displays bins where maxNR DL ic Threshold . Coverage consists of several independent layers that can be displayed
ic
and hidden on the map. There is a layer per user-defined noise rise threshold defined in the Display tab.
Colour per average noise rise
Atoll displays bins where averageNRDL ic Threshold . Coverage consists of several independent layers that can be
ic
displayed and hidden on the map. There is a layer per user-defined noise rise threshold defined in the Display tab.
When only one carrier is analysed, Atoll determines DL total noise or DL noise rise on this
carrier. In this case, the displayed coverage is the same for any selected display option
(average, minimum, maximum).
4.4.3.5 HSDPA Prediction Study

When calculating the HSDPA coverage prediction, either you can take all the possible HSDPA radio bearers into consideration,
or you can study a certain HSDPA radio bearer. Then, available display options depend on what you have selected.
When considering all the HSDPA radio bearers, you can set display parameters:
302
To analyse the uplink and downlink A-DPCH qualities on the map,

To analyse the HS-SCCH quality/power,
To model fast link adaptation for a single HSDPA bearer user or for a defined number of HSDPA bearer users.

AT332_TRR_E0
When studying a certain HSDPA radio bearer, you can display areas where a certain RLC peak throughput is available with
different cell edge coverage probabilities (i.e. the probability of having a certain RLC peak throughput). This type of analysis is
not relevant when modelling MC-HSDPA and DB-MC-HSDPA users.
Here we assume that each pixel on the map corresponds to one or several users with HSDPA capable terminal, mobility and
HSDPA service.The user does not create any interference. You can make the coverage prediction for a specific carrier, for all
carriers, for a specific layer, or for all layers. For DC-HSDPA, MC-HSDPA and DB-MC-HSDPA users, selecting one specific carrier
or one layer associated with one unique carrier is not suitable. To display the global throughput, you have to select several
carriers ("Best HSPA (All/Specific band)" as the carrier) or layers associated with several carriers.
For information on the best serving cell and secondary cells selection, see "MC-HSDPA Users" on page 294 and "DB-MCHSDPA Users" on page 294.
Note that the HSDPA service area is limited by the pilot quality, the A-DPCH quality and the HS-SCCH quality.
4.4.3.5.1

Parameters used as input for the HSDPA prediction study are:
The available HSDPA power of the cell,

The downlink total transmitted power of the cell,
The number of HSDPA bearer users within the cell if the study is calculated for several users.
These parameters can be either simulation outputs, or user-defined cell inputs. In the last case, when no value is defined in
the Cells table for the total transmitted power and the number of HSDPA bearer users, Atoll uses the following default values:
Total transmitted power = 50% of the maximum power (i.e, 40 dBm if the maximum power is set to 43 dBm)
Number of HSDPA bearer users = 1
On the other hand, no default value is used for the available HSDPA power; this parameter must be defined by the user.
4.4.3.5.2

When considering all the HSDPA radio bearers, several display options are available in the study properties dialog box. They
can be regrouped in four categories according to the objective of the study:
To analyse the uplink and downlink A-DPCH qualities on the map,

To analyse the HS-SCCH quality/power,
To model fast link adaptation for a single HSDPA bearer user,
To model fast link adaptation for a defined number of HSDPA bearer users.
When studying a certain HSDPA radio bearer, only one display option is available. It allows you to display where a certain RLC
peak throughput is available with different cell edge coverage probabilities.
Analysis of UL And DL A-DPCH Qualities
Colour per Max A-DPCH Eb/Nt DL

DL
Atoll displays the A-DPCH quality at the receiver ( Q MAX BS ) for the best serving cell (BS). No power control is performed as
in simulations. Here, Atoll determines downlink traffic channel quality at the receiver for a maximum traffic channel power
allowed for the best serving cell.
For information on calculation methods, see "Downlink R99 Sub-Menu" on page 286.
DL
Q MAX BS Threshold .
Colour per Max A-DPCH Eb/Nt UL

UL
Atoll displays the A-DPCH quality at the best serving cell ( Q MAX BS ). No power control is performed as in simulations. Here,
Atoll determines uplink traffic channel quality at the receiver for a maximum terminal power allowed.
For information on calculation methods, see "Uplink R99 Sub-Menu" on page 289.
UL
Q MAX BS Threshold .
Analysis of The HS-SCCH Quality/Power
Colour per HS-SCCH Power
303

This display option is relevant in case of dynamic HS-SCCH power allocation only. In this case, Atoll displays on each pixel the
HS-SCCH power per HS-SCCH channel. Coverage consists of several independent layers that can be displayed and hidden on
the map. There is a layer per threshold. For each layer, area is covered if P HS SCCH BS Threshold .
Colour per HS-SCCH Ec/Nt
This display option is relevant in case of static HS-SCCH power allocation only. In this case, Atoll displays on each pixel the HSSCCH quality per HS-SCCH channel. Coverage consists of several independent layers that can be displayed and hidden on the
Ec
map. There is a layer per threshold. For each layer, area is covered if ------ BS
Threshold .
Nt
HS SCCH
Fast Link Adaptation Modelling For A Single User
When you calculate the study with the following display options, Atoll considers one user on each pixel and determines the
best HSDPA bearer that the user can obtain. For MC-HSDPA and DB-MC-HSDPA users, Atoll determines the best HSDPA
bearers that the user can obtain in each serving cell. On each pixel, the user is processed as if he is the only user in the cell i.e.
he uses the entire HSDPA power available in the cell.
For further information on the fast link adaptation modelling, see "Fast Link Adaptation Modelling" on page 240.
Colour per HS-PDSCH Ec/Nt
Atoll displays on each pixel the HS-PDSCH quality. For MC-HSDPA and DB-MC-HSDPA users, it corresponds to the HS-PDSCH
Ec/Nt of the best serving cell. Coverage consists of several independent layers that can be displayed and hidden on the map.
Ec
Threshold .
There is a layer per threshold. For each layer, area is covered if ------ BS
Nt
HS PDSCH
Colour per CQI
Atoll displays either the CPICH CQI (see the calculation detail in "CPICH CQI Determination" on page 242) when the selected
option in Global parameters (HSDPA part) is CQI based on CPICH quality, or the HS-PDSCH CQI (see the calculation detail in
the section 10.7.1.2.2) when considering the CQI based on HS-PDSCH quality option.
For MC-HSDPA and DB-MC-HSDPA users, it corresponds to the CQI of the best serving cell.
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per CQI
threshold ( CQI threshold ). For each layer, area is covered if CQI CQI threshold .
Colour per Peak MAC Throughput

DL
Atoll displays the Peak MAC throughput ( TP P M AC ) provided on each pixel. The Peak MAC throughput is calculated as follows:
DL
TP P M AC =
c Serving cells
S block c
--------------------T TTI
Where,
S block c is the transport block size (in kbits) of the HSDPA bearer selected in the cell, c, for the user; it is defined for each
HSDPA bearer in the HSDPA Radio Bearers table.
3
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per possible
DL
Peak MAC throughput ( TP P M AC ). For each layer, area is covered if the Peak MAC throughput exceeds the user-defined
thresholds.
Colour per Effective MAC Throughput

DL
Atoll displays the Effective MAC throughput ( TP E M AC ) provided on each pixel. The Effective MAC throughput is calculated as
follows:
DL
TP E M AC =
c Serving cells
S block c
-------------------------T TTI TTI
Where,
S block c is the transport block size (in kbits) of the selected HSDPA bearer in the cell, c; it is defined for each HSDPA bearer
in the HSDPA Radio Bearers table.
304

AT332_TRR_E0
3
DL
Effective MAC throughput ( TP E M AC ). For each layer, area is covered if the Effective MAC throughput exceeds the userdefined thresholds.
Colour per Peak RLC Throughput
After selecting the bearer, Atoll reads the corresponding RLC peak throughput ( TP DL I
). This is the highest
P RLC HSDPABearer
throughput that the bearer can provide on each pixel. Then, it determines the peak RLC throughput provided by the serving
DL
cell, c, in the downlink, TP P RLC c .

DL
DL
For an HSDPA user, we have: TP P RLC = TP P RLC c

For MC-HSDPA and DB-MC-HSDPA users, the peak RLC throughput provided to the user is calculated as follows:
DL
TP P RLC =
DL
TP P RLC c
c Serving cell
DL
RLC peak throughput ( TP P RLC ). For each layer, area is covered if the peak RLC throughput can be provided.
Colour per Effective RLC Throughput

DL
Atoll displays the Effective RLC throughput ( TP E RLC ) provided on each pixel. The Effective RLC throughput is calculated as
follows:
DL
TP P RLC
DL
TP E RLC = ----------------TTI
DL
Effective RLC throughput ( TP E RLC ). For each layer, area is covered if the Effective RLC throughput exceeds the user-defined
thresholds.
Colour per Average Effective RLC Throughput

DL
Atoll displays the average effective RLC throughput ( TP Av E RLC ) provided on each pixel.
For an HSDPA user, we have:
DL

DL
TP Av E RLC = -----------------------------------------------------------------------TTI
For MC-HSDPA and DB-MC-HSDPA users, we have:
DL
DL
Serving cells
TP Av E RLC = c---------------------------------------------------------------------------------------------------------TTI
Where,
DL
average effective RLC throughput ( TP Av E RLC ). For each layer, area is covered if the average effective RLC throughput
exceeds the user-defined thresholds.
Colour per Application Throughput
305


DL
Atoll displays the application throughput ( TP A ) provided on each pixel. The application throughput represents the net
throughput after deduction of coding (redundancy, overhead, addressing, etc.).
It is calculated as follows:
DL
DL
TP A = TP Av E RLC f TP Scaling TP Offset

Where:
DL
TP Av E RLC is the average effective RLC throughput.

DL
application throughput ( TPA ). For each layer, area is covered if the application throughput exceeds the user-defined
thresholds.
Fast Link Adaptation Modelling For Several Users
When you calculate the study with the following display options, Atoll considers several users per pixel and determines the
best HSDPA bearer that each user can obtain. In this case, the cell available HSDPA power is shared between HSDPA bearer
users. When the coverage prediction is not based on a simulation, the number of HSDPA bearer users is taken from the cell
properties. The displayed results of the coverage prediction will be an average result for one user.
For further information on the HSDPA bearer allocation process when there are several users, see "HSDPA Bearer Allocation
Process" on page 237 For further information on the fast link adaptation modelling, see "Fast Link Adaptation Modelling" on
page 240.
Colour per Effective MAC Throughput per User

DL
Atoll displays the average Effective MAC throughput per user ( TP E M AC Av ) provided on each pixel. The average Effective
MAC throughput per user is calculated as follows:
n HSDPA
DL
TP E M AC x
DL
x=1
TP E M AC Av = ---------------------------------------------Max n HSDPA c
c Serving cells x
Where,
n HSDPA c is the number of HSDPA bearer users within the cell, c.
DL
TP E M AC x is the Effective MAC throughput of each HSDPA bearer user. For further information on the calculation of the
Effective MAC throughput, see "Colour per Effective MAC Throughput" on page 304.
DL
average Effective MAC throughput per user ( TP E M AC Av ). For each layer, area is covered if the average Effective MAC
throughput per user exceeds the user-defined thresholds.
Colour per RLC Throughput per User

DL
Atoll displays the average effective RLC throughput per user ( TP E R LC Av ) provided on each pixel. The average effective RLC
throughput per user is calculated as follows:
n HSDPA
DL
DL
TP E R LC x
x=1
TP E R LC Av = ----------------------------------------n HSDPA
306

AT332_TRR_E0
Where,
n HSDPA is the number of HSDPA bearer users within the cell.
DL
TP E R LC x is the Effective RLC throughput of each HSDPA bearer user. For further information on the calculation of the
Effective RLC throughput, see "Colour per Effective RLC Throughput" on page 305.
DL
average effective RLC throughput per user ( TP E R LC Av ). For each layer, area is covered if the average effective RLC
throughput per user exceeds the user-defined thresholds.
Colour per ApplicationThroughput per User

DL
Atoll displays the average application throughput per user ( TPA Av ) provided on each pixel. The average application
throughput per user is calculated as follows:
n HSDPA
DL
TP A x
DL
x=1
TP A Av = ---------------------------------n HSDPA
Where,
n HSDPA is the number of HSDPA bearer users within the cell.
DL
TP A x is the application throughput of each HSDPA bearer user. For further information on the calculation of the
application throughput, see "Colour per Application Throughput" on page 305.
DL
average application throughput per user ( TP A Av ). For each layer, area is covered if the average application throughput per
user exceeds the user-defined thresholds.
Probability of Having a Certain Peak RLC Throughput
This result can be obtained only if you have selected an HSDPA radio bearer in the Condition tab.
Colour per Cell Edge Coverage Probability
Atoll shows areas where the selected HSDPA radio bearer is available with different cell edge coverage probabilities. Coverage
consists of several independent layers that can be displayed and hidden on the map. There is a layer per cell edge coverage
probability defined in the Display tab. For each layer, area is covered if the selected HSDPA radio bearer is available.
4.4.3.6 HSUPA Prediction Study

A dedicated HSUPA study is available with different calculation and display options. Atoll determines on each pixel the best
HSUPA bearer that can be obtained; it can consider either a single HSPA user or several ones on each pixel. For further
information on the HSUPA bearer selection, see "HSUPA Bearer Allocation Process" on page 259. By calculating this study with
suitable display options, it is possible:
To analyse the power required by the selected terminal,

To analyse the required E-DPDCH quality,
To analyse peak and effective throughputs.
We assume that each pixel on the map corresponds to one or several users with HSUPA capable terminal, mobility and HSUPA
service. You can make the coverage prediction for a specific carrier, for all carriers, for a specific layer, or for all layers. The
user does not create any interference.
Note that the HSUPA service area is limited by the pilot quality and the A-DPCH-EDPCCH quality.
4.4.3.6.1

Parameters used as input for the HSUPA prediction study are:
The cell UL load factor,

The cell UL reuse factor,
The cell UL load factor due to HSUPA,
The maximum cell UL load factor,
The number of HSUPA bearer users within the cell if the study is calculated for several users.
307

These parameters can be either simulation outputs, or user-defined cell inputs. In the last case, When no value is defined in
the Cells table, Atoll uses the following default values:
4.4.3.6.2
Uplink load factor = 50%

Uplink reuse factor = 1
Uplink load factor due to HSUPA = 0%
Maximum uplink load factor = 75%
Number of HSUPA beare users = 1
Calculation Options
Atoll can calculate the HSUPA coverage prediction in one of two ways:
HSUPA resources can be dedictated to a single user: On each pixel, the user is processed as if he is the only user in
the cell i.e he will use the entire remaining load after allocating capacity to all R99 users.
HSUPA resources can be shared by HSUPA users defined or calculated per cell: Atoll considers several HSUPA bearer
users per pixel. After allocating capacity to all R99 users, the remaining load of the cell will be shared equally between
all the HSUPA bearer users. When the coverage prediction is not based on a simulation, the number of HSUPA bearer
users is taken from the cell properties. The displayed results of the coverage prediction will be an average result for
one user.
4.4.3.6.3
Display Options
The following display options are available in the prediction property dialog box.
Colour per Required E-DPDCH Ec/Nt
Atoll displays on each pixel the E-DPDCH Ec/Nt required to obtain the selected HSUPA bearer. Coverage consists of several
independent layers that can be displayed and hidden on the map. There is a layer per threshold. For each layer, area is covered
req
Ec
Threshold .
if ------
Nt E DPDCH
Colour per Required Terminal Power
Atoll displays on each pixel the terminal power required to obtain the selected HSUPA bearer. The required terminal power
is calculated from the required E-DPDCH Ec/Nt. Coverage consists of several independent layers that can be displayed and
req
hidden on the map. There is a layer per threshold. For each layer, area is covered if P term Threshold .
Colour per Peak MAC Throughput
UL
Atoll displays the Peak MAC throughput ( TP P M AC ) provided on each pixel. The Peak MAC throughput is calculated as follows:
UL
S block
UL
TP P M AC = -----------T TTI
Where,
UL
S block is the transport block size (in kbits) for the selected HSUPA bearer; it is defined for each HSUPA bearer in the HSUPA
Radio Bearers table.
T TTI is the duration of one TTI for the selected HSUPA bearer; it is defined for each HSUPA bearer in the HSUPA Radio Bearers
table.
UL
Peak MAC throughput ( TP P M AC ). For each layer, area is covered if the Peak MAC throughput exceeds the user-defined
thresholds.
Colour per Peak RLC Throughput
After selecting the HSUPA bearer, Atoll reads the corresponding RLC peak throughput. This is the highest throughput that the
selected HSUPA bearer can provide on each pixel.
UL
RLC peak throughput ( TP P RLC ). For each layer, area is covered if the peak RLC throughput can be provided.
308

AT332_TRR_E0
Colour per Minimum Effective RLC Throughput

UL
Atoll displays the minimum effective RLC throughput ( TP Min E RLC ) provided on each pixel. The minimum effective RLC
throughput corresponds to the RLC throughput obtained for a given BLER and the maximum number of retransmissions. It is
UL

UL
TP Min E RLC = ---------------------------------------------------------------N Rtx
Where,
BLER HSUPA is the residual BLER for the selected uplink transmission format (HSUPA bearer with N Rtx retransmissions). It is
read in the quality graph defined for the quartet reception equipment-selected bearer-number of retransmissions-mobility
(HSUPA Quality Graphs tab in the Reception equipment properties). This graph describes the variation of BLER as a function
of the measured quality (E-DPDCH Ec/Nt). Knowing the E-DPDCH Ec/Nt, Atoll finds the corresponding BLER.
N Rtx is the maximum number of retransmissions for the selected HSUPA bearer. This figure is read in the HSUPA Bearer
Selection table.
UL
minimum effective RLC throughput ( TP Min E RLC ). For each layer, area is covered if the minimum effective RLC throughput
Colour per Average Effective RLC Throughput
When HARQ (Hybrid Automatic Repeat Request) is used, the required average number of retransmissions is smaller and the
UL
Effective RLC throughput is an average effective RLC throughput ( TP Av E RL C ). This is the RLC throughput obtained for a given
BLER and the average number of retransmissions. It is calculated as follows:
UL

UL
TP Av E RL C = --------------------------------------------------------------- N Rtx av
BLER HSUPA is the residual BLER for the selected uplink transmission format (HSUPA bearer with N Rtx retransmissions). It is
read in the quality graph defined for the quartet reception equipment-selected bearer-number of retransmissions-mobility
(HSUPA Quality Graphs tab in the Reception equipment properties). This graph describes the variation of BLER as a function
of the measured quality (E-DPDCH Ec/Nt). Knowing the E-DPDCH Ec/Nt, Atoll finds the corresponding BLER.
The average number of retransmissions ( N Rtx av ) is determined from early termination probabilities defined for the selected
HSUPA bearer (in the HSUPA Bearer Selection table). The Early Termination Probability graph shows the probability of early
termination ( p ) as a function of the number of retransmissions ( N Rtx ). Atoll calculates the average number of
retransmissions ( N Rtx av ) as follows:
N Rtx
max
p N Rtx p N Rtx 1 N Rtx
=1
Rtx
N Rtx av = ----------------------------------------------------------------------------------------------p N Rtx max
UL
average effective RLC throughput ( TP Av E RL C ). For each layer, area is covered if the minimum effective RLC throughput
Colour per Application Throughput
UL
Atoll displays the application throughput ( TPA ) provided on each pixel. The application throughput represents the net
throughput after deduction of coding (redundancy, overhead, addressing, etc.). This one is calculated as follows:
UL
UL
TP A M b = TP Min E RLC f TP Scaling TP Offset

Where:
f TP Scaling and TP Offset respectively represent the scaling factor between the application throughput and the minimum RLC
(Radio Link Control) throughput and the throughput offset. These two parameters model the header information and other
309

UL
application throughput ( TP A ). For each layer, area is covered if the application throughput exceeds the user-defined
thresholds.
Colour per Average Application Throughput
UL
Atoll displays the average application throughput ( TP Av A ) provided on each pixel. It is calculated as follows:
UL
UL
TP Av A M b = TP Av E RL C f TP Scaling TP Offset
Where:
f TP Scaling and TP Offset respectively represent the scaling factor between the average application throughput and the
average RLC (Radio Link Control) throughput and the throughput offset. These two parameters model the header information
and other supplementary data that does not appear at the application level. They are defined in the service properties.
UL
average application throughput ( TP Av A ). For each layer, area is covered if the average application throughput exceeds the
user-defined thresholds.

Atoll permits the automatic allocation of intra-technology neighbours in the current network. Two allocation algorithms are
available, one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours.
The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all the
cells of TBC transmitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfil following conditions:
They are active,

Only TBA cells may be assigned neighbours.

zone.
"Neighbour Allocation for All Transmitters" on page 310.

"Neighbour Allocation for a Group of Transmitters or One Transmitter" on page 314.
"Importance Calculation" on page 314.

We assume that we have a reference, cell A, and a candidate neighbour, cell B. When the automatic neighbour allocation
starts, Atoll checks the following conditions:
The distance between both cells must be less than the user-definable maximum inter-site distance. If the distance
between the reference cell and the candidate neighbour is greater than this value, then the candidate neighbour is
discarded.
"Calculation of the Inter-Transmitter Distance" on page 317. Otherwise, it takes the real distance.
310

AT332_TRR_E0
[Neighbours]

Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose
one or more carriers. Atoll will allocate neighbours only to the cells using the selected carriers.
For inter-carrier neighbours, in addition to the carrier(s) on which you want to run the allocation, you can select
the carrier(s) of potential neighbours.
Force co-site cells as neighbours: This option enables you to force cells located on the reference cell site in the
candidate neighbour list. This constraint can be weighted among the others and ranks the neighbours through the
importance field (see after).
Force adjacent cells as neighbours (only for intra-carrier neighbours): This option enables you to force cells
geographically adjacent to the reference cell in the candidate neighbour list. This constraint can be weighted
among the others and ranks the neighbours through the importance field (see below).
Let CellA be a candidate neighbour cell of CellB. CellA is considered adjacent to CellB if
there exists at least one pixel in the CellB Best Server coverage area where CellA is Best
Server (if several cells have the same best server value) or CellA is the second best server
that enters the Active Set (respecting the HO margin of the allocation).
When Force adjacent cells as neighbours is selected, adjacent cells are sorted and listed
from the most adjacent to the least adjacent, depending on the above criterion.
Adjacency is relative to the number of pixels satisfying the criterion.
Force adjacent layers as neighbours: If selected, Atoll adds all the cells adjacent across network layers to the
reference cell to the candidate neighbour list. The weight of this constraint is always the average of the Min and
Max values defined for the adjacency factor. This weight is used to calculate the rank of each neighbour and its
importance. Cells are considered adjacent across layers if they belong to different layers and have a coverage
overlap of at least one pixel.
reference cell is a candidate neighbour of another cell, this one will be considered as candidate neighbour of the
reference cell.
If the neighbours list of a cell is full, the reference cell will not be added as a neighbour
of that cell and that cell will be removed from the reference cells neighbours list. You can
force Atoll to keep that cell in the reference cells neighbours list by adding the following
option in the Atoll.ini file:
[Neighbours]
you may force/forbid a cell to be candidate neighbour of the reference cell.
311

Intra-carrier neighbours: intra-carrier handover is a soft handover.
The reference cell A and the candidate cell B are using the carrier c1 (c1 is the selected carrier on which you run the allocation).
SA is the area where the cell A is the best serving cell. It means that the cell A is the first one in the active set.
The pilot signal level received from A is greater than the minimum pilot signal level (min RSCP).
The best server indicator of A ( I BS A ) exceeds the minimum pilot quality (min Ec/I0).
I BS A is the highest one.
For information on the best server indicator calculation, see "Best Serving Cell Determination in Monte Carlo
Simulations - Old Method" on page 281.
SB is the area where the cell B can enter the active set.
The pilot signal received from the cell B is greater than the minimum pilot signal level (min RSCP).
The pilot quality from B is greater than the pilot quality from A minus the Ec/I0 margin (AS Threshold).
Figure 4.16: Overlapping Zone for Intra-carrier Neighbours
Inter-carrier neighbours: inter-frequency handover is a hard handover, triggered in multi-carrier W-CDMA networks
for coverage reasons (1st case) and to balance the load between carriers (2nd case).
The reference cell A is using the carrier c1 (c1 is the carrier selected in Source) and the candidate cell B is using the carrier c2
(c2 is the carrier selected in Destination).
SA is the area where the reference cell A is either the best serving cell among all cells using c1 (1st case) or a cell that can enter
the active set of a user connected to c1 (2nd case).
1st case: The cell A is the best serving cell among all cells using c1 but its pilot quality starts significantly decreasing.
The best server indicator of A ( I BS A ) is the highest one.
I BS A is lower than the minimum pilot quality (min Ec/I0) plus the handover margin.
2nd case: The cell A is not the best serving cell among all cells using c1 but it can enter the active set of a user
connected to c1.
312
The best server indicator of A ( I BS A ) exceeds the minimum pilot quality (min Ec/I0).

AT332_TRR_E0
I BS A is not the highest one. It is strictly lower than the best server indicator of the best serving cell and
greater than the best server indicator of the best serving cell minus the handover margin.
SB is the area where the cell B is the best serving cell among all cells using c2.
The pilot signal level received from B is greater than the minimum pilot signal level (min RSCP).
The best server indicator of B ( I BS B ) exceeds the minimum pilot quality (min Ec/I0).
I BS B is the highest one.
Figure 4.17: Overlapping Zone for Inter-carrier Neighbours - 1st case
Figure 4.18: Overlapping Zone for Inter-carrier Neighbours - 1st Case

For information on the best server indicator calculation, see "Best Serving Cell Determination in Monte Carlo Simulations - Old
Method" on page 281.
313

Two ways enable you to determine the I0 value:

1. Global Value: A percentage of the cell maximum power is considered. If the % of
maximum power is too low, i.e. if % Pmax P pilot , Atoll takes into account the
pilot power of the cell. Then, I0 represents the sum of values calculated for each
cell.
2. Defined per Cell: Atoll takes into account the total downlink power defined per
cell. I0 represents the sum of total transmitted powers.
For calculating the overlapping coverage areas, Atoll uses the service with the lowest
body loss, the terminal that has the highest difference between gain and losses or the
lowest noise figure when all terminals have the same (gain-losses) value, and the
shadowing margin calculated using the defined cell edge coverage probability, if the
option is selected. The service and terminal are selected such that the selection gives
the largest possible coverage areas for the cells.
SA SB
- 100 ) and compares this value to the % minimum covered area. If
Atoll calculates the percentage of covered area ( ----------------SA
this percentage is not exceeded, the candidate neighbour B is discarded.
The importance of neighbours.
For information on the importance calculation, see "Importance Calculation" on page 314.
Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason. Atoll lists
all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum
number of neighbours to be allocated to each transmitter is exceeded. If we consider the case for which there are 15
candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8. Among these 15
candidate neighbours, only 8 (having the highest importance values) will be allocated to the reference cell. Note that specific
maximum numbers of neighbours (maximum number of intra-carrier neighbours, maximum number of inter-carrier
neighbours) can be defined at the cell level (property dialog box or cell table). If defined there, this value is taken into account
instead of the default one available in the Neighbour Allocation dialog box.
In the Results part, Atoll only displays the cells for which it finds new neighbours. For these cells, it provides the list of
neighbours, the number of neighbours and the maximum number of neighbours allowed for each cell. In addition, it indicates
the importance (in %) of each neighbour and the allocation reason. Therefore, a neighbour may be marked as exceptional pair,
co-site, adjacent, coverage or symmetric. For neighbours accepted for co-site, adjacency and coverage reasons, Atoll displays
the percentage of area meeting the coverage conditions and the corresponding surface area (km2), the percentage of area
meeting the adjacency conditions and the corresponding surface area (km2). Finally, if cells have previous allocations in the
list, neighbours are marked as existing.

Transmitter
Atoll allocates neighbours to:
TBA cells,
Neighbours of TBA cells marked as exceptional pair, adjacent and symmetric,
Neighbours of TBA cells that satisfy coverage conditions.
4.5.3 Importance Calculation

Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason and the
distance, and to quantify the neighbour importance.
4.5.3.1 Importance of Intra-carrier Neighbours

314
Neighbourhood cause
When
Importance value
Existing neighbour

Existing importance

AT332_TRR_E0
Neighbourhood cause
When
Importance value
Exceptional pair
100 %
Co-site cell

selected
Adjacent layer
Only if the Force adjacent layers as neighbours option is

selected
Adjacent cell
Only if the Force adjacent cells as neighbours option is

selected

relationship
using the Importance Function (IF). The IF considers fourfactors for calculating the importance:
transmitter.
d Di = 1 ---------d max
of antennas. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter Distance"
on page 317.
[Neighbours]

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
30%
Min(A)
30%
Max(A)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

Neighbourhood cause
Importance Function

Coverage
10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
Adjacent layer
(Min(A)+Max(A))/2
45%
Adjacent cells
30%+30%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Co-site cells
60%+40%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Where:
315

overlapping.
causes.
4.5.3.2 Importance of Inter-carrier Neighbours

As indicated in the table below, the neighbour importance depends on the distance and on the neighbourhood cause; this
value varies between 0 to 100%.
Neighbourhood cause
When
Importance value
Existing neighbour
If the Delete existing neighbours option is not selected
Existing importance
Exceptional pair
If the Force exceptional pairs option is selected
100 %
Co-site cell
If the Force co-site cells as neighbours option is selected

If the % minimum covered area is exceeded
relationship
If the Force neighbour symmetry option is selected
using the Importance Function (IF). The IF considers threefactors for calculating the importance:
transmitter.
d Di = 1 ---------d max
on page 317.
[Neighbours]

The IF is user-definable using the Min importance and Max importance fields.
316
Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

AT332_TRR_E0
The IF evaluates importance as follows:

Co-site Neighbourhood
cause
IF

No
Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)}
10%+50%{10%(Di)+90%(O)}
Yes
Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+
Max(O)(O)/(Max(Di)+Max(O))}
60%+40%{1/7%(Di)+6/7%(O)}
Where
co-site neighbours and neighbours allocated based on coverage overlapping.
causes.
4.5.4 Appendices
4.5.4.1 Calculation of the Inter-Transmitter Distance
where x = 0.3% so that the maximum D variation does not exceed 1%.

4.6 Primary Scrambling Code Allocation

Downlink primary scrambling codes enable you to distinguish cells from one another (cell identification).
By default, there are 512 primary scrambling codes numbered (0...511).
The cells to which Atoll allocates scrambling codes are referred to as the TBA cells (cells to be allocated). TBA cells fulfil
following conditions:
They are active,

317

They belong to the folder on which allocation has been executed. This folder can be either the Transmitters folder
or a group of transmitters or a single transmitter.
zone.
4.6.1 Automatic Allocation Description

4.6.1.1 Options and Constraints
The scrambling code allocation algorithm can take into account following constraints and options:
Neighbourhood between cells,
You may consider:
First order neighbours: The neighbours of TBA cells listed in the Intra-technology neighbours table,
Second order neighbours: The neighbours of neighbours,
Third order neighbours: The neighbours neighbours neighbours.
In the context of the primary scrambling code allocation, the term "neighbours" refers
to intra-carrier neighbours.
Atoll can take into account inter-technology neighbour relations as constraints to
allocate different scrambling codes to the UMTS neighbours of a GSM transmitter. In
order to consider inter-technology neighbour relations in the scrambling code
allocation, you must make the Transmitters folder of the GSM .atl document
accessible in the UMTS .atl document. For information on making links between GSM
and UMTS .atl documents, see the User Manual.
Atoll considers symmetry relationship between a cell, its first order neighbours, its
second order neighbours and its third order neighbours.
Cells fulfilling a criterion on Ec/I0 (option Additional Overlapping Conditions),
Atoll reuses the intra-carrier neighbour allocation algorithm to determine the list of cells which cannot be allocated the same
scrambling code, and to calculate their importance.
For a reference cell A, Atoll considers all the cells B that can enter the active set on the area where the reference cell is
the best server (area where (Ec/I0)A exceeds the minimum Ec/I0 and is the highest one and (Ec/I0)B is within a Ec/I0 margin
of (Ec/I0)A).
318

AT332_TRR_E0
Atoll considers either a percentage of the cell maximum powers or the total downlink
power used by the cells in order to evaluate I0. In this case, I0 equals the sum of total
transmitted powers. When this parameter is not specified in the cell properties, Atoll
uses 50% of the maximum power.
Reuse distance is a constraint on the allocation of scrambling codes. A code cannot be

reused at a cell that is not at least as far away as the reuse distance from the cell
allocated with the particular code.
Scrambling code reuse distance can be defined at cell level. If this value is not defined,
then Atoll will use the default reuse distance defined in the Scrambling Code
Automatic Allocation dialog box.
Reuse distance,
Exceptional pairs,
Domains of scrambling codes,

When no domain is assigned to cells, Atoll considers the 512 primary scrambling codes
available.
The number of primary scrambling codes per cluster. In Atoll, we call "cluster", a group of scrambling codes as defined
in 3GPP specifications. 3GPP specifications define 64 clusters consisting of 8 scrambling codes (in this case, clusters
are numbererd from 0 to 63). However, you can define another value (e.g. if you set the number of codes per cluster
to 4, scrambling codes will be distributed in 128 clusters).
When the allocation is based on a Distributed strategy (Distributed per Cell or Distributed per Site), this parameter can
also be used to define the interval between the primary scrambling codes assigned to cells on a same site. The defined
interval is applied by adding the following lines in the Atoll.ini file:
[PSC]
ConstantStep = 1
For more information about setting options in the Atoll.ini file, see the Administrator Manual.
The carrier on which the allocation is run: It can be a given carrier or all of them. In this case, either Atoll independently
plans scrambling codes for the different carriers, or it allocates the same primary scrambling code to each carrier of a
transmitter if the option "Allocate carriers identically" is selected.
The possibility to use a maximum of codes from the defined domains (option "Use a Maximum of Codes"): Atoll will
try to spread the scrambling code spectrum the most.
The "Delete All Codes" option: When selecting this option, Atoll deletes all the current scrambling codes and carries
out a new scrambling code allocation. If not selected, the existing scrambling codes are kept.
In addition, it depends on the selected allocation strategy. Allocation strategies can be:
Clustered allocation: The purpose of this strategy is to choose for a group of mutually constrained cells, scrambling
codes among a minimum number of clusters. In this case, Atoll will preferentially allocate all the codes within the
same cluster.
Distributed per cell allocation: This strategy consists in using as many clusters as possible. Atoll will preferentially
allocate codes from different clusters.
One cluster per site allocation: This strategy allocates one cluster to each site, then, one code from the cluster to each
cell of each site. When all the clusters have been allocated and there are still sites remaining to be allocated, Atoll
reuses the clusters as far as possible at another site.
Distributed per site allocation: This strategy allocates a group of adjacent clusters to each site, then, one cluster to
each transmitter on the site according to its azimuth and finally, one code from the cluster to each cell of each
transmitter. The number of adjacent clusters per group depends on the number of transmitters per site you have in
your network; this information is required to start allocation based on this strategy. When all the groups of adjacent
clusters have been allocated and there are still sites remaining to be allocated, Atoll reuses the groups of adjacent
clusters as far as possible at another site.
In the Results table, Atoll only displays scrambling codes allocated to TBA cells.
319

4.6.1.2 Allocation Process

For each TBA cell, Atoll lists all cells which have constraints with the cell. They are referred to as near cells. The near cells of a
TBA cell may be:
Its neighbour cells: the neighbours listed in the Intra-technology neighbours table (options Existing neighbours and
"First Order"),
The neighbours of its neighbours (options Existing neighbours and Second Order),
The third order neighbours (options Existing neighbours and Third Order),
The cells that fulfil Ec/I0 condition (option Additional Overlapping Conditions),
The cells with distance from the TBA cell less than the reuse distance,
The cells that make exceptional pairs with the TBA cell.
Additional constraints are considered when:
The cell and its near cells are neighbours of a same GSM transmitter (only if the Transmitters folder of the GSM .atl
document is accessible in the UMTS .atl document),
The neighbour cells cannot share the same cluster (for the "Distributed per site" allocation strategy only).
These constraints have a certain weight taken into account to determine the TBA cell priority during the allocation process
and the cost of the scrambling code plan. During the allocation, Atoll tries to assign different scrambling codes to the TBA cell
and its near cells. If it respects all the constraints, the cost of the scrambling code plan is 0. When a cell has too many
constraints and there are not anymore scrambling codes available, Atoll breaks the constraint with the lowest cost so as to
generate the scrambling code plan with the lowest cost. For information on the cost generated by each constraint, see "Cell
Priority" on page 321.
4.6.1.2.1
Single Carrier Network

The allocation process depends on the selected strategy. Algorithm works as follows:
Strategies: Clustered and Distributed per Cell
Atoll processes TBA cells according to their priority. It allocates scrambling codes starting with the highest priority cell and its
near cells, and continuing with the lowest priority cells not allocated yet and their near cells. For information on calculating
cell priority, see "Cell Priority" on page 321.
Strategy: One Cluster per Site
All sites which have constraints with the studied site are referred to as near sites.
Atoll assigns a cluster to each site, starting with the highest priority site and its near sites, and continuing with the lowest
priority sites not allocated yet and their near sites. When all the clusters have been allocated and there are still sites remaining
to be allocated, Atoll reuses the clusters at another site. When the Reuse Distance option is selected, the algorithm reuses
the clusters as soon as the reuse distance is exceeded. Otherwise, when the option is not selected, the algorithm tries to assign
reused clusters as spaced out as possible.
Then, Atoll allocates a primary scrambling code from the cluster to each cell located on the sites (codes belong to the assigned
clusters). It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet
and their near cells.
For information on calculating site priority, see "Site Priority" on page 324. For information on calculating cell priority, see
"Cell Priority" on page 321.
Strategy: Distributed per Site
Atoll assigns a group of adjacent clusters to each site, starting with the highest priority site and its near sites, and continuing
with the lowest priority sites not allocated yet and their near sites. When all the groups of adjacent clusters have been
allocated and there are still sites remaining to be allocated, Atoll reuses the groups of adjacent clusters at another site. When
the Reuse Distance option is selected, the algorithm reuses the groups of adjacent clusters as soon as the reuse distance is
exceeded. Otherwise, when the option is not selected, the algorithm tries to assign reused groups of adjacent clusters as
spaced out as possible. Then, Atoll assigns each cluster of the group to each transmitter of the site according to the
transmitter azimuth and selected neighbourhood constraints (options "Neighbours in Other Clusters" and "Secondary
Neighbours in Other Clusters"). Then, Atoll allocates a primary scrambling code to each cell located on the transmitters (codes
belong to the assigned clusters). It starts with the highest priority cell and its near cells and goes on with the lowest priority
cells not allocated yet and their near cells.
320

AT332_TRR_E0
Determination of Groups of Adjacent Clusters

In order to determine the groups of adjacent clusters to be used, Atoll proceeds as follows: It defines theoretical groups of
adjacent clusters, independently of the defined domain, considering the 512 primary scrambling codes available and the
specified number of codes per cluster (if this one is set to 8, 64 clusters are supposed to be available). It starts the division in
group from the cluster 0 (hard coded) and takes into account the maximum number of transmitters per site user-specified in
order to determine the number of clusters in each group and then, the number of possible groups.
Let us assume that the number of codes per cluster is set to 8 and the maximum number of transmitters per site in the
network is 3. In this case, we have the following theoretical groups:
Group 1
Group 2
Group 3
Group 4
...
Group 21
Cluster 0
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Cluster 5
Cluster 6
Cluster 7
Cluster 8
Cluster 9
Cluster 10
Cluster 11
...
Cluster 61
Cluster 62
Cluster 63
If no domain is assigned to cells, Atoll can use all these groups for the allocation. On the other hand, if a domain is used, the
tool compares adjacent clusters really available in the assigned domain to the theoretical groups and only keeps adjacent
clusters mapping the theoretical groups.
Let us assume that we have a domain consisted of 12 clusters: clusters 1 to 8 and clusters 12 to 15.
Therefore, Atoll will be able to use the following groups of adjacent clusters:
Group 2 with cluster 3, 4 and 5,

Group 3 with cluster 6, 7 and 8,
Group 6 with cluster 12, 13 and 14.
The clusters 1, 2 and 15 will not be used.
If a domain does not contain any adjacent clusters, the user is warned through the 'Event Viewer'.
4.6.1.2.2
Multi-Carrier Network
In case you have a multi-carrier network and you run the scrambling code allocation on all the carriers, the allocation process
depends on the allocation strategy as detailed above and in addition, wether the option "Allocate Carriers Identically" is
selected or not.
When the option is not selected, algorithm works for each strategy, as explained above. On the other hand, when the option
is selected, allocation order changes. It is no longer based on the cell priority but depends on the transmitter priority. All
transmitters which have constraints with the studied transmitter will be referred to as near transmitters.
In case of a "Per cell" strategy (Clustered and Distributed per cell), Atoll starts scrambling code allocation with the highest
priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their
near transmitters. The same scrambling code is assigned to each cell of the transmitter.
In case of the "One cluster per site" strategy, Atoll assigns a cluster to each site and then, allocates a scrambling code to each
transmitter. It starts with the highest priority transmitter and its near transmitters and continues with the lowest priority
transmitters not allocated yet and their near transmitters. The same scrambling code is assigned to each cell of the
transmitter.
In case of the "Distributed per site" strategy, Atoll assigns a group of adjacent clusters to each site, then a cluster to each
transmitter and finally, allocates a scrambling code to each transmitter. It starts with the highest priority transmitter and its
near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. The same
scrambling code is assigned to each cell of the transmitter.
For information on calculating transmitter priority, see "Transmitter Priority" on page 324.
When cells, transmitters or sites have the same priority, processing is based on an
alphanumeric order.
4.6.1.3 Priority Determination

4.6.1.3.1
Cell Priority
Scrambling code allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities
assigned to cells depend upon how much constrained each cell is and the cost defined for each constraint. A cell without any
constraint has a default cost, C , equal to 0. The higher the cost on a cell, the higher the priority it has for the scrambling code
allocation process.
There are six criteria employed to determine the cell priority:
Scrambling Code Domain Criterion
321

The cost due to the domain constraint, C i Dom , depends on the number of scrambling codes available for the allocation.
The domain constraint is mandatory and cannot be broken.
When no domain is assigned to cells, 512 scrambling codes are available and we have:
C i Dom = 0
When domains of scrambling codes are assigned to cells, each unavailable scrambling code generates a cost. The higher the
number of codes available in the domain, the less will be the cost due to this criterion. The cost is given as:
C i Dom = 512 Number of scrambling codes in the domain
Distance Criterion
The constraint level of any cell i depends on the number of cells (j) present within a radius of "reuse distance" from its centre.
The total cost due to the distance constraint is given as:
Cj Dist i
C i Dist =
Each cell j within the reuse distance generates a cost given as:
C j Dist i = w d ij c dis tan ce
Where
w d ij is a weight depending on the distance between i and j. This weight is inversely proportional to the inter-cell distance.
For a reuse distance of 2000m, the weight for an inter-cell distance of 1500m is 0.25, the weight for co-site cells is 1 and the
weight for two cells spaced out 2100m apart is 0.
c dis tan ce is the cost of the distance constraint. This value can be defined in the Constraint Cost dialog box.
Exceptional Pair Criterion
The constraint level of any cell i depends on the number of exceptional pairs (j) for that cell. The total cost due to exceptional
pair constraint is given as:
C i EP =
cEP i j
j
Where
c EP is the cost of the exceptional pair constraint. This value can be defined in the Constraint Cost dialog box.
Neighbourhood Criterion
The constraint level of any cell i depends on the number of its neighbour cells j, the number of second order neighbours k and
the number of third order neighbours l.
Lets consider the following neighbour schema:
Figure 4.20: Neighbourhood Constraints

The total cost due to the neighbour constraint is given as:
Ci N =
Cj N1 i + Cj j N1 i + Ck N2 i + Ck k N2 i + Cl N3 i + Cl l N3 i
j
Each first order neighbour cell j generates a cost given as:

C j N1 i = I j c N1
Where
I j is the importance of the neighbour cell j.
c N1 is the cost of the first order neighbour constraint. This value can be defined in the Constraint Cost dialog box.
322

AT332_TRR_E0
Because two first order neighbours must not have the same scrambling code, Atoll considers the cost created by two first
order neighbours to be each other.
C j N1 i + C j N1 i
C j j N1 i = ----------------------------------------------------2
Each second order neighbour cell k generates a cost given as:
C k N2 i = Max ( C j N1 i C k N1 j , C j N1 i C k N1 j ) c N2
Where
c N2 is the cost of the second order neighbour constraint. This value can be defined in the Constraint Cost dialog box.
Because two second order neighbours must not have the same scrambling code, Atoll considers the cost created by two
second order neighbours to be each other.
C k N2 i + C k N2 i
C k k N2 i = ------------------------------------------------------2
Each third order neighbour cell l generates a cost given as:
C N1 i C k N1 j C l N1 k C j N1 i C k N1 j C l N1 k
C l N3 i = Max j
c N3
C j N1 i C k N1 j C l N1 k C j N1 i C k N1 j C l N1 k
Where
c N3 is the cost of the third order neighbour constraint. This value can be defined in the Constraint Cost dialog box.
Because two third order neighbours must not have the same scrambling code, Atoll considers the cost created by two third
C l N3 i + C l N3 i
C l l N3 i = ----------------------------------------------------2
Atoll considers the highest cost of both links when a neighbour relation is symmetric and
the importance value is different.
In this case, we have:

C j N1 i = Max I i j I j i c N1
And
C k N2 i = Max (C j N1 i C k N1 j ,C j N1 k C i N1 j ) c N2
GSM Neighbour Criterion
This criterion is considered when the co-planning mode is activated (i.e. the Transmitters folder of the GSM .atl document is
made accessible in the UMTS .atl document) and inter-technology neighbours have been allocated. If the cell i is neighbour of
a GSM transmitter, the cell constraint level depends on how many cells j are neighbours of the same GSM transmitter. The
total cost due to GSM neighbour constraint is given as:
C i N 2G =
cN2G j Tx2G
j
Where
cN
2G
is the cost of the GSM neighbour constraint. This value can be defined in the Constraint Cost dialog box.
Cluster Criterion
When the "Distributed per Site" allocation strategy is used, you can consider additional constraints on allocated clusters (one
cell, its first order neighbours and its second order neighbours must be assigned scrambling codes from different clusters). In
this case, the constraint level of any cell i depends on the number of first and second order neighbours, j and k. The total cost
due to the cluster constraint is given as:
C i Cluster =
Cj N1 i cCluster + Ck N2 i cCluster
j
323

Where
c Cluster is the cost of the cluster constraint. This value can be defined in the Constraint Cost dialog box.
Therefore, the total cost due to constraints on any cell i is defined as:
C i = C i Dom + C i U
With
C i U = C i Dist + C i EP + C i N + C i N 2G + C i Cluster
4.6.1.3.2
Transmitter Priority
In case you have a multi-carrier network and you run scrambling code allocation on "all" the carriers with the option "allocate
carriers identically", algorithm in Atoll allots priorities to transmitters. Priorities assigned to transmitters depend on how
much constrained each transmitter is and the cost defined for each constraint. The higher the cost on a transmitter, the higher
the priority it has for the scrambling code allocation process.
Let us consider a transmitter Tx with two cells using carriers 0 and 1. The cost due to constraints on the transmitter is given as:
C Tx = C Tx Dom + C Tx U
With C Tx U =
Max C U and C Dom = 512 Number of scrambling codes in the domain

Tx
i Tx i
Here, the domain available for the transmitter is the intersection of domains assigned to cells of the transmitter. The domain
constraint is mandatory and cannot be broken.
4.6.1.3.3
Site Priority
In case of "Per Site" allocation strategies (One cluster per site and Distributed per site), algorithm in Atoll allots priorities to
sites. Priorities assigned to sites depend on how much constrained each site is and the cost defined for each constraint. The
higher the cost on a site, the higher the priority it has for the scrambling code allocation process.
Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. The cost due to constraints
on the site is given as:
C S = C S U + C S Dom
With C S U =

S
Tx S Tx
Here, the domain considered for the site is the intersection of domains available for transmitters of the site. The domain
4.6.2 Allocation Examples

4.6.2.1 Allocation Strategies and Use a Maximum of Codes
In order to understand the differences between the different allocation strategies and the behaviour of algorithm when using
a maximum of codes or not, let us consider the following sample scenario:
324

AT332_TRR_E0
Figure 4.21: Primary Scrambling Codes Allocation

Let Site0, Site1, Site2 and Site3 be four sites with 3 cells using carrier 0 whom scrambling codes have to be allocated out of
three clusters consisted of 8 primary scrambling codes. This implies that the domain of scrambling codes for the four sites is
from 0 to 23 (cluster 0 to cluster 2). The reuse distance is supposed to be less than the inter-site distance. Only co-site
neighbours exist.
The following section lists the results of each combination of options with explanation where necessary.
4.6.2.1.1
Strategy: Clustered
Since the restrictions of neighbourhood only apply to co-sites with the same importance and sites distances are greater than
reuse distances, every cell has the same priority. Then, scrambling code allocation to cells is performed in an alphanumeric
order.
Without Use a Maximum of Codes
With Use a Maximum of Codes
Atoll starts allocating the codes from the start of cluster 0 at As it is possible to use a maximum of codes, Atoll starts
each site.
allocation at the start of a different cluster at each site. When
a cluster is reused, and there are non allocated codes left in
the cluster, Atoll first allocates those codes before reusing
the already used ones.
325

4.6.2.1.2
Strategy: Distributed
reuse distances, every cell has the same priority. Then, scrambling code allocation to cells is performed in an alphanumeric
order.
Atoll allocates codes from different clusters to each cell of Atoll allocates codes from different clusters to each sites
the same site. Under given constraints of neighbourhood and cells. As it is possible to use a maximum of codes, Atoll
reuse distance, same codes can be allocated to each sites allocates the codes so that there is least repetition of codes.
cells.
4.6.2.1.3
Strategy: One Cluster per Site

reuse distances, every site has the same priority. Then, cluster allocation to sites is performed in an alphanumeric order.
In this strategy, a cluster of codes is limited to be used at just When it is possible to use a maximum of codes, Atoll can
one site at a time unless all codes and clusters have been allocate different codes from a reused cluster at another site.
allocated and there are still sites remaining to be allocated. In
this case Atoll reuses the clusters as far as possible at another
site.
4.6.2.1.4

reuse distances, every site has the same priority. Then, the group of adjacent clusters allocation to sites is performed in an
alphanumeric order.
326

AT332_TRR_E0
In this strategy, a group of adjacent clusters is limited to be When it is possible to use a maximum of codes, Atoll can
used at just one site at a time unless all codes and groups of allocate different codes from a reused group of adjacent
adjacent clusters have been allocated and there are still sites cluster at another site.
remaining to be allocated. In this case (here only one group
of adjacent clusters (clusters 0, 1 and 2) is available), Atoll
reuses the group at another site.
4.6.2.2 Allocate Carriers Identically

In order to understand the behaviour of algorithm when using the option "Allocate Carriers Identically" or not, let us consider
the following sample scenario:
Let Site0, Site1, Site2 and Site3 be four sites with 3 cells using carrier 0 and 3 cells using carrier 1. Scrambling codes have to
be allocated out of 3 clusters consisted of 8 primary scrambling codes. This implies that the domain of scrambling codes for
the five sites is from 0 to 23 (cluster 0 to cluster 2). The reuse distance is supposed to be less than the inter-site distance. Only
co-site neighbours exist. Allocation algorithm will be based on the "One Cluster per Site" strategy and the option "Use a
Maximum of Codes" is selected.
Without Allocate Carriers Identically
With Allocate Carriers Identically
Atoll allocates one cluster at each site as detailed in the In this case, Atoll allocates one cluster at each site and then,
previous section. Then, it allocates a code from the cluster to one code to each transmitter so as to use a maximum of
each cell of the site so as to use a maximum of codes.
codes. Then, the same code is given to each cell of the
transmitter.
In both cases (with and without Allocate Carriers Identically), every site has the same priority. Then, cluster allocation to sites
is performed in an alphanumeric order.
327

4.7 Automatic GSM-UMTS Neighbour Allocation

4.7.1 Overview
You can automatically calculate and allocate neighbours between GSM and UMTS networks. In Atoll, it is called intertechnology neighbour allocation.
Inter-technology handover is used in two cases:
When the UMTS coverage is not continuous. In this case, the UMTS coverage is extended by UMTS-GSM handover into
the GSM network,
And in order to balance traffic and service distribution between both networks.
Note that the automatic inter-technology neighbour allocation algorithm takes into account both cases.
In order to be able to use the inter-technology neighbour allocation algorithm, you must have:
An .atl document containing the GSM network, GSM.atl, and another one describing the UMTS network, UMTS.atl,
An existing link on the Transmitters folder of GSM.atl into UMTS.atl.
The external neighbour allocation algorithm takes into account all the GSM TBC transmitters. It means that all the TBC
transmitters of GSM.atl are potential neighbours. The cells to be allocated will be called TBA cells which, being cells of
UMTS.atl, satisfy following conditions:
They are active,

They satisfy the filter criteria applied to Transmitters folder,
They belong to the folder for which allocation has been executed. This folder can be either the Transmitters folder or
a group of transmitters subfolder.
Only UMTS TBA cells may be assigned neighbours.

The allocation algorithm takes into account criteria listed below:
The inter-transmitter distance,

The maximum number of neighbours fixed,
Allocation options,
The selected allocation strategy,
Two allocation strategies are available: the first one is based on distance and the second one on coverage overlapping.
We assume we have a UMTS reference cell, A, and a GSM candidate neighbour, transmitter B.
4.7.2.1 Algorithm Based on Distance

When the automatic allocation starts, Atoll checks the following conditions:
The distance between the UMTS reference cell and the GSM neighbour must be less than the user-definable maximum
inter-site distance. If the distance between the UMTS reference cell and the GSM neighbour is greater than this value,
then the candidate neighbour is discarded.
Atoll calculates the effective distance, which corresponds to the real inter-transmitter distance weighted by the
azimuths of antennas. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter
The calculation options,
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or more
carriers. Atoll will allocate neighbours to cells using the selected carriers.
Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site as the
reference UMTS cell in the candidate neighbour list. This option is automatically selected.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you may
force/forbid a GSM transmitter to be candidate neighbour of the reference UMTS cell.
Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current neighbours and
carries out a new neighbour allocation. If not selected, existing neighbours are kept.
Importance values are used by the allocation algorithm to rank the neighbours. Atoll lists all neighbours and sorts them by
importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be
328

AT332_TRR_E0
allocated to each cell is exceeded. If we consider the case for which there are 15 candidate neighbours and the maximum
number of neighbours to be allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8 (having the
highest importance values) will be allocated to the reference cell. Note that the maximum number of inter-technology
neighbours can be defined at the cell level (property dialog box or cell table). If defined there, this value is taken into account
Neighbourhood cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
Co-site transmitter
100 %

fulfils distance conditions
If the maximum distance is not exceeded
d1 ---------d max
Where d is the effective distance between the UMTS reference cell and the GSM neighbour and d max is the maximum intersite distance.For information on the effective distance calculation, see "Calculation of the Inter-Transmitter Distance" on
page 317.
In the Results part, Atoll provides the list of neighbours, the number of neighbours and the maximum number of neighbours
allowed for each cell. In addition, it indicates the importance (in %) of each neighbour and the allocation reason. Therefore, a
neighbour may be marked as exceptional pair, co-site, or distance. For neighbours accepted for distance reasons, Atoll
displays the distance from the reference cell (m). Finally, if cells have previous allocations in the list, neighbours are marked
as existing.
4.7.2.2 Algorithm Based on Coverage Overlapping

The distance between the UMTS reference cell and the GSM neighbour must be less than the user-definable maximum
inter-site distance. If the distance between the UMTS reference cell and the GSM neighbour is greater than this value,
then the candidate neighbour is discarded.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or more
carriers. Atoll will allocate neighbours to cells using the selected carriers.
reference UMTS cell in the candidate neighbour list. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference UMTS cell.
There must be an overlapping zone ( S A S B ) with a given cell edge coverage probability.
Four different cases may be considered for SA:
1st case: SA is the area where the cell A is the best serving cell of the UMTS network.
The pilot signal received from A is greater than the minimum pilot signal level,
The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0) and is the highest one.
In this case, the Ec/I0 margin must be equal to 0dB and the max Ec/I0 option disabled.
2nd case: SA represents the area where the pilot quality from the cell A strats decreasing but the cell A is still the
best serving cell of the UMTS network.
The Ec/I0 margin must be equal to 0dB, the max Ec/I0 option selected and a maximum Ec/I0 user-defined.
The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/I0.
The pilot quality from A is the highest one.
3rd case: SA represents the area where the cell A is not the best serving cell but can enter the active set.
Here, the Ec/I0 margin has to be different from 0dB and the max Ec/I0 option disabled.
329

The pilot quality from A is within a margin from the best Ec/I0, where the best Ec/I0 exceeds the minimum Ec/
I0.
4th case: SA represents the area where:
The pilot quality from A is within a margin from the best Ec/I0 (where the best Ec/I0 exceeds the minimum Ec/
I0) and lower than the maximum Ec/I0.
In this case, the margin must be different from 0dB, the max Ec/I0 option selected and a maximum Ec/I0 userdefined.
Two different cases may be considered for SB:
1st case: SB is the area where the cell B is the best serving cell of the GSM network.
In this case, the margin must be set to 0dB.
The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is the
highest one.
2nd case: The margin is different from 0dB and SB is the area where:
The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is
within a margin from the best BCCH signal level.
SA SB
Atoll calculates the percentage of covered area ( ------------------ 100 ) and compares this value to the % minimum covered area. If
SA
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to % of covered area.
When the automatic allocation is based on coverage overlapping, we recommend you to
perform two successive automatic allocations:
A first allocation in order to find handovers due to non-continuous UMTS coverage.

In this case, you have to select the max Ec/I0 option and define a high enough value.
A second allocation in order to complete the previous list with handovers motivated
for reasons of traffic and service distribution. Here, the max Ec/I0 option must be
disabled.
Importance values are used by the allocation algorithm to rank the neighbours according to the distance and the allocation
reason. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list
if the maximum number of neighbours to be allocated to each cell is exceeded. If we consider the case for which there are 15
candidate neighbours, only 8 (having the highest importance values) will be allocated to the reference cell. Note that the
maximum number of inter-technology neighbours can be defined at the cell level (property dialog box or cell table). If defined
there, this value is taken into account instead of the default one available in the Neighbour Allocation dialog box.
Neighbourhood reason
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
Co-site transmitter

using the Importance Function (IF). The IF considers threefactors for calculating the importance:
transmitter.
d
Di = 1 ----------d max
d is the distance between the UMTS reference cell and the GSM neighbour.
d max is the maximum inter-site distance defined.
330

AT332_TRR_E0

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

cause
IF

No
10%+50%{10%(Di)+90%(O)}
Yes
60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.
neighbour may be marked as exceptional pair, co-site or coverage. For neighbours accepted for co-site and coverage reasons,
Atoll displays the percentage of area meeting the coverage conditions and the corresponding surface area (km2). Finally, if
cells have previous allocations in the list, neighbours are marked as existing.
No prediction study is needed to perform an automatic neighbour allocation. When

starting an automatic neighbour allocation, Atoll automatically calculates the path
loss matrices if not found.
A forbidden neighbour must not be listed as neighbour except if the neighbourhood
relationship already exists and the Delete existing neighbours option is unchecked
when you start the new allocation. In this case, Atoll displays a warning in the Event
viewer indicating that the constraint on the forbidden neighbour will be ignored by
algorithm because the neighbour already exists.
In the Results, Atoll displays only the cells for which it finds new neighbours.
Therefore, if a TBA cell has already reached its maximum number of neighbours
before starting the new allocation, it will not appear in the Results table.
4.7.2.3 Appendices
4.7.2.3.1
Delete Existing Neighbours Option

As explained above, Atoll keeps the existing inter-technology neighbours when the Delete existing neighbours option is not
checked. We assume that we have an existing allocation of inter-technology neighbours.
A new TBA cell i is created in UMTS.atl. Therefore, if you start a new allocation without selecting the Delete existing
neighbours option, Atoll determines the neighbour list of the cell i.
If you change some allocation criteria (e.g. increase the maximum number of neighbours or create a new GSM TBC
transmitter) and start a new allocation without selecting the Delete existing neighbours option, it examines the neighbour list
of TBA cells and checks allocation criteria if there is space in their neighbour lists. A new GSM TBC transmitter can enter the
TBA cell neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list.
331

332
Chapter 5
CDMA2000 Networks
"General Prediction Studies" on page 335
"Definitions and Formulas" on page 338
"Active Set Management" on page 356
"CDMA2000 Prediction Studies" on page 390
"PN Offset Allocation" on page 428
"Automatic GSM-CDMA Neighbour Allocation" on

page 436

334

Chapter 5: CDMA2000 Networks
AT332_TRR_E0
5 CDMA2000 Networks
This chapter describes all the calculations performed in Atoll CDMA2000 documents.
All the calculations are performed on TBC (to be calculated) transmitters. For the definition of TBC transmitters please refer
to "Path Loss Matrices" on page 104.
5.1 General Prediction Studies

5.1.1 Calculation Criteria
Three criteria can be studied in point analysis (Profile tab) and in common coverage studies. Study criteria are detailed in the
table below:
Study criteria
Formulas
Signal level ( P rec ) in dBm

P rec ic = EIRP ic L path M Shadowing model L Indoor + G term L term
Path loss ( L path ) in dBm

Total losses ( L total ) in dBm
Tx
L total = L path + L Tx + L term + L indoor + M Shadowing model G Tx + G term
where,
ic is a carrier rank,
L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model,
L ant
Tx
M Shadowing model is the shadowing margin. This parameter is taken into account when the option Shadowing taken into
account is selected,
L term are the receiver losses,
G term is the receiver antenna gain,
G Tx is the transmitter antenna gain,
L Tx is the transmitter loss ( L Tx = L total DL ). For information on calculating transmitter loss, "UMTS, CDMA2000, TD-SCDMA,
WiMAX, and LTE Documents" on page 30.
For CDMA2000 1xRTT systems, EIRP ic = P pilot ic + G Tx L Tx (where, P pilot ic
is the cell pilot power).

For CDMA2000 1xEV-DO systems, EIRP ic = P max ic + G Tx L Tx (where P max ic
is the maximum cell power).

When you make the prediction, you can consider the best carrier of all bands or the
best carrier of a particular frequency band (Best (All Bands/Specific Band) option). In
this case, Atoll displays the best signal level received from a transmitter. Therefore, if
the network consists of 1xRTT and 1xEV-DO carriers, Atoll takes the highest power of
both cells for each transmitter (i.e. the highest value between the pilot power of the
1xRTT cell and the maximum power of the 1xEV-DO cell) to calculate the received
signal level.
Atoll considers that G term and L term equal zero.
335


5.1.2.1 Profile Tab
Atoll displays either the signal level received from the selected transmitter on a carrier ( P rec ic ), or the highest signal level
received from the selected transmitter on the best carrier.

Analysis provided in the Reception tab is based on path loss matrices. So, you can study reception from TBC transmitters for
which path loss matrices have been computed on their calculation areas.
For each transmitter, Atoll displays either the signal level received on a carrier, ( P rec ic ), or the highest signal level received
on the best carrier.
Reception bars are displayed in a decreasing signal level order. The maximum number of reception bars depends on the signal
level received from the best server. Only reception bars of transmitters whose signal level is within a 30 dB margin from the
best server can be displayed.
You can use a value other than 30 dB for the margin from the best server signal level,
for example a smaller value for improving the calculation speed. For more
information on defining a different value for this margin, see the Administrator
Manual.

For each TBC transmitter, Txi, Atoll determines the selected criterion on each bin inside the Txi calculation area. In fact, each
bin within the Txi calculation area is considered as a potential (fixed or mobile) receiver.
The study conditions in order to determine the service area of each TBC transmitter,
The display settings to select how to colour service areas.

Atoll uses parameters entered in the Condition tab of the coverage study property dialog box to predetermine areas where it
will display coverage.
We can distinguish three cases:
5.1.3.1.1
All Servers
Txi
Txi
Txi
Minimum threshold P rec ic or L total or L path Maximum threshold
5.1.3.1.2

Txi
Txi
Txi

And
Txi
Txj
P rec ic Best P rec ic M

ji

Best function: considers the highest value.
336

AT332_TRR_E0
5.1.3.1.3
Txi is the highest.
from Txi is either the highest or 2dB lower than the highest.
from Txi is 2dB higher than the signal levels from transmitters, which are 2nd best
servers.

Txi
Txi
Txi

And
Txi
nd
Txj
P rec ic 2 Best P rec ic M

ji

2nd Best function: considers the second highest value.
Txi is the second highest.
from Txi is either the second highest or 2dB lower than the second highest.
from Txi is 2dB higher than the signal levels from transmitters, which are 3rd best
servers.

5.1.3.2.1
Plot Resolution
Prediction plot resolution is independent of the matrix resolutions and can be defined on a per study basis. Prediction plots
are generated from multi-resolution path loss matrices using bilinear interpolation method (similar to the one used to
evaluate site altitude).
5.1.3.2.2
Display Types
It is possible to display the transmitter service area with colours depending on any transmitter attribute or other criteria such
as:
Atoll calculates signal level received from the transmitter on each bin of each transmitter service area. A bin of a service area
is coloured if the signal level is greater than or equal to the defined minimum thresholds (bin colour depends on signal level).
as transmitter service areas. Each layer shows the different signal levels available in the transmitter service area.
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where other service
areas overlap the studied one, Atoll chooses the highest value. A bin of a service area is coloured if the signal level is greater
than or equal to the defined thresholds (the bin colour depends on the signal level). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer
corresponds to an area where the signal level from the best server exceeds a defined minimum threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each bin of each transmitter service area. A bin of a service area is coloured
if path loss is greater than or equal to the defined minimum thresholds (bin colour depends on path loss). Coverage consists
of several independent layers whose visibility in the workspace can be managed. There are as many layers as service areas.
Each layer shows the different path loss levels in the transmitter service area.
337

Total Losses (dB)

Atoll calculates total losses from the transmitter on each bin of each transmitter service area. A bin of a service area is
coloured if total losses is greater than or equal to the defined minimum thresholds (bin colour depends on total losses).
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where other service
areas overlap the studied one, Atoll determines the best transmitter and evaluates path loss from the best transmitter. A bin
of a service area is coloured if the path loss is greater than or equal to the defined thresholds (bin colour depends on path
layers as defined thresholds. Each layer corresponds to an area where the path loss from the best server exceeds a defined
minimum threshold.
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where service areas
overlap the studied one, Atoll determines the best transmitter and evaluates total losses from the best transmitter. A bin of
a service area is coloured if the total losses is greater than or equal to the defined thresholds (bin colour depends on total
layers as defined thresholds. Each layer corresponds to an area where the total losses from the best server exceed a defined
minimum threshold.
Number of Servers
Atoll evaluates how many service areas cover a bin in order to determine the number of servers. The bin colour depends on
the number of servers. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the number of servers is greater than
or equal to a defined minimum threshold.
On each bin of each transmitter service area, the coverage corresponds to the pixels where the signal level from this
transmitter fulfils signal conditions defined in Conditions tab with different Cell edge coverage probabilities. There is one
coverage area per transmitter in the explorer.
On each bin of each transmitter service area, the coverage corresponds to the pixels where the best signal level received fulfils
signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability in the explorer.
5.2 Definitions and Formulas

5.2.1 Parameters Used for CDMA2000 1xRTT Modelling
5.2.1.1 Inputs
This table lists simulation and prediction inputs (calculation options, quality targets, active set management conditions, etc.)
338
Name
Value
Unit
Description
F ortho
Clutter parameter
None
F MUD
Tx
None
MUD factor
cn first
None
cn last
None
Last carrier number
cn
None
Carrier number step

AT332_TRR_E0
Name
Value
Unit
Description
cn cn first
- cn lower
ic = ------------------------ cn
ic
None

Q pilot
req
Q pilot txi ic + Q pilot
min
Q pilot txi ic
req
Min. Ec/I0 - Cell parameter
min
T_Drop - Cell parameter
Q pilot
Q pilot txi ic
req
req
min
min
Active set upper threshold

None (used to determine the best server in
the active set)
None
Active set lower threshold

(used to determine other members
of the active set)
Minimum Ec/I0 required from the

None cell to be the best server in the active
set
None

cell not to be rejected from the
active set
Variation of the minimum Ec/I0

None required from the cell to be the best
server in the active set
req
Delta Min. Ec/I0 - Mobility parameter
Q pilot
min
Delta T_Drop - Mobility parameter
None

required from the cell not to be
rejected from the active set
RSCP min Txi ic

None
Eb/Nt target for FCH channel on

downlink
None
Eb/Nt target for SCH channel on

downlink
None
Eb/Nt target for FCH channel on

uplink
None
Eb/Nt target for SCH channel on

uplink
Q pilot
DL
Q req FCH
FCH DL
E b
--- N t req
(Service, Terminal, Mobility) parameter

SCH DL
E b
--- N t req
DL
Q req SCH
(Service, Terminal, Mobility, SCH throughput multiple) parameter

UL
Q req FCH
FCH UL
E b
--- N t req
(Service, Terminal, Mobility) parameter

SCH UL
UL
Q req SCH
E b
--- N t req
(Service, Terminal, Mobility, SCH throughput multiple) parameter

Max
Site parameter
None

available for a site on uplink
N CE D L N I
Max
Site parameter
None

available for a site on downlink
N CE U L N I
Simulation result
None

site consumed by users on uplink
N CE D L N I
Simulation result
None

site consumed by users on downlink
Overhead
None

uplink
Overhead
None

downlink
N CE U L N I
N CE U L
N CE D L
339

Name
Value
Unit
Description
N CE U L
FCH
(Terminal, site equipment) parameter
None

for FCH on uplink
FCH
None

for FCH on downlink
N Codes txi ic
None
Maximum number of Walsh codes

available per cell (128)
N Codes txi ic
Simulation result
None
Number of Walsh codes used by the

cell
NF term
Terminal parameter
None
NF Tx

None
1.38 10-23
J/K
Boltzman constant
293
Ambient temperature
1.23 MHz
Hz
Spreading Bandwidth
N CE D L
Max
Tx DL
Cell parameter
Cell parameter
None
RF ic ic adj
Network parameter
interference
None


Tx UL
Tx m
ICP ic ic
i
Network parameter
and ic adj
None
network) and ic
UL
X max
DL
%Powermax
Maximum uplink load factor
Maximum percentage of used power
Tx UL
Tx
Term
Rc
bps
Chip rate
f rake efficiency
UL
Equipment parameter
DL
Terminal parameter
N0
N0
f rake efficiency
SCH
TPF DL
FCH
TPP DL
SCH
TPP DL
SCH
TPF UL
FCH
TP P UL
SCH
TP P UL
340
Tx DL
Simulation result
Terminal parameter
FCH
SCH
TP P DL TPF DL
Simulation result
Terminal parameter
FCH
SCH
TPP UL TPF UL

None
Downlink rake receiver efficiency

factor
SCH throughput factor (drawn

None following the SCH probabilities of the
service)
bps
Downlink FCH peak throughput
bps
Downlink SCH bit rate
SCH throughput factor (drawn

None following the SCH probabilities of the
service)
bps
Uplink FCH peak throughput
bps
Uplink SCH bit rate

AT332_TRR_E0
Name
Value
Unit
Description
W ----------------FCH
TP P DL
None
Downlink service processing gain on

FCH
W ----------------SCH
TP P DL
None
Downlink service processing gain on

SCH
W ----------------FCH
TP P UL
None
Uplink service processing gain on

FCH
W ----------------SCH
TP P UL
None

SCH
DL
Service parameter
None
Downlink activity factor on FCH
AF FCH
UL
Service parameter
None
Uplink activity factor on FCH
P Sync txi ic
Cell parameter
Cell synchronisation channel power
P paging txi ic
Cell parameter
Cell other common channels (except

CPICH and SCH) power
P pilot txi ic
Cell parameter
Cell pilot power
P max txi ic
Cell parameter
Maximum cell power
M pooling txi ic
Cell parameter
dB
Maximum amount of power

reserved for pooling
P FCH
min
Service parameter
Minimum power allowed for FCH
P FCH
max
Service parameter
Maximum power allowed for FCH
P SCH
min
Service parameter
Minimum power allowed for SCH
P SCH
max
Service parameter
Maximum power allowed for SCH
P FCH txi ic tch
Simulation result including the term AF FCH Serv
Cell FCH power for a traffic channel

on carrier ic
Total FCH power on carrier ic
Simulation result
Transmitter SCH power for a traffic

channel on carrier ic
Total SCH power on carrier ic
Transmitter total transmitted power

on carrier ic
FCH DL
Gp
SCH DL
Gp
FCH UL
Gp
SCH UL
Gp
AF FCH
P FCH txi ic
DL
P FCH txi ic tch
tch FCH ic
P SCH txi ic tch

P SCH txi ic
P SCH ic tch
tch SCH ic
P tx txi ic
P pilot txi ic + P Sync txi ic + P paging txi ic + P SCH txi ic

+ P FCH txi ic
P term
min
Terminal parameter
max
Terminal parameter
P term
FCH
Simulation result including the term AF FCH Serv
Terminal FCH power transmitted in

carrier ic
P term ic
SCH
Simulation result
Terminal SCH power transmitted on

carrier ic
BTS
BTS parameter

transmitted
term
Terminal parameter

P term ic
UL
341

Name
Value
Unit
Description
Clutter parameter

G Tx
Antenna parameter
None
G Term
Terminal parameter
None
Terminal gain
L Tx

None
Transmitter lossa
L body
Service parameter
None
Body loss
L Term
Terminal parameter
None
Terminal loss
L indoor
Clutter (and, optionally, frequency band) parameter
L path
None
Path loss
Terminal parameter
None
Number of fingers
Terminal parameter
Pilot power percentage
M Shadowing model

standard deviation
None

M Shadowing Ec Io

standard deviation
None

None

DL
M Shadowing Eb Nt
npaths
DL
G macro diversity
M Shadowing Eb Nt
Indoor loss
n=2 or 3
DL

standard deviation
None
DL Eb/Nt Shadowing margin

UL

standard deviation
None

None

in soft handoffc.
None

None
P pilot txi ic
------------------------------LT
Chip power received at terminal
P FCH txi ic tch

----------------------------------------LT
Bit received power at terminal for

FCH on carrier ic
UL
UL
G macro diversity
E Shadowing
npaths
UL
M Shadowing Eb Nt
n=2 or 3
Simulation result
UL
For Ec/I0 calculation
------------------------------------------------------------------------------------------------------------------------------------G Tx G term
LT

DL
---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
UL
---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
In simulations
-------------------------------------------------------------------------------------------------------------------G Tx G term
P c txi ic
FCH DL
Pb
342
txi ic tch

AT332_TRR_E0
Name
SCH DL
Pb
txi ic tch
FCH DL
DL
P b txi ic tch
Pb
Value
Unit
Description
P SCH txi ic tch

----------------------------------------LT

SCH on carrier ic

FCH+SCH on carrier ic
Total received power at terminal


from traffic channels of a transmitter
on carrier ic
Bit received power at transmitter for

FCH on carrier ic

SCH on carrier ic

SCH+FCH on carrier ic
Total power transmitted by the

Chip received power at transmitter
SCH DL
txi ic tch + P b
P tx txi ic
-------------------------LT
DL
P tot txi ic
DL
P traf txi ic
tch ic
P FCH txi ic + P SCH txi ic

-----------------------------------------------------------------LT
FCH
FCH UL
ic
P term
-----------LT
SCH UL
ic
P term
-----------LT
Pb
Pb
txi ic tch
SCH
FCH UL
UL
P b ic
Pb
SCH UL
ic + P b
ic
UL
P b ic
UL
UL
P b ic + P c ic = ---------------1 p
UL
P tot ic
UL
UL
P c ic
p Ptot ic
L Tx = L total UL on uplink and L Tx = L total DL on downlink. For information on calculating transmitter losses on uplink and downlink,
a.
see "UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents" on page 30.
npaths
b.

npaths
c.
M Shadowing Eb Nt
d.

UL
UL
). In downlink
prediction studies, carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing Eb Nt
M Shadowing Ec Io ) while extra-cell interference level is not. Therefore, M Shadowing Eb Nt
DL
DL
or
or M Shadowing Ec Io is set to 1 in downlink
extra-cell interference calculation.
5.2.1.2 Ec/I0 Calculation

This table details the pilot quality ( Q pilot or Ec Io ) calculations.
Name
Value
I intra txi ic
P tot txi ic
DL
DL
DL
I extra ic
DL
I inter carrier ic
DL
txj j i
DL
P tot txj
RF ic ic adj
DL
Description



ic adj
txj j
---------------------------------------------
ni
I 0 ic
DL
P tot txj ic
Unit
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
a
Term
DL
DL
DL
DL
I intra txi ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0 W

carrier ic b
343

Name
Value
Unit
Description
E
Q pilot txi ic ----c
I0
BTS P c txi ic
-------------------------------------------------DL
I 0 ic
None
Quality level at terminal on pilot for

carrier ic
a.
b.
In an active set, N 0
Term
is calculated for all its members with Inter-technology downlink noise rise of the best server.
5.2.1.3 DL Eb/Nt Calculation

Eb
DL
This table details calculations of downlink traffic channel quality ( Q tch (tch could be FCH or SCH) or ------ ).
Nt DL
Name
Value
Unit
Description
I intra txi ic
1 BTS F ortho P DL txi ic

tot



DL
DL
I extra ic
txj j i
DL
P tot txj
DL
I inter carrier ic
RF ic ic adj
ni
DL
ic adj
txj
j
---------------------------------------------
DL
N tot ic
DL
P tot txj ic
DL
DL
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
DL
Term
DL
I intra ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0

carrier ic
None

channel from one transmitter for a
FCH channel on carrier ic b

FCH DL
Pb
txi
E b DL
DL
Q FCH txi ic -----
N t FCH
DL
Q FCH ic
BTS
ic tch
DL
------------------------------------------------------------------------------------------------------ G FCH
p
DL
DL
FCH DL
BTS P b
txi ic tch
DL
- G FCH
Total noise: -----------------------------------------------------------------p
DL
N tot ic
DL
f rake efficiency
DL
Q FCH tx k ic
tx k ActiveSet FCH
Quality level at terminal for FCH

using carrier ic due to combination of
None
all transmitters of the active set

SCH DL
Pb
txi
DL
Q SCH txi
E DL
ic ----b-
N t SCH
DL
Q SCH ic
BTS
ic tch
DL
------------------------------------------------------------------------------------------------------ G SCH
p
DL
DL
None
SCH DL
BTS P b
txi ic tch
DL
- G SCH
Total noise: -----------------------------------------------------------------p
DL
N tot ic
DL
f rake efficiency
DL
Q SCH tx k ic
tx k ActiveSet SCH
Quality level at terminal for SCH

using carrier ic due to combination of
None
all transmitters of the active set
DL
DL
G SHO FCH
Q FCH ic
---------------------------------------------------DL
Q FCH BestServer ic
None
Downlink soft handover gain for FCH

None
Downlink soft handover gain for SCH

DL
DL
G SHO SCH
344
Q SCH ic
---------------------------------------------------DL
Q SCH BestServer ic

channel from one transmitter for a
SCH channel on carrier icc

AT332_TRR_E0
Name
Value
Unit
Description
P FCH txi ic
Q req FCH
----------------------- P FCH txi ic
DL
Q FCH ic
Required transmitter FCH traffic

channel power to achieve Eb/Nt
target at terminal on carrier ic
req
P SCH txi
Q req SCH
----------------------- P SCH txi ic
DL
Q SCH ic
Required transmitter SCH traffic

channel power to achieve Eb/Nt
target at terminal on carrier ic
Required transmitter traffic channel

power on carrier ic
DL
req
DL
ic
req
req
P tch txi ic
a.
b.
req
P FCH txi ic + P SCH txi ic
c.
5.2.1.4 UL Eb/Nt Calculation

Eb
UL
This table details calculations of uplink traffic channel quality ( Q tch (tch could be FCH or SCH) or ------ ).
Nt UL
Name
Value
Pb
UL
UL intra
I tot
txi ic
UL extra
I tot
UL
UL

ic

carrier ic

UL
P b ic + P c ic
term
txj j i
UL
P b ic adj
UL
+ P c ic adj
term
txj j
----------------------------------------------------------------------
Description
UL
ic + P c ic
term
txi
txi ic
Unit
RF ic ic adj
UL
I tot txi ic
UL
N tot txi ic
UL extra
I tot
UL intra
Tx

UL
UL
W
txi ic +I inter carrier txi ic
tx
I tot txi ic + N 0


ic (Uplink interference) a
None

traffic channel for the FCH channel
on carrier icb
None

traffic channel for the SCH channel
on carrier icc

FCH UL
E
UL
Q FCH txi ic ----b-
N t UL
term P b
ic
UL
-------------------------------------------------------------------------------------------------------- G FCH
p
UL
Tx
UL
FCH UL
term P b
ic
FCH UL
Total noise: ----------------------------------------------- Gp
UL
N tot txi ic
SCH UL
E
UL
Q SCH txi ic ----b-
N t UL
term P b
ic
UL
-------------------------------------------------------------------------------------------------------- G SCH
p
UL
Tx
UL
SCH UL
term P b
ic
SCH UL
Gp
Total noise: -----------------------------------------------UL
N tot txi ic
345

Name
Value
Unit
Description
UL
No HO: Q tch txi ic
UL
UL
Q tch txk ic
tx k ActiveSet
samesite
Soft, Softer/Soft HO (No MRC):

Max Q UL
tch tx k
tx k ActiveSet
ic
UL
G macro diversity
Softer/Soft HO (MRC):
UL
Q tch ic
UL
UL
UL
f
tx
ic
tx
ic
tch
k
tch
l
tx k ,tx l ActiveSet rake efficiency
tx k samesite
tx k
Max

account increase of the quality due
None to macro-diversity (macro-diversity
gain).
tch could be FCH or SCH
In simulations,
UL
G macro diversity
tx othersite
l
= 1.
UL
G macro diversity
UL
Q FCH ic
---------------------------------------------------UL
Q FCH BestServer ic
UL
G SHO FCH
None
Uplink soft handover gain for FCH

None
Uplink soft handover gain for SCH


Eb/Nt target at transmitter for FCH
on carrier ic

Eb/Nt target at transmitter for SCH
on carrier ic
Required terminal power on carrier

ic
UL
G SHO SCH
Q SCH ic
---------------------------------------------------UL
Q SCH BestServer ic
FCH req
P term
ic
Q req FCH
---------------------- P FCH
term ic
UL
Q FCH ic
SCH req
P term
ic
Q req SCH
----------------------- P SCH
term ic
UL
Q SCH ic
UL
UL
UL
FCH req
req
P term ic
P term
SCH req
ic + P term
ic
tx
a.
In an active set, N 0 is calculated for all its members with Inter-technology uplink noise rise of the best server.
b.
c.
5.2.1.5 Simulation Results

This table contains some simulation results provided in the Cells and Mobiles tabs of the simulation property dialog box.
Name
DL
I intra txi ic
Value
Unit
Description
DL
DL
P tot txi ic F ortho BTS P tot txi ic
None


DL
1 F ortho BTS P b txi ic

DL
I extra ic
346
txj j i
DL
P tot txj ic

AT332_TRR_E0
Name
Value
DL
I inter carrier ic
DL
DL
DL
DL
I tot
DL
DL
Term
Pb
Total effective interference at

terminal on carrier ic (after
unscrambling)

carrier ic

ic

carrier ic

UL
ic + P c ic
term
txi
UL extra
I tot
txi ic
UL
UL
P b ic + P c ic
term
txj j i
UL
P b ic adj
UL
UL
+ P c ic adj
term
txj j
----------------------------------------------------------------------
RF ic ic adj
UL
I tot txi ic
UL extra
I tot
UL intra
Tx
UL
UL
N tot txi ic
a
I tot ic + N 0
txi ic

UL
UL intra
DL
DL
Description
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
N tot ic
Unit
ic adj
txj j
--------------------------------------------RF ic ic adj
I tot ic
DL
P tot txj
UL
W
tx
I tot txi ic + N 0


ic
None
UL
UL
I tot txi ic
---------------------------UL
N tot txi ic
UL
I tot txi ic
--------------------------------------------------------------------------------------UL intra
Tx
I tot
txi ic 1 F MUD term
None
Cell uplink reuse factor on carrier ic
E txi ic
UL
1
--------------------------UL
F txi ic
None

carrier ic
DL
P tx txi ic
---------------------------- P max txi ic 100
None
Percentage of max transmitter

power used.
None
Downlink load factor on carrier ic
None
Downlink reuse factor on a carrier ic
dB
X txi ic
UL
F txi ic
%Power txi ic
Simulation result available per cell

DL
DL
I extra ic + I inter carrier ic L T

---------------------------------------------------------------------------------- + 1 F ortho BTS
P tx txi ic
-------------------------------------------------------------------------------------------------------------------------------1
---------+
tch
ortho
BTS
DL
CI req
DL
X txi ic
with
DL
CI req
SCH DL
FCH DL
Q req
Q req
= -------------------+ -------------------SCH DL
FCH DL
Gp
Gp
DL
I tot ic
Simulation result available per mobile: -----------------DL
N tot ic
DL
DL
F txi ic
DL
NR txi ic
I tot ic
----------------------------DL
I intra txi ic
DL
10 log 1 X txi ic
347

Name
Value
UL
UL
10 log 1 X txi ic
NR txi ic
a.
Unit
Description
dB
5.2.2 Parameters Used for CDMA2000 1xEV-DO Modelling

5.2.2.1 Inputs
This table lists simulation and prediction inputs (calculation options, quality targets, active set management conditions, etc.)
Name
Value
Unit
Description
F ortho
Clutter parameter
None
F MUD
Tx
None
MUD factor
cn first
None
cn last
None
Last carrier number
cn
None
Carrier number step

cn cn first
- cn lower
ic = ------------------------ cn
ic
None

Q pilot
req
min
Q pilot txi ic
req
Min. Ec/I0 - Cell parameter
min
T_Drop - Cell parameter
Q pilot
Q pilot txi ic
req
min
min
Active set upper threshold

None (used to determine the best server in
the active set)
None
Active set lower threshold

(used to determine other members
of the active set)

None cell to be the best server in the active
set
None

cell not to be rejected from the
active set

None required from the cell to be the best
server in the active set
req
Delta Min. Ec/I0 - Mobility parameter
Q pilot
min
Delta T_Drop - Mobility parameter
None

required from the cell not to be
rejected from the active set
RSCP min Txi ic

Ec
--- N t min Rev0
Mobility parameter for 1xEV-DO Rev. 0 users
None
Minimum pilot quality required in

the uplink to operate EV-DO Rev. 0
Ec
--- N t min RevB
None
Minimum pilot quality required in

the uplink to operate multi-carrier
EV-DO
Ec
--- N t min
Parameter read in the 1xEV-DO Radio Bearer Selection (Uplink) table

for 1xEV-DO Rev. A and Rev. B users
None
Minimum pilot quality level required

to obtain a radio bearer in the uplink
Q pilot
UL
UL
UL
348
req

AT332_TRR_E0
Name
Value
n SF
1xEV-DO Radio Bearer Selection (Uplink) table
Number of subframes associated

None with the 1xEV-DO radio bearer in the
uplink
TP P R LC
UL
1xEV-DO Radio Bearer Selection (Uplink) table
None
Uplink RLC peak throughput

provided by the 1xEV-DO radio
bearer
Ec
--- N t min
Parameter read in the 1xEV-DO Radio Bearer Selection (Downlink)

table for 1xEV-DO Rev. A and Rev. B users
None
Minimum pilot quality level required

to obtain a radio bearer in the
downlink
n TS
1xEV-DO Radio Bearer Selection (Downlink) table
None
Number of timeslots associated with

the 1xEV-DO radio bearer in the
downlink
DL
Downlink 1xEV-DO Radio Bearer Table
None
Downlink RLC peak throughput

provided by the 1xEV-DO radio
bearer
N EVDO CE N I
Site parameter
None
Number of EVDO channel elements

available for a site on uplink and
downlink
N EVDO CE N I
Simulation result
None
Total number of EVDO channel

elements of a site consumed by
users on uplink and downlink
N CE UL
TCH
None

for TCH on uplink
N MacIndexes txi ic
None
Maximum number of MAC indexes

available per cell (59 for Rev0 and
114 for RevA)
N MacIndexes txi ic
Simulation result
None
Number of MAC indexes used by the

cell
n EVDO txi ic
Simulation constraint (cell parameter)
None
Maximum number of EVDO users

that can be connected to the cell
n EVDO txi ic
Simulation result
None
Number of EVDO users connected to

the cell
NF term
Terminal parameter
None
NF Tx

None
1.38 10-23
J/K
Boltzman constant
293
Ambient temperature
1.23 MHz
Hz
Spreading Bandwidth
Cell parameter
Cell parameter
None
RF ic ic adj
Network parameter
interference
None

DL
TP P R LC
Max
Max
Max
Tx DL
Tx UL
Tx m
ICP ic ic
i
Unit
Network parameter
Description
and ic adj
None
network) and ic
UL
X max
Tx
N0

Tx UL
349

Name
Unit
Description
Rc
bps
Chip rate
f rake efficiency
Equipment parameter
Term
N0
UL
UL
Value
Tx DL
Simulation result
bps
Uplink throughput
TP TCP ACK
Simulation result
bps
Uplink throughput due to TCP

aknowledgements
TPBCMCS
Cell parameter
bps
Downlink throughput for Broadcast/

Multicast services
TP max DL
Simulation result
bps
Downlink peak throughput supplied

to the terminal
TP avg
Simulation result
bps
Downlink average cell throughput
TPD min UL
Service parameter
kbps

available in the uplink
TPD min DL
Service parameter
kbps

available in the downlink
TPA
f TP Scaling TP max DL TP Offset
bps
Downlink user application

throughput
f TP Scaling
Service parameter
Scaling factor
TPOffset
Service parameter
kbps
Offset
C DL Bearer
TPD min DL
-----------------------------------------------------------DL
TP P R LC Index DL Bearer
Downlink radio bearer consumption

for a (1xEV-DO Rev. A - Guaranteed
Bit Rate) service user
C UL Bearer
TPD min UL
-----------------------------------------------------------UL
TP P R LC Index UL Bearer
Uplink radio bearer consumption for

a (1xEV-DO Rev. A - Guaranteed Bit
Rate) service user
Gp
W---------UL
TP
None

FCH
G idle power
Cell parameter
None
Idle power gain
G MU
Cell parameter
None
Multi user gain
P max txi ic
Cell parameter
Max cell power
P tx txi ic b pilot
P max txi ic
Pilot burst transmitted by the

transmitter on carrier ic.
Traffic burst transmitted by the

transmitter on carrier ic.
TP
UL
DL
DL
UL
P tx txi ic b traffic
350
P max txi ic if users to support

P max txi ic G idle power if no user to support
ER DRC
Cell parameter
Error rate on the DRC channel
TS BCMCS
Cell parameter
Pourcentage of EVDO timeslots

dedicated to Broadcast/Multicast
services
TS EVDO CCH
Cell parameter
Pourcentage of EVDO timeslots

dedicated to control channels
P term ic
Simulation result
Terminal power transmitted on

carrier ic

AT332_TRR_E0
Name
Value
Unit
Description
P term
min
Terminal parameter
P term
max
Terminal parameter
BTS
BTS parameter

transmitted
term
Terminal parameter

Clutter parameter

G Tx
Antenna parameter
None
G Term
Terminal parameter
None
Terminal gain
L Tx

None
Transmitter lossa
L body
Service parameter
None
Body loss
L Term
Terminal parameter
None
Terminal loss
L indoor
L path
None
Path loss
G ACK
Terminal parameter
None
Acknowledgement Channel gain
G RRI
Terminal parameter (for 1xEV-DO Rev A terminals only)
None
Reverse Rate Indicator Channel gain
G DRC
Terminal parameter
None
Data Rate Control Channel gain
G Auxiliary pilot
Terminal parameter (for 1xEV-DO Rev A terminals only)
None
Auxiliary Pilot Channel gain
G TCH
Terminal parameter
None
Traffic data Channel gain
carriers
Terminal parameter
None
Maximum number of carriers in

multi-carrier mode
M Shadowing model

standard deviation
None

M Shadowing Ec Io

standard deviation
None

None

None

None

in soft handoffc.
None

n max
DL
n=2 or 3
UL

standard deviation
UL
UL
G macro diversity
E Shadowing
npaths
DL
G macro diversity
M Shadowing Eb Nt
Indoor loss
npaths
UL
M Shadowing Eb Nt
n=2 or 3
Simulation result
UL
351

Name
Value
Unit
Description
None
P tx txi ic b pilot
----------------------------------------LT
Pilot burst received at terminal from

a transmitter on carrier ic
P tx txi ic b traffic
--------------------------------------------LT
Traffic burst received at terminal

P b ic
P term
-----------LT
Bit received power at transmitter on

carrier ic
NRthreshold txi ic
Cell parameter
dB
Cell uplink noise rise threshold
Cell parameter
dB
Cell uplink noise rise upgrading/

downgrading delta
For Ec/I0 and Ec/Nt calculations
------------------------------------------------------------------------------------------------------------------------------------G Tx G term
UL
---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
LT
In simulations
-------------------------------------------------------------------------------------------------------------------G Tx G term
DL
P tot txi ic b pilot

DL
P tot txi ic b traffic

UL
UL
UL
NR threshold txi ic
a.
L Tx = L total UL on uplink and L Tx = L total DL on downlink.
b.
npaths

npaths
c.
M Shadowing Eb Nt
d.

UL
UL
). In downlink
prediction studies, carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing Ec Io ) while
extra-cell interference level is not. Therefore, M Shadowing Ec Io is set to 1 in downlink extra-cell interference calculation.
5.2.2.2 Ec/I0 and Ec/Nt Calculations

Ec
Ec
Ec
This table details ---- txi ic b pilot , ----- txi ic b pilot and ----- txi ic b traffic calculations.
I0
Nt
Nt
Name
Value
Unit
Description
txi ic
DL
I intra
b pilot or b traffic

terminal on carrier ic (only one
mobile is served at a time)
Downlink extra-cell interference

based on pilot at terminal on carrier
ic

based on traffic at terminal on
carrier ic

based on pilot at terminal on carrier
ic
DL
I extra ic b pilot
DL
I extra ic b traffic
P tot txj ic b pilot
P tot txj ic b traffic
DL
txj j i
DL
txj j i
Ptot txj icadj bpilot

DL
DL
I inter carrier ic
352
b pilot
txj j
-------------------------------------------------------------
RF ic ic adj

AT332_TRR_E0
Name
DL
I inter carrier ic
Value
b traffic
DL
P tot txj
DL
DL
DL
I 0 ic
b pilot
b traffic
DL
DL
Total noise based on pilot received at

Total noise based on traffic received

I extra ic b pilot + N 0
Total noise based on pilot received at

DL

None
Pilot quality level at terminal on

carrier ic
None
Pilot quality level at terminal on

carrier ic
None
Traffic quality level at terminal on

carrier ic
DL
DL
term
N0
DL
DL
DL
N tot ic b pilot
DL
N tot ic b traffic
term
N0
term
term
I extra ic b traffic + N 0
Q pilot txi ic
DL
BTS P tot txi ic b pilot

---------------------------------------------------------------------DL
I 0 ic b pilot
Ec
---- txi ic b pilot
I0
DL
E
----c- txi ic b pilot
Nt
BTS P tot txi ic b pilot

---------------------------------------------------------------------------------------------------------------DL
DL
N tot ic b pilot + 1 BTS P tot txi ic b pilot
E
----c- txi ic b traffic
Nt
BTS P tot txi ic b traffic

----------------------------------------------------------------------------------------------------------------------DL
DL
N tot ic b traffic + 1 BTS P tot txi ic b traffic
a.
a
P tot txi ic b traffic + I extra ic b traffic + I inter carrier ic b traffic

+

carrier ic
DL
DL
DL
I 0 ic
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
P tot txi ic b pilot + I extra ic b pilot + I inter carrier ic b pilot

+
Description
ic adj b traffic
txj j
---------------------------------------------------------------RF ic ic adj
Unit
DL
5.2.2.3 UL Eb/Nt Calculation

This table details calculations of uplink quality ( Q
Name
I tot
Pb
txi ic
UL extra
I tot
Eb
or ------ ).
Nt UL
Value
UL
UL intra
UL
ic
term
txi
txi ic
Unit
Description

ic

carrier ic

UL
P b ic
term
txj j i
UL
P b ic adj
UL
term
txj j
-----------------------------------
RF ic ic adj
UL
I tot txi ic
UL
N tot txi ic
UL extra
I tot
UL intra
Tx

UL
tx
I tot ic + N 0
UL
W

ic
353

Name
Value
Unit
Description
None
Quality level at transmitter on carrier

ica

UL
term P b ic
-------------------------------------------------------------------------------------------------------- G UL
p
UL
Tx
UL
E
UL
Q txi ic ----b-
N t UL
UL
term P b ic
UL
Total noise: ----------------------------------- Gp
UL
N tot txi ic
UL
No HO: Q txi ic
UL
UL
Q tch txk ic
tx k ActiveSet
samesite
Soft, Softer/Soft HO (No MRC):

UL
Max Q UL
tch tx k ic G macro diversity
tx k ActiveSet
Softer/Soft HO (MRC):
UL
Q total ic
UL
UL
UL
Q tch tx k ic Q tch tx l ic
tx ,tx ActiveSet f rake efficiency
k l
tx k samesite
tx k
Max

account increase of the quality due
None to macro-diversity (macro-diversity
gain).
In simulations,
UL
G macro diversity
tx l othersite
= 1.
UL
G macro diversity
UL
Q total ic
------------------------------------------------UL
Q BestServer ic
UL
G SHO
None
Uplink soft handover gain on carrier

ic
None
Eb/Nt target on uplink

Eb/Nt target at transmitter on carrier
ic
For 1xEV-DO Rev 0 terminal

UL
E
UL
c
-----
G p 1 + G ACK + G DRC + G TCH
N t min
For 1xEV-DO Rev A terminalb
When the acknoledgement signal is considered
UL
Q req
Ec
UL
--- N t min G p 1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary Pilot
UL
When the acknoledgement signal is not considered

UL
E
UL
----c-
G p 1 + G RRI + G DRC + G TCH + G Auxiliary Pilot
N t min
UL
Q req
---------------------- P term
UL
Q total ic
req
P term ic
a.
In simulations, the uplink Eb/Nt target is calculated without considering the acknowledgement signal.
b.
5.2.2.4 Simulation Results

This table contains some simulation results provided in the Cells and Mobiles tabs of the simulation property dialog box.
354

AT332_TRR_E0
Name
Value
Unit
Description
I intra txi ic b traffic
1 F ortho BTS P tot txi ic b traffic = 0
DL

terminal on carrier ic (only one
mobile is served at a time)

carrier ic

carrier ic
DL
DL
I extra ic b traffic
DL
P tot txj ic b traffic
txj j i
Ptot txj icadj btraffic

DL
DL
I inter carrier ic
b traffic
txj j
----------------------------------------------------------------
RF ic ic adj
DL
ni
DL
DL
I tot ic b traffic
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP n ic
DL
DL
DL
term
I tot ic b traffic + N 0
Pb
UL
I tot
txi ic
UL extra
I tot
Total effective interference based on

traffic at terminal on carrier ic (after
unscrambling)


ic

carrier ic

DL
+ I inter techno log y ic
DL
ic
term
txi
txi ic
UL
P b ic
term
txj j i
UL
P b ic adj
UL
term
txj j
-----------------------------------
RF ic ic adj
UL
I tot txi ic
UL extra
I tot
UL intra
Tx
a
I intra ic b traffic + I extra ic b traffic + I inter carrier ic b traffic
N tot ic b traffic
UL intra
UL
W

N tot txi ic
I tot txi ic + N 0

ic
N mobiles txi ic
Simulation result
None
Number of mobiles connected to

transmitter txi on carrier ic
UL
UL
tx
N GBR m obiles txi ic
Simulation result
None
Number of (1xEV-DO Rev. A Guaranteed bit rate) service users

connected to transmitter txi on
carrier ic
N VBR m obiles txi ic
Simulation result
None
Number of (1xEV-DO - Variable bit

rate) service users connected to
transmitter txi on carrier ic
DL
X txi ic
DL
I tot ic b traffic
------------------------------------DL
N tot ic b traffic
None Cell downlink load factor on carrier ic
UL
UL
I tot txi ic
---------------------------UL
N tot txi ic
UL
UL
X txi ic
None
I tot txi ic
--------------------------------------------------------------------------------------UL
intra
Tx
I tot
txi ic 1 F MUD term
None
Cell uplink reuse factor on carrier ic
1
-------------------------UL
F txi ic
None

carrier ic
UL
F txi ic
E txi ic
355

Name
Value
Unit
Description
DL
dB
UL
dB
DL
10 log 1 X txi ic
UL
10 log 1 X txi ic
NR txi ic
NR txi ic
a.
5.3 Active Set Management

The mobile active set is the list of the transmitters to which the mobile is connected. The active set may consist of one or more
transmitters; depending on whether the service supports soft handoff and on the terminal active set size. Transmitters in the
mobile active set must use a frequency band with which the terminal is compatible and the pilot signal level received from
these transmitters must exceed the defined minimum RSCP threshold.
It is, however, the quality of the pilot (EcI0) that finally determines whether or not a transmitter can belong to the active set.
Cells entering the mobiles active set must fulfill the following conditions:
The best server (first cell entering active set)

In order for a given transmitter to enter the mobile active set as best server, the quality of this transmitters pilot must
be the highest one and it must exceed an upper threshold equal to the sum of the minimum Ec/I0 defined in the
properties of the best serving cell and the Delta minimum Ec/I0 defined in the properties of the mobility type. The
upper threshold is set for the carrier as defined in the cell properties and can also take into account the user mobility
type if the Delta minimum Ec/I0 defined in the mobility type is different from 0. The carrier used by the transmitters
in the active set corresponds to the best carrier of the best server. For information on the best carrier selection, see
the Technical Reference Guide.
In order for a transmitter to enter the active set (other cells of active set):
They must use the same carrier as the best server cell,
The pilot quality from other candidate cells must exceed a lower threshold. The lower threshold depends both on
the type of carrier and the mobility type. It is equal to the sum of T_Drop defined in the properties of the best
server and the Delta T_Drop defined in the properties of the mobility type.
If you have selected to restrict the active set to neighbours, the cell must be a neighbour of the best server (the
"restricted to neighbours option is selected in the equipment properties).
For multi-carrier EVDO Rev.B users, the active set may consist of several sub-active sets, each one being associated with one
carrier. The number of sub-active sets depends on the maximum number of carriers supported by the terminal. As detailed
above, the quality of the pilot (EcI0) determines whether or not a transmitter can belong to a sub-active set. The sub-active
set associated with the best carrier is the same as the active set of a single-carrier user. For the other carriers, the uplink EcNt
received by the best server on the best carrier and on the studied carrier determines whether or not a carrier can have a subactive set, and the transmitters in the sub-active sets depend on the mode supported by the terminal (locked mode or
unlocked mode):
The Ec/Nt received by the best serving transmitter on the best carrier must exceed the minimum uplink Ec/Nt defined
in the properties of the transmitter.
The Ec/Nt received by the best serving transmitter on the studied carrier must exceed the minimum uplink Ec/Nt
defined in the properties of the transmitter.
When the locked mode is used, the serving transmitters must be the same in all sub-active sets. With the unlocked
mode, the serving transmitters may be different from one sub-active set to another.
5.4 Simulations
The simulation process is divided into two steps:
1. Obtaining a realistic user distribution
Atoll generates a user distribution using a Monte-Carlo algorithm, which requires traffic maps and data as input. The
resulting user distribution complies with the traffic database and maps provided to the algorithm.
the simulation and on the network interferences. A user may be either active or inactive. Both active and inactive
users consume radio resources and create interference.
Additionally, each 1xEV-DO Rev. 0 user is assigned a transition flag ("True" or "False") for each possible throughput
transition (from 9.6 to 19.2 kbps, 19.2 to 38.4 kbps, 38.4 to 76.8 kbps, and 76.8 to 153.6 kbps for throughput upgrading
356

AT332_TRR_E0
and from 153.6 to 76.8 kbps, 76.8 to 38.4 kbps, 38.4 to 19.2 kbps, and 19.2 to 9.6 kbps for throughput downgrading).
These transition flags are based on the throughput downgrading and upgrading probabilities. If a transition flag is
"True," the user throughput can be downgraded or upgraded if necessary.
Then, Atoll randomly assigns a shadowing error to each user using the probability distribution that describes the
shadowing effect.
Finally, another random trial determines user positions in their respective traffic zone and whether they are indoors
or outdoors (according to the clutter weighting and the indoor ratio per clutter class defined for the traffic maps).
2. Modelling the network regulation mechanism
This algorithm depends on the network. Atoll uses a power control algorithm in case of CDMA2000 1xRTT networks
and a different algorithm, which mixes throughput control on downlink and power control on uplink, for CDMA2000
1xEV-DO networks.

5.4.1.1 Number of Users, User Activity Status and User Throughput
During the simulation, a first random trial is performed to determine the number of users and their activity status. The
determination of the number of users and the activity status allocation depend on the type of traffic cartography used.
[Simulation]
RandomTotalUsers=0
5.4.1.1.1
Simulations Based on User Profile Traffic Maps

user profiles with an associated mobility type and a given density (number of subscribers with the same profile per km).
User profile traffic maps: Each polygon and line of the map is assigned a density of subscribers with given user profile and
mobility type. If the map is composed of points, each point is assigned a number of subscribers with given user profile and
mobility type.
The user profile models the behaviour of the different subscriber categories. Each user profile contains a list of services and
their associated parameters describing how these services are accessed by the subscriber.
From environment (or polygon) surface (S) and user profile density (D), a number of subscribers (X) per user profile is inferred.
X = SD
In case of user profile traffic maps composed of lines, the number of subscribers (X)
per user profile is calculated from the line length (L) and the user profile density (D)
(nb of subscribers per km) as follows: X = L D
The number of subscribers (X) is an input when a user profile traffic map is composed
of points.
For each behaviour described in a user profile, according to the service, frequency use and exchange volume, Atoll calculates
the probability for the user being connected in uplink and in downlink at an instant t.
N call d
p 0 = ------------------3600
where N call is the number of calls per hour and d is the average call duration (in second).
Then, Atoll calculates the total number of users trying to access a certain service.
Calculation of the number of users trying to access the service j ( n j ):
nj = X p0
357

The next step determines the activity status of each user.
This steps depends on the type of service (Voice, 1xRTT data, 1xEV-DO data).
CDMA2000 1xRTT Services
Activity status of voice and data service users is determined as follows.

Users are always active on FCH in both directions, uplink and downlink. Therefore, we have:
Probability of being active on UL: p UL = 0
Probability of being active on DL: p DL = 0
Probability of being active both on UL and DL: p UL + DL = 1
Probability of being inactive: p inactive = 0
Thus, for voice and data services, we have:
Number of inactive users: n j inactive = n j p inactive = 0
Number of users active on UL: n j UL = n j p UL = 0
Number of users active on DL: n j DL = n j p DL = 0
Number of users active on UL and DL both: n j UL + DL = n j p UL + DL = n j
n j = n j UL + n j DL + n j UL + DL + n j inactive = n j UL + DL
Voice Users
Voice users are active on uplink and downlink. However, the FCH can have inactivity periods on both links. This is modelled by
UL
DL
the FCH activity factor, AF FCH and AF FCH . Therefore, all voice service users try to access the service with the following FCH
FCH
UL
FCH
DL
throughputs, TP P UL AF FCH on uplink and TP P DL AF FCH on downlink.

FCH
FCH
TP P UL and TP P DL are respectively the uplink and downlink FCH peak throughputs.
Data Users
Data service users are active on uplink and downlink. FCH is always allocated but can have inactivity periods on both links; this
UL
DL
is modelled by the FCH activity factor, AF FCH and AF FCH . SCH may be allocated with four possible throughputs (2x, 4x, 8x and
16xFCH peak throughput).
Therefore, data service users can access the service with different throughputs. Possible throughputs are detailed in the table
below:
Only FCH is used
SCH throughput factor

rk
On UL
TP P UL AF FCH
2x
TP P UL AF FCH + 2
4x
TP P UL AF FCH + 4
8x
TP P UL AF FCH + 8
16x
TP P UL AF FCH + 16
Both FCH and SCH are used
FCH
Allocated throughputs
FCH
On DL
UL
FCH
DL
TP P DL AF FCH
FCH
UL
TPP DL AF FCH + 2
FCH
UL
TPP DL AF FCH + 4
FCH
UL
TPP DL AF FCH + 8
FCH
UL
FCH
DL
FCH
DL
FCH
DL
FCH
DL
TP P DL AF FCH + 16
FCH
TP P UL and TP P DL are respectively the uplink and downlink FCH peak throughputs.
Then, Atoll determines the distribution of users between the different possible throughputs.
UL
In case of a data service, j, several data throughput probabilities, P k

factors, r k , for SCH channel.
358
DL
and P k , can be assigned to different throughput

AT332_TRR_E0
For non-data services, these probabilities are 0.
For data service users, a random trial compliant with throughput probabilities is performed for each link in order to determine
the throughput for each user.
On uplink, we have:
r
FCH
UL
FCH
DL
For each SCH throughput factor, r k , the number of users n j with the throughput TP P UL AF FCH + r k is calculated as
follows,
rk
UL
nj = Pr nj
k
FCH
Therefore, the number of users n j

FCH
nj
= nj
FCH
UL
with the throughput, TP P UL AF FCH , is:
rk
nj
rk
On downlink, we have:
rk
For each SCH throughput factor, r k , the number of users, n j with the throughput, TP P DL AF FCH + r k , is calculated as
follows,
rk
DL
nj = Pr nj
k
FCH

FCH
nj
= nj
FCH
DL
with the throughput, TP P DL AF FCH , is:
rk
nj
rk
CDMA2000 1xEV-DO Services
As power control is performed in the uplink only, 1xEV-DO data service users will be considered either active in the uplink or
inactive. 1xEV-DO data Rev. 0 service users can access the service with uplink throughputs of 9.6, 19.2, 38.4, 76.8 and 153.6
kbps. 1xEV-DO data Rev. A and Rev. B service users can access the service with uplink throughputs of 4.8, 9.6, 19.2, 38.4, 76.8,
115.2, 153.6, 230.4, 307.2, 460.8, 614.4, 921.6, 1,228.8 and 1,848.2 kbps.
UL
UL
For each service, j, several throughput probabilities, P k , can be assigned to different throughputs TP k . The number of users
active on uplink ( n j UL ) and the number of inactive users ( n j inactive ) are calculated as follows:
Probability of being active on UL: p UL =
Pk
UL
UL
TP k
UL
Rk
Pk
UL
UL
TP k
UL
k
Probability of being active on DL: p DL = 0

Probability of being active on UL and DL both: p UL + DL = 0
Therefore, we have:
Number of users active on UL: n j UL = n j p UL
Number of inactive users: n j inactive = n j p inactive
Number of users active on DL: n j DL = n j p DL = 0
Number of users active on UL and DL both: n j UL + DL = n j p UL + DL = 0
n j = n j UL + n j DL + n j UL + DL + n j inactive = n j UL + n j inactive
359


UL
Then, Atoll determines the distribution of users between the different possible throughputs, TP k . The number of users with
UL
UL
the throughput TP k , n j TP k , is calculated as follows:

UL
UL
n j TP k = P k n j
Inactive users have a requested throughput equal to 0.
The user distribution per service is an average distribution and the service of each
user is randomly drawn in each simulation. Therefore, if you compute several
simulations at once, the average number of users per service will correspond to the
calculated distribution. But if you check each simulation, the user distribution
between services is different in each of them.
It is the same for the SCH throughput distribution between 1xRTT data service users
and the traffic throughput distribution between 1xEV-DO data service users.
In calculations detailed above, we assume that the sum of throughput probabilities is
less than or equal to 1. If the sum of throughput probabilities exceeds 1, Atoll
considers normalised throughput probabilities values, P r P r , instead of

k
k
rk
specified throughput probabilities P r .

k
5.4.1.1.2
Simulations Based on Sector Traffic Maps

Sector traffic maps can be based on live traffic data from OMC (Operation and Maintenance Centre). Traffic is spread over the
best server coverage area of each transmitter and each coverage area is assigned either the throughputs in the uplink and in
the downlink, or the number of users per activity status or the total number of users (including all activity statuses).
CDMA2000 1xRTT Services
Voice Service (j)
For each transmitter, Txi, Atoll proceeds as follows:
When selecting Throughputs in Uplink and Downlink, you can input the throughput demands in UL ( TPD
DL ( TPD
DL
UL
) and
) for each sector.
Atoll calculates the number of users active in UL and DL using the voice service in the Txi cell as follows:
UL
DL
N UL = TPD
--------------- and N DL = TPD
--------------UL
DL
TP j
TP j
Where,
UL
TPD is the number of kbits per second transmitted in UL in the Txi cell to provide the service j to the users (userdefined value in the traffic map properties)
DL
TPD is the number of kbits per second transmitted in DL in the Txi cell to provide the service j to the users (userdefined value in the traffic map properties).
UL
TP j
DL
and TP j
correspond to the UL and DL throughputs of a user. FCH is always allocated to active users but
UL
can have inactivity periods on both links. Therefore, we have TP j
FCH
UL
FCH
= TP P UL AF FCH (where TP P UL is the
UL
service FCH peak throughput on UL and AF FCH corresponds to the FCH activity factor on UL) and
DL
TP j
FCH
DL
FCH
DL
= TPP DL AF FCH (where TP P DL is the service FCH peak throughput on DL and AF FCH corresponds to the
FCH activity factor on DL).

Users are always active on FCH for both links. Therefore, we have following activity probabilities.
Probability of being active in UL: p UL = 0
Probability of being active in DL: p DL = 0
Probability of being active in UL and DL both: p UL + DL = 1
360

AT332_TRR_E0

Number of users active in UL and DL both: n j UL + DL = max (N UL,N DL)
Number of users active in UL and inactive in DL: n j UL = 0
Number of users active in DL and inactive in UL: n j DL = 0
inactive
Number of inactive users in UL and DL: n j
= 0
Therefore, all connected voice users ( n j ) are active in both links.
When selecting Total Number of Users (All Activity Statuses), you can input the number of connected users for
each sector ( n j ).
inactive
= n j p inactive = 0
Number of users active in UL and inactive in DL: n j UL = n j p UL = 0

Number of users active in DL and inactive in UL: n j DL = n j p DL = 0
Number of users active in UL and DL both: n j UL + DL = n j p UL + DL = n j
Therefore, all connected users ( n j ) are active in both links.
When selecting Number of Users per Activity Status, you can directly input the number of users active in the
uplink and downlink ( n j UL + DL ), for each sector.
FCH
UL
FCH
DL
Voice service users try to access the service with the FCH throughputs, TPP UL AF FCH on uplink and TP P DL AF FCH on
downlink.
All user characteristics determined, a second random trial is performed to obtain their geographical positions.
Data Service Users (j)
FCH is always allocated to active users but can have inactivity periods on both links. This is modelled by the FCH activity factors,
UL
DL
AF FCH and AF FCH . SCH may be allocated with four possible throughputs (2x, 4x, 8x, 16xFCH peak throughput). Several
UL
DL
throughput probabilities, P k and P k , can be assigned to different throughputs factor, r k , for SCH channel.
For non-data services, these probabilities are 0.
For each transmitter, Txi, Atoll proceeds as follows:
DL ( TPD
DL
UL
) and
) for each sector.
Atoll calculates the number of users active in UL and DL using the service in the Txi cell as follows:
UL
DL
N UL = TPD
--------------- and N DL = TPD
--------------UL
DL
TP j
TP j
361

Where,
UL
TPD is the number of kbits per second transmitted in UL in the Txi cell to provide the service j to the users (userdefined value in the traffic map properties)
DL
TPD is the number of kbits per second transmitted in DL in the Txi cell to provide the service j to the users (userdefined value in the traffic map properties).
UL
TP j
UL
Pj
DL
and TP j
rk
DL
Pj
rk
correspond to uplink and downlink throughputs of a user.
UL
UL
UL
r k + AF FCH R FCH P r + 1
DL
DL
DL
r k + AF FCH R FCH P r + 1
k
FCH
rk
rk
UL
FCH
UL
P r TP P UL AF FC
k
DL
FCH
DL
P r TP P DL AF FC
k
FCH
TP P UL and TP P DL are the uplink and downlink FCH peak throughputs respectively.
In calculations detailed above, we assume that the sum of throughput probabilities is

less than or equal to 1. If the sum of throughput probabilities exceeds 1, Atoll
considers normalised throughput probabilities values, P r P r , instead of

k
k
rk
specified throughput probabilities P r .

k
Then, Atoll calculates the number of users per activity status and the total number of users:
Number of users active in UL and DL both: n j UL + DL = max (N UL,N DL)
Number of users active in UL and inactive in DL: n j UL = 0
Number of users active in DL and inactive in UL: n j DL = 0
inactive
= 0
inactive
362
= n j p inactive = 0

AT332_TRR_E0
Number of users active in UL and inactive in DL: n j UL = n j p UL = 0

Number of users active in DL and inactive in UL: n j DL = n j p DL = 0
Number of users active in UL and DL both: n j UL + DL = n j p UL + DL = n j
When selecting Number of Users per Activity Status, you can directly input the number of users active in the
uplink and downlink ( n i UL + DL ), for each sector.
As explained above, data service users can access the service with different throughputs. Possible throughputs are detailed in
the table below:
Only FCH is used
SCH throughput factor

rk
Allocated throughputs
On UL
TP P UL AF FCH
2x
TPP UL AF FCH + 2
4x
TPP UL AF FCH + 4
8x
TPP UL AF FCH + 8
16x
TPP UL AF FCH + 16
FCH
Both FCH and SCH are used
On DL
UL
FCH
DL
TP P DL AF FCH
FCH
UL
TP P DL AF FCH + 2
FCH
UL
TP P DL AF FCH + 4
FCH
UL
TP P DL AF FCH + 8
FCH
UL
FCH
DL
FCH
DL
FCH
DL
FCH
DL
TP P DL AF FCH + 16
Atoll determines the distribution of users with the different possible throughputs. A random trial compliant with throughput
probabilities is performed for each link in order to determine the throughput of each user.
On uplink, we have,
rk
FCH
UL
FCH
DL
For each SCH throughput factor, r k , the number of users n j with the throughput TP P UL AF FCH + r k is calculated as
follows,
r
UL
nj = Pr nj
k
FCH

FCH
nj
= nj
FCH
UL
with the throughput, TP P UL AF FCH , is,
rk
nj
rk
On downlink, we have,
rk
For each SCH throughput factor, r k , the number of users, n j with the throughput, TP P DL AF FCH + r k , is calculated as
follows,
k
DL
nj = Pk nj
FCH

FCH
nj
= nj
FCH
DL
with the throughput, TP P DL AF FCH , is,
rk
nj
rk
CDMA2000 1xEV-DO Services

As power control is performed in the uplink only, 1xEV-DO data service users will be considered either active in the uplink or
inactive. 1xEV-DO data Rev. 0 service users can access the service with uplink throughputs of 9.6, 19.2, 38.4, 76.8 and 153.6
kbps. 1xEV-DO data Rev. A and Rev. B service users can access the service with uplink throughputs of 4.8, 9.6, 19.2, 38.4, 76.8,
115.2, 153.6, 230.4, 307.2, 460.8, 614.4, 921.6, 1,228.8 and 1,848.2 kbps.
UL
UL
For each service, j, several throughput probabilities, P k , can be assigned to different uplink throughputs TP k . The number
of users active in uplink ( n j UL ) and the number of inactive users ( n j inactive ) are calculated into several steps. First of
all, Atoll determines the number of users active in UL using the service j in the Txi cell.
363

For each transmitter, Txi, and each service j:
each sector.
UL
) for
Atoll calculates the number of users active in UL using the service j in the Txi cell as follows:
UL
N UL = TPD
--------------UL
TP j
UL
Where TPD is the number of kbits per second transmitted on UL in the Txi cell to provide the service j (userdefined value in the traffic map properties).
UL
TP j
UL
TP j
corresponds to the uplink throughput for a user.

=
Pk
UL
UL
TP k
In the above calculations, we assume that the sum of throughput probabilities is less than
or equal to 1. If the sum of throughput probabilities exceeds 1, Atoll considers
normalised throughput probabilities values, P r P r , instead of specified

k
k
rk
throughput probabilities P r .
k
We have the following activity probabilities:

Probability of being active in UL: p UL =
Pk
UL
UL
TP k
UL
Rk
Pk
UL
UL
TP k
UL
Rk

Therefore, we have:
Number of users active in UL: n j UL = N UL p UL
Number of inactive users: n j inactive = N UL p inactive
Number of users active in DL: n j DL = 0
Number of users active in UL and DL both: n j UL + DL = 0
Total number of connected users: n j = n j UL + n j inactive
We have the following activity probabilities:
Probability of being active in UL: p UL =
Pk
UL
UL
TP k
UL
Rk
Pk
UL
UL
TP k
UL
Rk

364

AT332_TRR_E0
Therefore, we have:
Number of users active in UL: n j UL = n j p UL
Number of inactive users: n j inactive = n j p inactive
Number of users active in DL: n j DL = 0
Number of users active in UL and DL both: n j UL + DL = 0
When selecting Number of Users per Activity Status, you can directly input the number of inactive users
( n j inactive ) and the number of users active in the uplink ( n j UL ), for each sector.
The total number of connected users ( n j ) is calculated as follows
n j = n j UL + n j inactive
Then, Atoll determines the distribution of users with the different possible throughputs. The number of users with the
UL
UL
throughput TP k , n j TP k , is calculated as follows:

UL
UL
n j TPk = P k n j
Inactive users have a requested throughput equal to 0.
The user distribution per service is an average distribution and the service of each user is
randomly drawn In each simulation. Therefore, if you compute several simulations at
once, the average number of users per service will correspond to the calculated
distribution. But if you check each simulation, the user distribution between services is
different in each of them.
It is the same for the SCH throughput distribution between 1xRTT data service users and
the traffic throughput distribution between 1xEV-DO data service users.
5.4.1.2 Transition Flags for 1xEV-DO Rev.0 User Throughputs

For 1xEV-DO Rev. 0 services supporting throughput downgrading, you can define the probability of the service being upgraded
UL
UL
UL
UL
UL
( P Upg k TP k ) or downgraded ( P Downg k TP k ) on the uplink (reverse link) for each throughput ( TP k ). The probabilities
are taken into account in order to determine if a user with a certain throughput can be upgraded or downgraded. User
throughput downgrading and upgrading occur during congestion control when the cell is over- or underloaded.
The following table shows the throughput changes that are possible when a throughput is upgraded or downgraded. The
probabilities are defined with a number from 1 to 255 for each throughput.
Possible Throughput Changes
During Upgrading
Possible Throughput Changes

During Downgrading
From
To
From
To
9.6 kbps
19.2 kbps
153.6 kbps
76.8 kbps
19.2 kbps
38.4 kbps
76.8 kbps
38.4 kbps
38.4 kbps
76.8 kbps
38.4 kbps
19.2 kbps
76.8 kbps
153.6 kbps
19.2 kbps
9.6 kbps
During the generation of the user distribution, each 1xEV-DO Rev. 0 user is assigned a random number between 1 and 255 for
each possible throughput transition. When this number is lower or equal to the value of the probability, the transition flag for
this throughput transition is set to "True" meaning that this throughput transition can be performed if necessary.
UL
The number of 1xEV-DO Rev. 0 users with a certain throughput that can be downgraded ( n j TP k Downg ) and upgraded
UL
( n j TP k Upg ) are calculated as follows:

UL
UL
UL
P Upg k TP k n j TP k
UL
n j TPk Upg = -----------------------------------------------------------255
And
365

UL

UL
UL
P Downg k TP k n j TP k
UL
n j TP k Downg = -----------------------------------------------------------------255
The number of users with a certain throughput that can be downgraded or upgraded is
an average. Therefore, if you compute several simulations at once, the average number
of users with a certain throughput that can be downgraded or upgraded will correspond
to the calculated value. But if you check each simulation, this number is different in each
of them.
5.4.1.3 User Geographical Position

Once all the user characteristics determined, another random trial is performed to obtain their geographical positions and
whether they are indoors or outdoors according to the percentage of indoor users per clutter class defined for the traffic
maps.
5.4.2 Network Regulation Mechanism

5.4.2.1 CDMA2000 1xRTT Power Control Simulation Algorithm
CDMA2000 1xRTT network automatically regulates itself using traffic driven uplink and downlink power control on the
fundamental and supplemental channels (FCH and SCH respectively) in order to minimize interference and maximize capacity.
Atoll simulates this network regulation mechanism with an iterative algorithm and calculates, for each user distribution,
network parameters such as base station power, mobile terminal power, active set and handoff status for each terminal.
The power control simulation is based on an iterative algorithm, where in each iteration, all the mobiles selected during the
user distribution generation (1st step) try to connect to network active transmitters with a calculation area. The process is
repeated from iteration to iteration until convergence is achieved. The algorithm steps are detailed below.
Figure 5.1: CDMA2000 1xRTT Power Control Algorithm
5.4.2.1.1
Algorithm Initialization
Total power on carrier ic, P Tx ic , of base station Sj is initialised to P pilot ic + P sync ic + P paging ic .
UL
intra
Uplink received powers on carrier ic, I tot

(no connected mobile).
366
UL
extra
ic , I tot
UL
ic and I inter carrier ic , at base station Sj are initialised to 0 W

AT332_TRR_E0
UL
I tot S j ic
UL
X k S j ic = ------------------------- = 0
UL
N tot S j ic
5.4.2.1.2
Presentation of the Algorithm

UL
The algorithm is detailed for any iteration k. Xk is the value of the variable X at the iteration k. In the algorithm, all Q req and
DL
Q req thresholds depend on user mobility type and are defined in Service and Mobility parameters tables. All variables are
described in Definitions and formulas part.
The algorithm applies to single frequency band networks and to multi-band networks (dual-band and tri-band networks).
Multi-band terminals can have the following configurations:
Configuration 1: The terminal can work on f1, f2 and f3 without any priority (select "All" as main frequency band
in the terminal property dialog box).
Configuration 2: The terminal can work on f1, f2 and f3 but f1 has a higher priority (select "f1" as main frequency
band, "f2" as secondary frequency band and "f3" as third frequency band in the terminal property dialog box).
For each mobile (Mi), Atoll only considers the cells (Sj,ic) for which the pilot RSCP exceeds the minimum pilot RSCP:
P c Sj M i ic RSCP min Sj ic .
For each mobile Mi, we have the following steps:
Determination of Mis Best Serving Cell
For each transmitter Sj containing Mi in its calculation area and working on the main frequency band supported by the Mis
terminal (i.e. either f1 for a single frequency band network, or f1, f2 or f3 for a multi-band terminal with the configuration 1,
or f1 for a multi-band terminal with the configuration 2).
BTS P c Sj M i ic
Calculation of Q pilot Sj ic M i = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------Term
k
DL
DL
DL
DL
P tot Sj ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0
Determination of the candidate cells, (SBS,ic).
For each carrier ic, selection of the transmitter with the highest Q pilot Sj M i ic , S BS ic M i .
k
Analysis of candidate cells, (SBS,ic).

For each pair (SBS,ic), calculation of the uplink load factor:
UL
I tot S BS ic
UL
UL
X k S BS ic = ----------------------------+ X
UL
N tot S BS ic
pilot
If Q pilot S BS M i ic Q req then (SBS,ic) is rejected by Mi

k
UL
UL
If X k S BS ic X max , then (SBS,ic) is rejected by Mi

Else
Keep (SBS,ic) as good candidate cell
For multi-band terminals with the configuration 1 or terminals working on one frequency band only, if no good candidate cell
has been selected, Mi has failed to be connected to the network and is rejected.
For multi-band terminals with the configuration 2, if no good candidate cell has been selected, try to connect Mi to
transmitters txi containing Mi in their calculation area and working on the secondary frequency band supported by the Mis
terminal (i.e. f2). If no good candidate cell has been selected, try to connect Mi to transmitters txi containing Mi in their
calculation area and working on the third frequency band supported by the Mis terminal (i.e. f3). If no good candidate cell
Determination of the best carrier, icBS.
If a given carrier is specified for the service requested by Mi
ic BS M i is the carrier specified for the service
367

UL
ic BS M i is the cell with the lowest X k S BS ic

ic BS M i is the cell with the lowest P tx S BS ic k
ic BS M i is randomly selected
UL
UL
ic BS M i is the first carrier where X k S BS ic X max

Endif
Determination of the best serving cell, (SBS,icBS).
max
(S BS,ic BS) k M i is the best serving cell ( BestCell k M i ) and its pilot quality is Q pilot M i .
k
In the following lines, we will consider ic as the carrier used by the best serving cell.
Determination of the Active Set
For each station Sj containing Mi in its calculation area, using ic, and if neighbours are used, neighbour of BestCell k M i
BTS P c M i S j
Calculation of Q pilot M i S j ic = ------------------------------------------------DL
k
I 0 ic
Rejection of station Sj if the pilot is not received
pilot
If Q pilot M i S j ic Q min then Sj is rejected by Mi

k
Else Sj is included in the Mi active set
Rejection of Sj if the Mi active set is full
k
EndFor
req
Calculation of the required power for Mi, P term M i ic k

For each cell (Sj,ic) present in the Mi active set
Calculation of quality level on Mi traffic channel at (Sj,ic), with the minimum power allowed on traffic channel for the Mi
service
FCH U L
Pb
FCH r eq
SCH r eq
P term M i ic k 1
P term M i ic k 1
SCH U L
- and P b
M i S j ic = --------------------------------------------- M i S j ic = ---------------------------------------------L T M i S j
L T M i S j
FCH U L
term P b
M i S j ic
UL
UL
- G FCH
Q FCH M i S j ic k = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Service
p
UL
FCH U L
SCH U L
N tot ic 1 F MUD term P b
M i S j ic + P b
M i S j ic
SCH U L
term P b
M i S j ic
UL
UL
- G SCH
Q SCH M i S j ic k = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Service
p
UL
FCH U L
SCH U L
N tot ic 1 F MUD term P b
M i S j ic + P b
M i S j ic
If the user selects the option Total noise
FCH U L
term P b
M i S j ic
UL
UL
- G FCH
Q FCH M i S j ic k = ------------------------------------------------------------- Service
p
UL
N tot ic
368

AT332_TRR_E0
SCH U L
term P b
M i S j ic
UL
UL
- G SCH
Q SCH M i S j ic k = ------------------------------------------------------------- Service
p
UL
N tot ic
End For
If (Mi is not in handoff)
UL
UL
UL
UL
Q FCH M i = Q FCH M i S j ic k and Q SCH M i = Q SCH M i S j ic k

k

UL
UL
Q FCH M i = f rake efficiency

k
UL
UL
Q FCH M i S j ic k
S j ActiveSet
UL
Q SCH M i = f rake efficiency

k
UL
Q SCH M i S j ic k
S j ActiveSet
Else if (Mi is in soft or softer/soft without MRC)

UL
UL
UL
Q FCH M i = G macro diversity 2 links Max Q FCH M i S j ic k

k
S j ActiveSet
UL
UL
UL
Q SCH M i = G macro diversity 2 links Max Q SCH M i S j ic k

k
S j ActiveSet
Else if (Mi is in soft/soft)

UL
UL
UL
Q FCH M i = G macro diversity 3 links Max Q FCH M i S j ic k

k
UL
S j ActiveSet
UL
UL
Q SCH M i = G macro diversity 3 links Max Q SCH M i S j ic k

k
S j ActiveSet
Else if (Mi is in softer/soft with MRC)
UL
UL
UL
UL
Q FCH M i = Max f rake efficiency
Q FCH ic Q FCH
k
other site
i AS ActiveSet
UL
ic G macro diversity 2 links
UL
UL
UL
UL
Q SCH M i = Max f rake efficiency
Q SCH ic Q SCH
k
other site
i AS ActiveSet
UL
ic G macro diversity 2 links
(same site)
(same site)
EndIf
FCH r eq
P term
UL
Q req Service M i Term M i Mobility M i FCH

r eq
- P FCH
M i ic k = -------------------------------------------------------------------------------------------------------------------------- M i ic k 1
term
UL
Q FCH M i
k
SCH r eq
P term
UL
Q req Service Mi Term M i Mobility M i SCH_rate_multiple SCH

r eq
- P SCH
M i ic k = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------- M i ic k 1
term
UL
Q SCH M i
k
req
FCH r eq
P term M i ic k = P term
req
SCH r eq
M i ic k + P term
M i ic k
min
If P term M i ic k P term M i then

FCH r eq
P term
SCH r eq
P term
min
P term M i S j
r eq
- P FCH
M i ic k = ----------------------------- M i ic k
term
req
P term M i k
min
P term M i S j
r eq
- P SCH
M i ic k = ----------------------------- M i ic k
term
req
P term M i k
EndIf
369

FCH r eq
If P term
max
M i ic k P term M i then Mi cannot select any station and its active set is cleared
req
max
If P term M i ic k P term M i and Mi uses SCH then:

Downgrading the service SCH throughput:
req
max
SCH
FCH
While P term M i ic k P term M i and TP P UL Service M i TPP UL Service M i 2

SCH
TP P UL Service M i
SCH
TP P UL Service M i ----------------------------------------------------2
SCH r eq
SCH r eq
P term M i
UL
SCH
P term M i ic
Q req Service M i Term M i Mobility M i TP P UL Service Mi SCH
ic k = -----------------------------------------k ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------UL
SCH
2
Q Service M Term M Mobility M TP
Service M 2
req
FCH r eq
req
SCH r eq
M i ic k + P term
P UL
SCH
M i ic k
EndWhile
req
max
If P term M i ic k P term M i then Mi will not use SCH

Endif
Endif
If the required number of channel elements exceeds the available quantity in the site of Sj (Best server of Mi) and Mi uses SCH
then:
Downgrading the service SCH throughput:
Max
SCH
FCH
While N CE U L M i N CE U L S j and TP P UL Service M i TP P UL Service M i 2

SCH
TP P UL Service M i
SCH
TP P UL Service M i ----------------------------------------------------2
SCH
N CE U L M i k
SCH
N CE U L M i k = ----------------------------2
SCH r eq
SCH r eq
P term
SCH UL
SCH
P term M i ic
Service M i Term M i Mobility M i TP P UL Service Mi
Q req
M i ic k = -----------------------------------------k -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------SCH UL
SCH
2
Q
Service M Term M Mobility M TP
Service M 2
req
FCH r eq
req
SCH r eq
M i ic k + P term
SCH
P UL
M i ic k
FCH
N CE U L M i k = N CE U L M i k + N CE U L M i k
EndWhile
Endif
If Mi uses an SCH on the downlink
For each cell (Sj,ic) in Mi FCH active set
Calculation of quality level on (Sj,ic) FCH at Mi, with the minimum power allowed on FCH for the Mi service
FCH D L
Pb
min
P FCH Service M i
M i S j ic = ---------------------------------------------L T M i S j
FCH D L
BTS P b
M i S j
DL
DL
- G FCH
Q FCH M i S j ic k = ------------------------------------------------------------------------------------------------------------ Service M i
p
DL
DL
N tot ic 1 F ortho BTS P b M i S j ic
FCH D L
BTS P b
M i S j
DL
Q FCH M i S j ic k = ----------------------------------------------------DL
N tot ic
370

AT332_TRR_E0
If cell (Sj,ic) in Mi SCH active set

Calculation of quality level on (Sj,ic) SCH at Mi, with the minimum power allowed on SCH for the Mi service
SCH D L
Pb
min
P SCH Service M i
M i S j ic = ---------------------------------------------L T M i S j
SCH D L
BTS P b
M i S j
DL
DL
- G SCH
Q SCH M i S j ic k = ------------------------------------------------------------------------------------------------------------ Service Mi
p
DL
DL
N tot ic 1 F ortho BTS P b M i S j ic
SCH D L
BTS P b
M i S j
DL
Q SCH M i S j ic k = ----------------------------------------------------DL
N tot ic
EndIf
End For
Recombination of the first f active set links (f is the number of fingers of the Mi terminal): only quality levels from the first f
cells (Sf,ic) of active set are recombined.
DL
DL

k
Q FCH M i S j ic k
Q SCH M i S j ic k
DL
S f ActiveSet FCH
DL
DL

k
DL
S f ActiveSet SCH
Do
For each cell (Sj,ic) in Mi FCH active set
Calculation of the required power for DL traffic channel between (Sj,ic) and Mi:
DL
FCH
Q req Service M i Term M i Mobility M i TP P DL Service M i FCH

req
- P min
P FCH M i S j ic k = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------FCH Service M i
DL
Q FCH M i
k
req
max
If P FCH M i S j ic k P FCH Service M i then S j ic is excluded from Mi active set

DL
Recalculation of a decreased Q req

If cell (Sj,ic) in Mi SCH active set
Calculation of the required power for DL traffic channel between (Sj,ic) and Mi:
DL
SCH
Q req Service M i Term M i Mobility M i TP P DL Service M i SCH

req
- P min
P SCH M i S j ic k = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------SCH Service M i
DL
Q SCH M i
k
Downgrading the service SCH throughput (only for (Sj,ic) best server cell of Mi):
req
max
SCH
While P SCH M i S j ic k P SCH Service M i TP P DL Service M i

req
SCH
FCH
Or P tx S j ic k + P tch M i S j ic k P max S j ic and TPP DL Service M i TP P DL Service M i 2

SCH
TP P DL Service M i
SCH
TP P DL Service M i = ----------------------------------------------------2
req
DL
SCH
P SCH M i S j ic k
Q req Service M i Term M i Mobility M i TP P DL Service M i SCH
req
P SCH M i S j ic k = ------------------------------------- ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------DL
SCH
2
Service M 2
req
req
req
P DL
SCH
req
P tch M i S j ic k = P SCH M i S j ic k + P FCH M i S j ic k

EndWhile
req
max
req
If P SCH M i S j ic k P SCH Service M i or P tx S j ic k + P tch M i S j ic k P max S j ic then Mi will not use SCH
371

Endif
Max
SCH
FCH
While N CE D L M i N CE D L S j and TP P DL Service M i TP P DL Service Mi 2

SCH
TP P DL Service M i
SCH
TP P DL Service M i = ----------------------------------------------------2
SCH
N CE D L M i k
SCH
N CE D L M i k = ----------------------------2
req
DL
SCH
P SCH M i S j ic k
Q req Service M i Term M i Mobility M i TPP DL Service M i SCH
req
P SCH M i S j ic k = ------------------------------------- ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------DL
SCH
2
Service M 2
req
req
req
P DL
SCH
req

FCH
SCH
N CE D L M i k = N CE D L M i k + N CE D L M i k
EndWhile
Max
If N CE D L M i N CE D L S j then Mi will not use SCH

Endif
Max
SCH
FCH
While N Codes M i N Codes S j ic and TP P DL Service M i TP P DL Service Mi 2

SCH
TP P DL Service M i
SCH
TP P DL Service M i = ----------------------------------------------------2
SCH
N Codes M i k
SCH
N Codes M i k = ---------------------------2
req
DL
SCH
P SCH M i S j ic k
Q req Service M i Term M i Mobility M i TPP DL Service M i SCH
req
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------P SCH M i S j ic k = -------------------------------------DL
SCH
2
Service M 2
req
req
req
P DL
SCH
req

FCH
SCH
N Codes M i k = N Codes M i k + N Codes M i k

EndWhile
Max
If N Codes M i N Codes S j ic then Mi will not use SCH

Endif
Endif
EndFor
Recombination of the first f active set links (f is the number of fingers of the Mi terminal): only quality levels from the first f
cells (Sf,ic) of active set are recombined.
DL
DL

k
Q FCH M i S f ic k
Q SCH M i S f ic k
DL
S f ActiveSet FCH
DL
DL

k
DL
S f ActiveSet SCH
DL
DL
While Q k M i Q req Service M i Mobility M i and Mi FCH active set is not empty
DL
DL
And Q k M i Q req Service M i Mobility M i (if SCH active set is not empty)
Endif
Uplink and Downlink Interference Updates
Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones)
372

AT332_TRR_E0
For each cell (Sj,ic)

UL
Update of N tot S j ic
EndFor
For each mobile Mi
DL
Update of N tot ic
EndFor
Control of Radio Resource Limits (Walsh Codes, Cell Power and Site Channel Elements)
For each cell (Sj,ic) on a site Nl
P tx S j ic
DL
While -------------------------k %Power max
P max
req
Rejection of mobile with highest P tch S j M b ic k for the lowest service priority
EndWhile
EndFor
For each site Nl
The list of rejected mobiles for the site Nl is L rejected N l
If the equipment installed on Nl supports power pooling between transmitters
Activation of power pooling between transmitters for each cell (Sj,ic) containing rejected users
Control of the available power for the other cells (Si,ic) of the site where power pooling between transmitters is not activated
If
DL
%Power max P max P tx S i ic k 0
S ic
i
Si Nl
Then, the power unused by the cells (Si,ic) of the site can be allocated to cells (Sj,ic)
Sort of all the rejected mobiles by priority in a descending order and by simulation rank in a descending order
For the first mobile Mb of the list ( M b L rejected N l )
req
DL
If P tx S j ic k + P tch S j M b ic k %Power max P max + M Pooling S j ic

Mb is reconnected
EndIf
EndFor
EndIf
EndFor
Max
While N Codes S j ic k N Codes S j ic

Rejection of last admitted mobile
EndFor
For each site (Node B) Nl
Max
While N CE DL N I k N CE DL N I
req
Rejection of mobile with highest P tch M i S j k for the lowest service priority
Max
While N CE UL N I k N CE UL N I
req
Rejection of mobile with highest P term M i ic k for the lowest service priority
373

EndFor
UL
UL
For each cell (Sj,ic) with X S j ic X max

Rejection of a mobile with the lowest service priority
EndFor
UL
UL
While at least one cell with X S j ic X max exists
5.4.2.1.3
Convergence Criterion
The convergence criteria are evaluated at each iteration, and can be written as follow:
DL
DL

P tx ic k P tx ic k 1
DL = max int ma x ------------------------------------------------ 100 int ma x ----------------------------------------------------------- 100
DL
P tx ic k
Stations
Stations
N
ic
user
UL
UL
UL
UL

UL = max int ma x ------------------------------------------------- 100 int ma x ----------------------------------------------------------- 100
UL
UL
Stations
Stations

I ic
N
ic
tot
user

1st case: Between two successive iterations, UL and DL are lower ( ) than their respective thresholds (defined when
creating a simulation).
Example: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5. If
UL 5 and DL 5 between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th iteration. Convergence has
been achieved.
2nd case: After 30 iterations, UL or/and DL are still higher than their respective thresholds and from the 30th iteration, UL
or/and DL do not decrease during the next 15 successive iterations.
Examples: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5.
1. After the 30th iteration, UL and/or DL equal 100 and do not decrease during the next 15 successive iterations: Atoll stops
the algorithm at the 46th iteration. Convergence has not been achieved.
2. After the 30th iteration, UL and/or DL equal 80, they start decreasing slowly until the 40th iteration (without going under
the thresholds) and then do not change during the next 15 successive iterations: Atoll stops the algorithm at the 56th iteration
without achieving convergence.
If UL and/or DL are still strictly higher than their respective thresholds, the simulation has not converged (specific
divergence symbol).
If UL and DL are lower than their respective thresholds, the simulation has converged.
374

AT332_TRR_E0
5.4.2.2 CDMA2000 1xEV-DO Power/Data Rate Control Simulation Algorithm
Figure 5.2: CDMA2000 1xEVDO Power Control Algorithm

In a CDMA2000 1xEV-DO system, power control is performed in the uplink only. In the downlink, the transmitter transmits at
the full power (Pmax) when a connection is established. Instead of power control, there is a data rate control based on the C/
I ratio calculated at the mobile. For each distribution of users, Atoll simulates the power control mechanism for the UL and
the data rate control for the DL.
The simulation uses an iterative algorithm, where in each iteration, all the 1xEV-DO data service users selected during the user
distribution generation (1st step) try to connect to network active transmitters with a calculation area. Atoll considers the
guaranteed bit rate service users first, in the order established during the generation of the user distribution, and then, it
processes the variable bit rate service users, in the order established during the generation of the user distribution.
The process is repeated from iteration to iteration until convergence is achieved. The algorithm steps are detailed below.
5.4.2.2.1
Algorithm Initialization
UL
intra
Uplink received powers on carrier ic, I tot
UL
extra
ic , I tot
UL
ic and I inter carrier ic , at base station Sj are initialised to 0 W
(no connected mobile).

UL
I tot S j ic
UL
X k S j ic = ------------------------- = 0
UL
N tot S j ic
5.4.2.2.2
Presentation of the Algorithm

The algorithm is detailed for any iteration k. Xk is the value of the variable X at the iteration k.
UL
Ec
In the algorithm, -----
is the minimum pilot quality level required in the uplink to operate 1xEV-DO Rev. 0. This
N t min Rev0
threshold depends on the user mobility type and is defined in the Mobility parameters table.
Ec
--- N t min RevB is the minimum pilot quality level required in the uplink to operate EV-DO multi-carrier. This threshold is
UL
defined in the Transmitter properties dialog box.

E c UL
For 1xEV-DO Rev. A and Rev. B users, the value of -----
depends on the user requested throughput. This throughput can
N t min
be obtained by using a certain uplink 1xEV-DO radio bearer ( Index UL Bearer ) in a certain number of subframes ( n SF ).
Ec
---is the value defined in the 1xEV-DO Radio Bearer Selection (Uplink) table for the combination (radio bearer Index,
N t min
UL
375

mobility and number of subframe) providing the user requested throughput. Two values are available for this parameter, one
when the service uplink mode is "Low Latency" and another one for high capacity services.
All variables are described in Definitions and formulas part (see "Definitions and Formulas" on page 338).
The algorithm applies to single frequency band networks and to multi-band networks (dual-band and tri-band networks).
Multi-band terminals can have the following configurations:
Configuration 1: The terminal can work on f1, f2 and f3 without any priority (select "All" as main frequency band
in the terminal property dialog box).
Configuration 2: The terminal can work on f1, f2 and f3 but f1 has a higher priority (select "f1" as main frequency
band, "f2" as secondary frequency band and "f3" as third frequency band in the terminal property dialog box).
For each mobile (Mi), Atoll only considers the cells (Sj,ic) for which the pilot RSCP exceeds the minimum pilot RSCP:
P c Sj M i ic b pilot RSCP min Sj ic .
For each mobile Mi, we have the following steps:
Determination of Mis Best Serving Cell
For each transmitter Sj containing Mi in its calculation area and working on the main frequency band supported by the Mis
terminal (i.e. either f1 for a single frequency band network, or f1, f2 or f3 for a multi-band terminal with the configuration 1,
or f1 for a multi-band terminal with the configuration 2).
BTS P c Sj M i ic ,b pilot
Calculation of Q pilot Sj ic M i = -----------------------------------------------------------------------------------------------------------------------------------------------------------------Term
k
DL
DL
DL
P tot Sj ic ,b pilot + I extra ic ,b pilot + I inter carrier ic ,b pilot + N 0
Determination of the candidate cells, (SBS,ic).
For each carrier ic, selection of the transmitter with the highest Q pilot Sj M i ic , S BS ic M i .
k
Analysis of candidate cells, (SBS,ic).

For each pair (SBS,ic), calculation of the uplink load factor:
UL
I tot S BS ic
UL
UL
X k S BS ic = ----------------------------+ X
UL
N tot S BS ic
pilot
If Q pilot S BS M i ic Q req then (SBS,ic) is rejected by Mi

k
UL
UL
If X k S BS ic X max , then (SBS,ic) is rejected by Mi

Else
Keep (SBS,ic) as good candidate cell
For multi-band terminals with the configuration 1 or terminals working on one frequency band only, if no good candidate cell
For multi-band terminals with the configuration 2, if no good candidate cell has been selected, try to connect Mi to
transmitters txi containing Mi in their calculation area and working on the secondary frequency band supported by the Mis
terminal (i.e. f2). If no good candidate cell has been selected, try to connect Mi to transmitters txi containing Mi in their
calculation area and working on the third frequency band supported by the Mis terminal (i.e. f3). If no good candidate cell
Determination of the best carrier, icBS.
If a given carrier is specified for the service requested by Mi
ic BS M i is the carrier specified for the service
UL
ic BS M i is the cell with the lowest X k S BS ic

376

AT332_TRR_E0
ic BS M i is the cell with the lowest P tx S BS ic k

ic BS M i is randomly selected
UL
UL
ic BS M i is the first carrier where X k S BS ic X max

Endif
Determination of the best serving cell, (SBS,icBS).
max
(S BS,ic BS) k M i is the best serving cell ( BestCell k M i ) and its pilot quality is Q pilot M i .
k
In the following lines, we will consider ic as the carrier used by the best serving cell.
Determination of the Active Set
For each station Sj containing Mi in its calculation area, using ic, and if neighbours are used, neighbour of SBS(Mi)
DL
BTS P tot M i S j ic b pilot

Calculation of Q pilot M i S j ic = ---------------------------------------------------------------------------DL
k
I 0 ic b pilot
min
If Q pilot M i S j ic Q pilot then Sj is rejected by Mi

k
Else Sj is included in the Mi active set

Rejection of Sj if the Mi active set is full
k
EndFor
Determination of the Sub-active Sets of a EVDO Multi-carrier User
For multi-carrier EV-DO Rev.B service users with a 1xEV-DO Rev. B capable terminal, calculation of the quality level received
by the best serving cell (SBS,ic)
Ec
--- N t
UL
max
term P term M i
S BS ic = -----------------------------------------UL
L T N tot S BS ic
E UL
E UL
If ----c- S BS ic ----c- S BS then EV-DO multi-carrier is not activated.
Nt
N t min
For each transmitter Sj containing Mi in its calculation area and using other EV-DO carriers, icn (either icn belongs to f1 for a
single frequency band network, or it belongs to f1, f2 or f3 for a multi-band terminal)
Calculation of Q pilot Sj ic n M i
k
Ranking of carriers, icn,according to Q pilot Sj ic n M i , from the highest to the lowest value.
k
For each received carrier, icn, in the defined order:

carriers
While n max
M i is not exceeded
Determination of the best transmitter of the sub-active set, based on the received pilot quality, Q pilot Sj ic n M i .
k
Determination of the other transmitters of the sub-active set, based on the received pilot quality, Q pilot Sj ic n M i .
k
Calculation of the quality level received by the best serving cell (SBS,icn)
Ec
--- N t
UL
max
term P term M i
S BS ic n = -----------------------------------------UL
L T N tot S BS ic n
377

E UL
E UL
If ----c- S BS ic n ----c- S BS , then no sub-active set is associated with icn
N t
N t min
If the user terminal supports the Locked mode, analysis of the sub-active set
If a transmitter of the studied sub-active set does not belong to the sub-active set associated with the best carrier, then it is
removed.
If the studied sub-active set does not contain the same transmitters as the sub-active set associated with the best carrier, then
the studied sub-active set is removed.
EndIf
Endwhile
EndFor
req
Calculation of the required power for Mi, P term M i ic k

For each cell (Sj,ic) present in the Mi active set or sub-active set
Calculation of quality level on Mi traffic channel at (Sj,ic), with the minimum power allowed on traffic channel for the Mi
service
req
P term M i ic k 1
UL
P b M i S j ic = --------------------------------------L T M i S j
UL
term P b M i S j ic
- G UL
Q M i S j ic k = -------------------------------------------------------------------------------------------------------------p Service
UL
Tx
UL
N tot ic 1 F MUD term P b M i S j ic
UL

UL
term P b M i S j ic
UL
- G UL
Q M i S j ic k = ---------------------------------------------------p Service
UL
N tot ic
End For
If (Mi is not in handoff)
UL
UL
Q total M i = Q M i S j ic
k

UL
UL
Q total M i = f rake efficiency

k
UL
Q M i S j ic k
S j ActiveSet
Else if (Mi is in soft or softer/soft without MRC)

UL
Q total M i =
k
UL
UL
Max Q M i S j ic k G macro diversity 2 links
I AS ActiveSet
Else if (Mi is in soft/soft)

UL
Q total M i =
k
UL
UL
Max Q M i S j ic k G macro diversity 3 links
I AS ActiveSet
Else if (Mi is in softer/soft with MRC)

UL
Q total M i
k
UL
UL
UL
= Max f rake efficiency
Q M i S j ic k Q M i S j ic k
othersite
i AS ActiveSet
(same site)
G UL
macro diversity 2 links
EndIf
UL
Q req Service M i Term M i Mobility M i

req
- P req
P term M i ic k = --------------------------------------------------------------------------------------------------------------term M i ic k 1
UL
Q total M i
k
378

AT332_TRR_E0
If the service of Mi uses Transmission Control Protocol (TCP)

For the best server cell (Sk,ic) of Mi
Calculation of the Mi downlink application throughput
DL
Calculation of N tot ic b traffic
Ptot txj icadj btraffic

DL
DL
txj j
- + N0
Ptot Sj i c btraffic + ---------------------------------------------------------------RF ic ic adj
DL
N tot ic b traffic =
term
j j k
Calculation of the maximum throughput supplied to Mi, TP max DL M i S k

Calculation of pilot quality level at Mi
DL
E
P tot M i S k ic b pilot
----c- M i S k ic b pilot = -------------------------------------------------DL
Nt
N tot ic b pilot
If Mi is a 1xEV-DO Rev. 0 service user, determination of the peak throughput from the graph (Peak throughput=f(C/I)) specified
for the mobility type of Mi
E
TP max DL M i S k = f ----c- M i S k ic b pilot
Nt
If Mi is a 1xEV-DO Rev. A service user, selection of the downlink 1xEV-DO radio bearer ( Index DL Bearer ): Index DL Bearer
DL
E
E
where ----c- M i S k ic b pilot ----c- Index DL Bearer
Nt
min
Nt
If Mi is a 1xEV-DO Rev. B service user, selection of the downlink 1xEV-DO radio bearer ( Index DL Bearer ): Index DL Bearer
DL
E
E
where ----c- M i S k ic b pilot ----c- Index DL Bearer
and the modulation scheme is supported by the terminal.
Nt
min
Nt
DL
Determination of the peak throughput: TP max DL M i S k = -----------------------------------------------------------n TS Index DL Bearer
DL
TP A M i S k = TP max DL M i S k f TP Scaling Service Mi TP Offset Service Mi

UL
Determination of the uplink throughput due to TCP acknowledgements, TP TCP ACK M i S k from the graph (UL Thr due to
TCP=f(DL Thr) specified for the service of Mi
UL
DL
TP TCP ACK M i S k = f TP A M i S k
UL
UL
UL
Determination of the nearest lower and higher supported throughputs ( TP low and TPhigh ) for TP TCP ACK M i S k
UL
UL
UL
UL
UL
UL
For TPlow and TP high , calculation of CI req TP low and CI req TP high
E
UL
CI req = ----c-
1 + G DRC + G TCH for DO Rev.0 terminals
N t min
UL
And
E UL
UL
CI req = ----c-
1 + G DRC + G TCH + G RRI + G Auxiliary pilot for DO Rev.A and DO Rev.B terminals
N t min
EndFor
UL
UL
UL
UL
UL
UL
UL
UL
Linear interpolation of CI req TP TCP ACK between CI req TP low and CI req TP high
UL
UL
UL
CI req = CI req TP + CI req TP TCP ACK

W
UL
UL
Q req = CI req ----------------------------------------------UL
UL
TP + TPTCP ACK
EndIf
379

req
req
min
P term M i ic k = Max P term M i ic k P term M i S j

For DO Rev.0 and DO Rev.A users
req
max
If P term M i ic k P term M i then:

Downgrading the traffic channel throughput
req
max
While P term M i ic k P term M i

And
UL
TP Service M i 9,6kbps for 1xEV-DO Rev. 0 users,

UL
TP Service M i 4,8kbps for (1xEV-DO Rev. A - Variable bit rate) service users,
UL
TP Service M i 4,8kbps for single-carrier 1xEV-DO Rev. B service users,

UL
TP Service M i TPD min UL Service M i for (1xEV-DO Rev. A - Guaranteed bit rate) service users,
req
P term M i ic k
req
- TP UL
P term M i ic k = ---------------------------------------------low Service M i
UL
TP Service M i
UL
( TP low Service M i
is the nearest lower supported
throughput)
For 1xEV-DO Rev. 0, (1xEV-DO Rev. A - Variable bit rate) and single-carrier 1xEV-DO Rev. B service users,
UL
UL
TP Service M i = TP low Service M i

UL
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, TP Service M i = TPD min UL Service M i
EndWhile
req
max
If P term M i ic k P term M i then Mi is rejected

req
P term M i ic = P term M i ic k
req
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, P term M i ic = P term M i ic k C UL Bearer
Endif
Endif
For multi-carrier 1xEV-DO Rev. B service users, load balancing between carriers is performed. The available terminal power is
shared between each carrier as follows:
The maximum terminal power is allocated to the best carrier ( ic 1 ).
UL
Calculation of the traffic channel throughput ( TP Service M i ic 1 )

UL
Downgrading the traffic channel throughput ( TP Service M i ic 1 )
While
req
max
P term M i ic 1 k P term M i
UL
and TP Service M i ic 1 153 6kbps
req
P term M i ic 1 k
req
UL
- TP UL
P term M i ic 1 k = ------------------------------------------------------------low Service M i ( TP low Service M i is the nearest lower supported
UL
TP Service M i ic 1
throughput)
UL
UL
TP Service M i ic 1 = TP low Service M i

EndWhile
req
max
If P term M i ic 1 k P term M i , then Mi is not connected to cells of the sub-active set associated with ic 1 .
380

AT332_TRR_E0
Else the remaining terminal power is allocated to the second best carrier ( ic 2 ) and the traffic channel throughput
UL
TP Service M i ic 2 is calculated.
The same process is repeated for the other carriers in Mi s active set as long as the remaining terminal power is sufficient to
obtain the lowest bearer allowed.
If no sub-active set can be used, then Mi is rejected.
Endif
UL
Calculation of TP Service M i for each combination of carriers

n
UL
TP Service M i =
TP
UL
Service M i ic where n corresponds to the number of carriers in the combination.
ic = 1
UL
Selection of the configuration providing the highest throughput, Max TP Service M i .

UL
UL
UL
If Max TP Service M i TP high Service M i ( TP high Service M i is the nearest supported throughput higher than
the requested throughput)
Downgrading the traffic channel throughput
UL
UL
UL
While Max TP Service M i TP high Service M i and TP Service M i ic 153 6kbps

EndWhile
EndIf
Endfor
Uplink Interference Updates
Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones)
UL
EndFor
Control of Radio Resource Limits (Number of EVDO users, MAC Indices and Site Channel Elements)
Max
While n EVDO S j ic n EVDO S j ic

Rejection of the last admitted mobile
EndFor
Max
While N MacIndexes S j ic N MacIndexes S j ic

EndFor
For each site (Node B) Nl
Max
While N EVDO CE N I k N EVDO CE N I

EndFor
UL
UL
UL
For each cell (Sj,ic) with NR S j ic NR threshold S j ic + NRthreshold S j ic

UL
UL
UL
While NR S j ic NR threshold S j ic + NRthreshold S j ic and there is at least one mobile that can be downgraded
381

Downgrading the traffic channel throughput for all 1xEV-DO Rev. 0 mobiles for which the throughput transition flag is set to
"True".
UL
Endwhile
UL
UL
UL
For each cell (Sj,ic) with NR S j ic NR threshold S j ic NRthreshold S j ic

UL
UL
UL
While NR S j ic NR threshold S j ic NRthreshold S j ic and there is at least one mobile that can be upgraded
Upgrading the traffic channel throughput for all 1xEV-DO Rev. 0 mobiles for which the throughput transition flag is set to
"True". (only 1xEV-DO Rev. 0 mobiles which have not been downgraded can be upgraded. In addition, the upgraded
throughput cannot exceed the initial user throughput drawn by the Monte-Carlo algorithm. This means that only mobiles
downgraded during the uplink power control step can be upgraded).
UL
Endwhile
UL
UL
For each cell (Sj,ic) with X S j ic X max

Rejection of a mobile with the lowest service priority
EndFor
UL
UL
While at least one cell with X S j ic X max exists

Downlink Data Rate Control
For each mobile Mi connected to a cell (Sk,ic)
DL
Calculation of N tot ic b traffic

For each cell (Sj,ic) ( k j )
Determination of the number of mobiles connected to the cell (Sj,ic), N mobiles S j ic
If N mobiles S j ic = 0 then, P tx S j ic b traffic = G idle power P max S j ic
Else P tx S j ic b traffic = P max S j ic
EndFor
DL
N tot ic b traffic =
Ptot Sj ic btraffic + N0
DL
term
j j k
EndFor
Calculation of the maximum throughput supplied to Mi, TP max DL
For the Mis best server cell (Sk,ic) (in the active set or each sub-active set)
Calculation of pilot quality level at Mi
DL
E
P tot M i S k ic b pilot
----c- M i S k ic b pilot = -------------------------------------------------DL
Nt
N tot ic b pilot
If Mi is a 1xEV-DO Rev. 0 service user, determination of the peak throughput from the graph (Peak throughput=f(C/I)) specified
for the mobility type of Mi
E
TP max DL M i S k = f ----c- M i S k ic b pilot
Nt
If Mi is a 1xEV-DO Rev. A service user, selection of the downlink 1xEV-DO radio bearer ( Index DL Bearer ) for which
DL
E
E
----c- M i S k ic b pilot ----c- Index DL Bearer
Nt
min
Nt
382

AT332_TRR_E0
If Mi is a 1xEV-DO Rev. B service user, selection of the downlink 1xEV-DO radio bearer ( Index DL Bearer ) for which
DL
E
E
----c- M i S k ic b pilot ----c- Index DL Bearer
and the modulation is supported by Mis terminal.
Nt
min
Nt
DL
If Mi is a (1xEV-DO Rev. A - Guaranteed bit rate) service user and TP P R LC Index DL Bearer TPD min DL Service M i , Mi is
rejected.
DL
Determination of the peak throughput: TP max DL M i S k ic = -----------------------------------------------------------n TS
TP max DL M i = TP max DL M i S k ic
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, TP max DL M i = TPD min DL Service M i
For multi-carrier 1xEV-DO Rev. B service users, TP max DL M i =
DL
TP max DL max M i S k ic
S k ic
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, calculation of C DL Bearer
EndFor
DL
Calculation of the average cell throughput, TP av

G MU N mobiles S j ic
TP max DL M i S j ic
Mi NVBR m obiles Sj ic
C DL Bearer M k S j ic
----------------------------------------------------------------------------------------- 1
N VBR m obiles S j ic
M k N GBR m obiles Sj ic
DL
TP av S j ic =
TPD min DL M k
Mk NGBR m obiles Sj ic
-
+ -------------------------------------------------------------------------------------------C DL Bearer M k S j ic
N
ic
GBR m obiles j
Mk NGBR m obiles Sj ic
1 ER S ic
DRC j
N mobiles
1 TS
BCMCS S j ic TS EVDO CCH S j ic + TP BCMCS S j ic TS BCMCS S j ic
If N mobiles S j ic = 1 , then G MU = 1
Else if N mobiles S j ic 1 , G MU is determined from the graph (MUG table=f(nb users)) specified for (Sj,ic). If the transmitter
supports the multi-carrier EV-DO mode, G MU is determined from the graph (MUG table=f(nb users)) specified for Sj.
EndIf
EndFor
5.4.2.2.3
Convergence Criterion
The algorithm convergence is studied on uplink only. The uplink convergence criterion is evaluated at each iteration, and can
be written as follow:
UL
UL
UL
UL

UL = max int ma x ------------------------------------------------- 100 int ma x ----------------------------------------------------------- 100
UL
UL
Stations
Stations
ic
I ic
N
tot
user

1st case: Between two successive iterations, UL is lower ( ) than the threshold (defined when creating a simulation).
Example: Let us assume that the maximum number of iterations is 100, UL convergence threshold is set to 5. If UL 5
between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th iteration. Convergence has been achieved.
383

2nd case: After 30 iterations, UL is still higher than the threshold and from the 30th iteration, UL does not decrease during
the next 15 successive iterations.
Examples: Let us assume that the maximum number of iterations is 100, UL convergence threshold is set to 5.
1. After the 30th iteration, UL equals 100 and do not decrease during the next 15 successive iterations: Atoll stops the
algorithm at the 46th iteration. Convergence has not been achieved.
2. After the 30th iteration, UL equals 80, it starts decreasing slowly until the 40th iteration (without going under the
threshold) and then does not change during the next 15 successive iterations: Atoll stops the algorithm at the 56th iteration
without achieving convergence.
If UL is still strictly higher than the threshold, the simulation has not converged (specific divergence symbol).
If UL is lower than the threshold, the simulation has converged.
5.4.3 Appendices
5.4.3.1 Admission Control
During admission control, Atoll calculates the uplink load factor of a considered cell assuming the mobile concerned is
connected with it. Here, activity status assigned to users is not taken into account. So even if the mobile is not active on UL,
it can be rejected due to cell load saturation. To calculate the cell UL load factor, either Atoll takes into account the mobile
power determined during power control if mobile was connected in previous iteration, or it estimates a load rise due to the
mobile and adds it to the current load. The load rise ( X
X
UL
UL
1
= ------------------------------------W
1 + --------------------------UL
UL
Q req TP
In case of CDMA2000 1xRTT networks, we have:

UL
UL
UL
Q req = Q req FCH + Q req SCH and TP
UL
FCH
SCH
= TP P UL + TP P UL
5.4.3.2 Resources Management

5.4.3.2.1
Walsh Code Management

Walsh codes are managed in the downlink during the simulation in case of CDMA2000 1xRTT networks. Atoll performs Walsh
code allocation during the radio resource control step.
Walsh codes form a binary tree with codes of a longer length generated from codes of a shorter length. Length-k Walsh codes
are generated from length-k/2 Walsh codes. Therefore, if a channel needs 1 length-k/2 Walsh code, it is equivalent to using 2
length-k Walsh codes, or 4 length-2k Walsh codes and so on.
Figure 5.3: Walsh Code Tree Indices (Not Walsh Code Numbers)
128 128-bit-length Walsh codes per cell are available in CDMA2000 documents.
384

AT332_TRR_E0
During the resource control, Atoll determines the number of 128-bit-length Walsh codes that will be consumed by each cell.
Therefore, it allocates :
A code with the longest length (i.e. a 128 bit-length code) per common channel for each cell. The number of common
channels per cell corresponds to the value defined for the DL overhead resources for common channels per cell
parameter available in the site equipment properties.
Two 128 bit-length codes per cell-receiver link for FCH in RC1, RC2, RC3 or RC5 and only one for FCH in RC4.
The number of 128 bit-length codes to be allocated per cell-receiver link for SCH (in case SCH is supported by the user
Walsh codes
radio configuration), N 128 bits

Walsh codes
N 128 bits
, is determined as follows:
SCH
= TPF DL 2 for RC1, RC2, RC3 and RC5,
And
Walsh codes
N 128 bits
SCH
= TPF DL for RC4.
Where
SCH
TPF DL
is the SCH throughput factor.
The Walsh code allocation follows the Buddy algorithm, which guarantees that:
If a k-length Walsh code is used, all of its children with lengths 2k, 4k, , cannot be used as they are not orthogonal.
If a k-length Walsh code is used, all of its ancestors with lengths k/2, k/4, , cannot be used as they are not orthogonal.
5.4.3.2.2
The Walsh code allocation follows the mobile connection order (mobile order in the
Mobiles tab).
The Walsh code and channel element management is dealt with differently in case of
softer handoff. Atoll allocates Walsh codes for each transmitter-receiver link while
it assigns channel elements globally to a site.
Channel Element Management

Channel elements are controlled in the simulation.
CDMA2000 1xRTT networks
Atoll checks the availability of this resource on uplink and downlink.
On uplink, Atoll consumes N CE UL j channel elements for each cell j on a site NI. This figure includes:
Overhead
N CE UL
FCH
N CE UL
channel elements for control channels (Pilot channel),

SCH
1 + TPF UL per cell-receiver link, for TCH (TCH correspond to Traffic channels i.e. FCH and SCH).
Therefore, the number of channel elements required on uplink at the site level, N CE UL N I , is:
N CE UL N I =
NCE UL j
j NI
In the downlink, Atoll consumes N CE DL j channel elements for each cell j on a site NI. This figure includes:
Overhead
N CE DL
FCH
N CE DL
channel elements for control channels (Pilot channel, Synchronisation channel, Paging channel),
SCH
1 + TPF DL per cell-receiver link, for TCH (TCH correspond to Traffic channels i.e. FCH and SCH).
Therefore, the number of channel elements required on downlink at the site level, N CE DL N I , is:
N CE DL N I =
NCE DL j
j NI
In case of softer handover (the mobile has several links with co-site cells), Atoll
allocates channel elements for the best serving cell-mobile link only.
385

CDMA2000 1xEV-DO networks

In the uplink, Atoll consumes N CE UL j channel elements for each cell j on a site NI. This figure includes:
2 channel elements for control channels (Pilot channel, Data Rate Control channel, etc ). This value is fixed and hardcoded.
N CE UL per cell-receiver link, for (EV-DO - Variable bit rate) service users.
N CE UL C UL Bearer per cell-receiver link, for (EV-DO - Guaranteed bit rate) service users.
TCH
TCH
Therefore, the number of channel elements required on uplink at the site level, N CE UL N I , is:
N CE UL N I =
NCE UL j
j NI
In the downlink, only one user can be served by a cell at a time, so this resource is not limited.
5.4.3.3 Downlink Load Factor Calculation

Atoll calculates the downlink load factor for each cell (available in the Cells tab of any given simulation results) and each
connected mobile (available in the Mobiles tab of any given simulation results).
5.4.3.3.1
Downlink Load Factor per Cell

The downlink load factor is calculated for each CDMA2000 1xRTT cell.
Approach for downlink load factor evaluation is highly inspired by the downlink load factor defined in the book WCDMA for
UMTS by Harry Holma and Antti Toskala.
DL FCH
DL SCH
Q req
Q req
+ -------------------Let CI req = -------------------be the required quality.
DL FCH
DL SCH
Gp
Gp
FCH
SCH
So, we have CI req = CI req + CI req
In case of soft handoff, required quality is limited to the effective contribution of the transmitter.
DL
ortho
P tx ic = P pilot ic + P sync ic + P paging ic + P SCH ic + P FCH ic = P CCH ic +
Ptch ic
tch
where
ortho
P CCH ic = P pilot ic + P sync ic + P paging ic
Ptch ic
= P SCH ic + P FCH ic
tch
At mobile level, we have a required power, Ptch:

term
P tch ic = CI req I extra ic + I intra ic + I inter carrier ic + I inter techno log y ic + N 0
P tch ic = CI req
I extra ic + I inter carrier ic + I inter techno log y ic

DL
P tx ic P tch ic
- + N term
+ 1 F ortho BTS ----------------------------------------0
LT
LT
L
T
DL
term
I extra ic + I inter carrier ic + I inter techno log y ic L T + 1 F ortho BTS P tx ic + N 0 L T

P tch ic = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------1
----------- + 1 F ortho BTS
CIreq
where
DL
I intra ic is the total power received at receiver from the cell to which it is connected.
DL
I extra ic is the total power received at receiver from other cells.

I inter carrier ic is the inter-carrier interference received at receiver.
I inter techno log y ic is the inter-technology interference received at receiver.
386

AT332_TRR_E0
I
ic + I inter carrier ic + I inter techno log y ic L T
extra
DL
term
1
F
+
ic
+
N
ortho
BTS
tx
0
T
DL
ortho
P tx ic = P CCH ic +
--------------------------------------------------------------------------------------------------------------------------------------------------
1
----------- + 1 F ortho BTS

tch
CI req
We have:
I
extra ic + I inter carrier ic + I inter techno log y ic L T
--------------------------------------------------------------------------------------------------------------------------------------DL
P tx ic
DL
term
+ 1 F ortho BTS P tx ic + N 0 L T
DL
ortho
P tx ic = P CCH ic +
--------------------------------------------------------------------------------------------------------------------------------------------------
1 - + 1 F
---------ortho BTS
tch
CI req
I extra ic + I inter carrier ic + I inter techno log y ic L T

--------------------------------------------------------------------------------------------------------------------------------------+ 1 F ortho BTS

DL
P tx ic
tch
DL
DL
P tx ic --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- P tx ic
1
----------- + 1 F ortho BTS
CI req
ortho
= P CCH ic +
term
N0 LT
--------------------------------------------------------------
1
tch ----------- + 1 F ortho BTS

CI req
term
ortho
N0 LT
---------------------------------------------------------------
P CCH ic +
1
tch ----------- + 1 F ortho BTS
CI req
DL
P tx ic = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ I extra ic + I inter carrier ic + I inter techno log y ic L T
--------------------------------------------------------------------------------------------------------------------------------------- + 1 F ortho BTS

DL
P tx ic
1
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 - + 1 F
---------
tch
ortho BTS
CI req
Therefore, the downlink load factor can be expressed as:
DL
I extra ic + I inter carrier ic + I inter techno log y ic L T

--------------------------------------------------------------------------------------------------------------------------------------+ 1 F ortho BTS
DL
ic
tx
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
=
1 - + 1 F
---------
ortho BTS
tch
CI
req
The downlink load factor represents the signal degradation in relative to the reference interference (thermal noise).
5.4.3.3.2
Downlink Load Factor per Mobile

Atoll evaluates the downlink load factor for any connected mobile (CDMA2000 1xRTT 1xEV-DO user) as follows,
X
DL
DL
I tot ic
= -----------------DL
N tot ic
5.4.3.4 Best Server Determination in Monte Carlo Simulations - Old Method

Before Atoll 2.8.0, best server determination used to be performed by selecting the best carrier within transmitters according
to the selected method (site equipment) and then the best transmitter using the best carrier. To switch back to this method,
add the following lines in the Atoll.ini file:
[CDMA]
MultiBandSimu = 0
387

The method is described below:

For each station Sj containing Mi in its calculation area and using the main frequency band supported by the Mis terminal (i.e.
either f1 for a single frequency band network, or f1, f2 or f3 for a multi-band terminal without any priority on frequency bands,
or f1 for a multi-band terminal with f1 as main frequency band).
Determination of BestCarrier k S j M i .
If a given carrier is specified for the service requested by Mi and if it is used by Sj
BestCarrier k S j M i is the carrier specified for the service.
Else the carrier selection mode defined for Sj is considered.
For each carrier ic used by Sj, we calculate current loading factor:
UL
I tot S j ic
UL
UL
- + X
X k S j ic = ------------------------UL
N tot S j ic
EndFor
UL
BestCarrier k S j M i is the carrier with the lowest X k S j ic

BestCarrier k S j M i is the carrier with the lowest P tx S j ic k
BestCarrier k S j M i is randomly selected
UL
UL
BestCarrier k S j M i is the first carrier so that X k S j ic X max

BTS P c M i S j BestCarrier
Calculation of Q pilot M i S j BestCarrier = ----------------------------------------------------------------------------------DL
k
I 0 BestCarrier k S j M i
pilot
If Q pilot M i S j BestCarrier Q req then Sj is rejected by Mi

k
max
If Q pilot M i S j BestCarrier Q pilot M i

k
Admission control (If simulation respects a load factor constraint and Mi was not connected in previous iteration).
UL
UL
If X k S j BestCarrier k S j M i X max , then Sj is rejected by Mi

Else
max
Q pilot M i = Q pilot M i S j BestCarrier

k
S BS M i = S j
Endif
EndFor
If no SBS has been selected and Mis terminal can work on one frequency band only, Mi has failed to be connected to the
network and is rejected.
If no SBS has been selected and Mis terminal can work on another frequency band.
Determination of BestCarrier k Sj M i for each station txj containing Mi in its calculation area and using another frequency
band supported by the Mis terminal (i.e. f1, f2 or f3 for a multi-band terminal without any priority on frequency bands, or f2
for a multi-band terminal with f2 as secondary frequency band)
388

AT332_TRR_E0

UL
I tot S j ic
UL
UL
X k S j ic = ------------------------- + X
UL
N tot S j ic
EndFor
UL

UL
UL

k
pilot

k
max

k
UL
UL

Else
max

k
S BS M i = S j
Endif
EndFor
If no SBS has been selected and Mis terminal can work on two frequency bands only, Mi has failed to be connected to the
network and is rejected.
If no SBS has been selected and Mis terminal can work on another frequency band.
Determination of BestCarrier k Sj M i for each station txj containing Mi in its calculation area and using another frequency
band supported by the Mis terminal (i.e. f1, f2 or f3 for a multi-band terminal without any priority on frequency bands, or f3
for a multi-band terminal with f3 as third frequency band)
UL
I tot S j ic
UL
UL
X k S j ic = ------------------------- + X
UL
N tot S j ic
EndFor
389


UL

UL
UL

k
pilot

k
max

k
UL
UL

Else
max

k
S BS M i = S j
Endif
EndFor
If no SBS has been selected, Mi has failed to be connected to the network and is rejected.
5.4.3.5 Radio Bearer Allocation Algorithm for Multi-carrier EVDO Rev.B - Old
Method
Before Atoll 3.2.1, radio bearer allocation for multi-carrier EVDO Rev.B used to be performed by equally sharing the available
terminal power between the carriers.
To switch back to this method, add the following lines in the Atoll.ini file:
[CDMA]
SharingEquallyPower = 1
UsingPreviousIterationPowerWeight = 1
5.5 CDMA2000 Prediction Studies

5.5.1 Point Analysis: The AS Analysis Tab
Let us assume a receiver with a terminal, a mobility type and a service with certain UL and DL throughputs. This receiver does
not create any interference. You can make the prediction for a specific carrier or for the best 1xRTT or 1xEV-DO carrier. The
type of carrier and the carriers you can select depend on the service and on the frequency band(s) supported by the terminal.
The analysis is based on the uplink load percentage and the downlink total power of cells. These parameters can be either
outputs of a given simulation, average values calculated from a group of simulations, or user-defined cell inputs.
Results are displayed for any point of the map where the pilot signal level exceeds the defined minimum RSCP.
5.5.1.1 Bar Graph and Pilot Sub-Menu

We can consider the following cases:
390

AT332_TRR_E0

The carrier that can be used by transmitters is fixed. In this case, for each transmitter i containing the receiver in its calculation
area and using the selected carrier, Atoll calculates the pilot quality at the receiver on this carrier. Then, it determines the best
serving transmitter using the selected carrier ic.
2nd case: Analysis based on the best carrier of all frequency bands
Atoll determines the best carrier for each transmitter i which contains the receiver in its calculation area and uses a frequency
band supported by the receivers terminal. The best carrier selection depends on the option selected for the site equipment
(UL minimum noise, DL minimum power, random, sequential). Then, Atoll calculates the pilot quality at the receiver from
these transmitters on their best carriers (ic) and defines the best server (on its best carrier).
3rd case: Analysis based on the best carrier of any frequency band (for multi-band terminals with priority defined on frequency
bands only)
The frequency band that can be used is fixed. Atoll determines the best carrier for each transmitter i containing the receiver
in its calculation area and using the selected frequency band. The best carrier selection depends on the option selected for
the site equipment (UL minimum noise, DL minimum power, random, sequential). Then, Atoll calculates the pilot quality at
the receiver from these transmitters on their best carriers (ic) and defines the best server (on its best carrier).
Atoll provides the same outputs in the bar graph and pilot sub-menu whichever the studied network, CDMA2000 1xRTT or
1xEV-DO.
Ec/I0 (or Q pilot ic ) evaluation
We assume that ic is the best carrier of a transmitter i containing the receiver in its calculation radius.
For CDMA2000 1xRTT users we have,
BTS P c i ic
Q pilot i ic = --------------------------------------------DL
I 0 ic
DL
DL
DL
DL
term
DL
with I 0 ic = P tot i ic + I extra ic + I inter carrier ic + I inter techno log y ic + N 0

For CDMA2000 1xEV-DO users, we have,
DL
BTS P tot i ic b pilot

Q pilot i ic = ---------------------------------------------------------------DL
I 0 ic b pilot
DL
DL
DL
DL
DL
term
With I 0 ic b pilot = P tot i ic b pilot + I extra ic b pilot + I inter carrier ic b pilot + I inter techno log y ic + N 0
The calculation of Q pilot i ic can be divided into 6 steps explained in the table below.
CDMA2000 1xRTT users
CDMA2000 1xEV-DO users

DL
P tot i ic b pilot calculation for each cell (i,ic)

P c i ic calculation for each cell (i,ic)
P c i ic is the pilot power from a transmitter i on the carrier ic at
DL
P tot i ic b pilot is the pilot burst from the transmitter i on the

carrier ic at the receiver.
1st step
the receiver.
P pilot i ic
P c i ic = ------------------------LT
I
P tx i ic b pilot
DL
P tot i ic b pilot = ----------------------------------LT
I
and
P tx i ic b pilot = P max i ic

L T is the total loss between the transmitter i and the receiver: -----------------------------------------------------------------------------------------------------------------------------------I
G Tx G term
391


DL
DL
DL
I extra ic , I inter carrier ic and I inter techno log y ic calculation

We have,
DL
I extra ic
Ptot j ic
DL
DL
2nd step
For each transmitter of the network,
DL
P tot j
DL
I extra ic b pilot and I inter carrier ic b pilot calculation
j j i
ic is the total power

We have,
received at the receiver from the transmitter j on the best carrier ic

of the transmitter i.
P tx j ic
DL
P tot j ic = -------------------LT
DL
j j i
P tx j ic is the total power transmitted by the transmitter j on the
Ptot j icadj bpilot
DL
and
Finally, we have,
DL
DL
j j
I inter carrier ic b pilot = ---------------------------------------------------RF ic ic adj
best carrier of the transmitter i.
DL
P tot j
Ptot j ic bpilot
DL
I extra ic b pilot =
ic adj
DL
j
I inter carrier ic = j-----------------------------------RF ic ic adj
ni
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
and
DL
ni
Tx
P Transmitted ic i
-------------------------------------Tx
Tx m
L total ICP ic ic
i
3rd step
term
N0
calculation
Tx DL
4th step

DL
I 0 ic and Q pilot i ic evaluation based on formulas defined above

DL
G macro diversity calculation

DL
The macro-diversity gain, G macro diversity , models the decrease in shadowing margin due to the fact there are several pilot signals at the
5th step
mobile.
DL
G macro diversity
npaths
M Shadowing Ec Io
M Shadowing Ec Io
npaths
M Shadowing Ec Io is the shadowing margin for the mobile receiving n pilot signals (not necessarily from transmitters belonging to the
mobile active set).
Note: This parameter is determined from the fixed cell edge coverage probability and the model standard deviation. When the model
standard deviation is set to 0, the macro-diversity gain equals 0.
392

AT332_TRR_E0
Determination of active set

Atoll takes the transmitter i with the highest Q pilot i ic and calculates the best pilot quality received with a fixed cell edge coverage
Resulting
probability Q pilot
Resulting
Q pilot
ic
Resulting
Q pilot
DL
G macro diversity
req
ic .
max Q pilot i ic
Resulting
Q pilot means that the pilot quality at the receiver exceeds Q pilot
ic x% of times (x is the fixed cell edge coverage
probability). The cell with the highest Q pilot i ic enters the active set as best server ( Q pilot BS ic ) and the best carrier (icBS) of the
6th step
best server
BS will be the one used by other transmitters of active set (when active set size is greater than 1). Pilot is available.
Resulting
If Q pilot
req
ic Q pilot , no cell (i,ic) can enter the active set. Pilot is unavailable.
Then, pilot qualities at the receiver from transmitters i (other than the best server) on the best carrier of the best server, icBS, are
recalculated to determine the entire receiver active set (when active set is greater than 1). Same formulas and calculation method are
used to update
DL
I 0 ic BS and determine Q pilot i ic BS .

Other cells (i,icBS) in active set must fulfill the following criteria:
pilot
Q pilot i ic BS Q min
i ic BS neighbour list BS ic BS (optional)

For multi-carrier 1xEV-DO Rev.B service users, these results are detailed for each sub-active set. For each carrier, Atoll displays
the thermal noise, I0 (Best server), the pilot quality from the best server and from the other servers of the sub-active set, and
the downlink macro-diversity gain. They are calculated as described above.
Number of cells in active set
This is a user-defined input in the terminal properties. It corresponds to the active set size.
Number of fingers
The number of fingers, f, of the rake receiver. This parameter is defined in the terminal properties. It is relevant in CDMA2000
1xRTT only11. This is the maximum number of active set links that the terminal (rake) can combine.
Thermal noise
This parameter is calculated as described above (3rd step).
I0 (Best server)
I0 (Best server) is the total noise received at the receiver on icBS.
Downlink macro-diversity gain
This parameter is calculated as described above (5th step).
5.5.1.2 Downlink Sub-Menu

Outputs calculated by Atoll depend on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO).
5.5.1.2.1
CDMA2000 1xRTT
Let mFCH and mSCH respectively denote the number of cells in the receiver active set for the fundamental channel (FCH) and
the supplemental channel (SCH) and f be the number of rake fingers defined for the terminal. We assume that f is less than
or equal to mFCH and mSCH.
Among the mFCH cells of the receiver active set, only the first f cells will be considered in order to determine the FCH
availability on downlink. In the same way, only the first f cells among the mSCH cells of the receiver active set will be considered
in order to determine the SCH availability on downlink. Each of these cells is noted (k,icBS).
Atoll calculates the traffic channel quality on FCH from each cell (k,icBS). No power control is performed as in simulations.
Here, Atoll determines the downlink traffic channel quality on FCH at the receiver for the maximum traffic channel power per
transmitter allowed on FCH. Then, after combination, the total downlink traffic channel quality on FCH is evaluated and
compared with the specified target quality.
11.
CDMA2000 1xEV-DO systems do not support soft handover on downlink.
393

Atoll calculates the traffic channel quality on SCH from each cell (k,icBS). No power control is performed as in simulations.
Here, Atoll determines the downlink traffic channel quality on SCH at the receiver for the maximum traffic channel power per
transmitter allowed on SCH. This value depends on the downlink throughput specified in the analysis. Then, after
combination, the total downlink traffic channel quality on SCH is evaluated and compared with the specified target quality.
Eb/Nt target on FCH and Eb/Nt target on SCH

DL
Eb/Nt target on FCH ( Q req FCH ) is the downlink traffic data quality target on the fundamental channel (FCH). This value is
user-defined for a given service and terminal.
DL
Eb/Nt target on SCH ( Q req SCH ) is the downlink traffic data quality target on the supplemental channel (SCH). This value is
specified for a given service, terminal and SCH throughput.
Required transmitter powers on FCH and SCH

req
req
The calculation of the required transmitter powers on FCH and SCH ( P FCH and P SCH ) may be divided into three steps.
1st step: Eb/Nt max for the first f (number of fingers) cells of active set
DL
DL
Let us assume the following notations: Eb/Nt max on FCH and SCH respectively correspond to Q max FCH and Q max SCH .
DL FCH
BTS P b max k ic BS
DL
FCH
- G DL
Q max k ic BS FCH = -------------------------------------------------------p
DL
N tot ic BS
And
DL SCH
DL
Q max k
BTS P b max k ic BS
SCH
- G DL
= -------------------------------------------------------p
DL
N tot ic BS
ic BS SCH
DL FCH
With P b
DL
max
max
P FCH DL SCH
P SCH
k ic BS = ---------- , P b max k ic BS = ---------LT
LT
DL
DL
DL
DL
term
And N tot ic BS = I intra ic BS + I extra ic BS + I inter carrier ic BS + I inter techno log y ic BS + N 0

Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table for a certain terminal.
max
P SCH is the maximum power allowed on SCH for the specified downlink throughput. This parameter is user-defined in the
Services table for a certain terminal and SCH throughput.
L T is the total loss between the transmitter i and the receiver.
k

DL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
DL
N tot ic BS is the total noise at the receiver on the best carrier of the best server.
With
DL
I intra ic BS = 1 BTS F ortho P DL k ic

tot
BS
And
DL
I extra ic BS =
Ptot j icBS
DL
j j k
DL
For each transmitter in the network, P tot ic BS is the total power received at the receiver from this transmitter on icBS.
DL
I inter carrier ic BS is the inter-carrier interference at the receiver on the best carrier of the best server.
394

AT332_TRR_E0
Ptot j icadj
DL
DL
j
I inter carrier ic BS = txj
---------------------------------------RF ic BS ic adj

DL
I inter techno log y ic BS is the inter-technology interference at the receiver on the best carrier of the best server.
DL
ni
ic i is the i
Tx m
ICPic ic
i
BS
th
Tx
P Transmitted ic i
----------------------------------------Tx
Tx m
L total ICP ic ic
i
BS

2nd step: Calculation of the total traffic channel quality on FCH and SCH
DL
Q MAX FCH is the traffic channel quality on FCH at the receiver on icBS after combining the signal from each cell (k,icBS).
DL
DL
Q MAX ic BS FCH = Q max k ic BS FCH

DL
DL
Q MAX ic BS FCH = f rake efficiency
Qmax k icBS FCH

DL
Where
DL
DL
Q MAX SCH is the traffic channel quality on SCH at the receiver on icBS after combining the signal from each cell (k,icBS).
DL
DL
Q MAX ic BS SCH = Q max k ic BS SCH

DL
DL
Q MAX ic BS SCH = f rake efficiency
Qmax k icBS SCH

DL
k
req
req
3rd step: P FCH and P SCH calculation

DL
Q req FCH
req
- P max
P FCH = --------------------------------------FCH
DL
Q MAX ic BS FCH
DL
Q req SCH
req
- P max
P SCH = --------------------------------------SCH
DL
Q MAX ic BS SCH
Eb/Nt max on FCH for the first f (number of fingers) cells of active set
DL
Let us assume the following notation: Eb/Nt max on FCH corresponds to Q max FCH .
DL FCH
BTS P b max k ic BS
DL
FCH
- G DL
Q max k ic BS FCH = -------------------------------------------------------p
DL
N tot ic BS
max
P FCH
DL FCH
DL
DL
DL
DL
term
- and N tot ic BS = I intra ic BS + I extra ic BS + I inter carrier ic BS + N 0
k
395

Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table for a certain terminal.
k

DL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
DL
With
max
req
DL
P FCH P FCH
I intra ic BS = 1 BTS F ortho P DL
tot k ic BS 1 BTS max (----------------------------,0)
LT
k
And
DL
I extra ic BS =
Ptot j icBS
DL
j j k
DL
For each transmitter in the network, P tot ic BS is the total power received at the receiver from the transmitter on icBS.
DL
Ptot j icadj
DL
DL
txj j
I inter carrier ic BS = ---------------------------------------RF ic BS ic adj

DL
Tx
DL
P Transmitted ic i
L----------------------------------------Tx
Tx m
ICP
n
ic i is the i
Tx m
ICP ic ic
i
BS
th
total
ic ic
i BS
Eb/Nt max on SCH for the first f (number of fingers) cells of active set
DL
Let us assume the following notation: Eb/Nt max on SCH corresponds to Q max SCH .
DL SCH
DL
Q max k
ic BS SCH
BTS P b max k ic BS
SCH
- G DL
= -------------------------------------------------------p
DL
N tot ic BS
max
P SCH
DL SCH
k
DL
DL
DL
DL
DL
term
and N tot ic BS = I intra ic BS + I extra ic BS + I inter carrier ic BS + I inter techno log y ic BS + N 0

Where
max
P SCH is the maximum power allowed on SCH for the specified downlink throughput. This parameter is user-defined in the
Services table for a certain terminal and SCH throughput.
396

AT332_TRR_E0

k

DL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
DL
With
max
req
DL
P SCH P SCH
I intra ic BS = 1 BTS F ortho P DL
tot k ic BS 1 BTS max (----------------------------,0)
LT
k
And
DL
I extra ic BS =
Ptot j icBS
DL
j j k
DL
For each transmitter in the network, P tot ic BS is the total power received at the receiver from the transmitter on icBS.
DL
Ptot j icadj
DL
DL
j
---------------------------------------I inter carrier ic BS = txj
RF ic BS ic adj

DL
Tx
DL
P Transmitted ic i
L----------------------------------------Tx
Tx m
ICP
ni
ic i is the i
Tx m
ICPic ic
i
BS
th
ic i ic BS
total
Eb/Nt max on FCH and Eb/Nt max on SCH
DL
Q MAX FCH is the traffic channel quality on FCH at the receiver on icBS after combining the signal from each cell (k,icBS).
DL
DL
Q MAX ic BS FCH = Q max k ic BS FCH

DL
DL
Qmax k icBS FCH

DL
Where
DL
DL
Q MAX SCH is the traffic channel quality on SCH at the receiver on icBS after combining the signal from each cell (k,icBS).
DL
DL
Q MAX ic BS SCH = Q max k ic BS SCH

397

DL
DL
Q MAX ic BS SCH = f rake efficiency
Qmax k icBS SCH

DL
Therefore, the service on the downlink traffic channel is available if

DL
DL
DL
Q MAX ic BS FCH Q req FCH
and
DL
Q MAX ic BS SCH Q req SCH .
Effective Eb/Nt on FCH and Eb/Nt on SCH
DL
DL
Q eff FCH and Q eff SCH are respectively effective traffic channel qualities at the receiver on icBS supplied on FCH and SCH.
DL
DL
DL
DL
DL
Q eff FCH = min Q MAX FCH Q req FCH

And
DL
Q eff SCH = min Q MAX SCH Q req SCH
Downlink soft handover gain on FCH and downlink soft handover gain on SCH
DL
DL
G SHO FCH and G SHO SCH respectively correspond to DL soft handover gains on FCH and SCH.
DL
Q MAX ic BS FCH
DL
G SHO FCH = -------------------------------------------------------------DL
max Q max k ic BS FCH
k
And
DL
Q MAX ic BS SCH
DL
G SHO SCH = -------------------------------------------------------------DL
max Q max k ic BS SCH
k
DL
DL

k
5.5.1.2.2
CDMA2000 1xEV-DO
Atoll calculates the effective pilot quality level at the receiver and compares this value with the required quality level.
1xEV-DO Rev.0 and 1xEV-DO Rev. A Service Users
For 1xEV-DO Rev.0 and 1xEV-DO Rev. A users, Atoll displays the following results:
Required throughput
DL
The required throughput, TP req , is the downlink throughput selected for the analysis.
Required C/I
C
For 1xEV-DO Rev. 0 users, the required C/I ( --- ) is determined from the graph Peak throughput=f(C/I) defined for the
I req
mobility type selected in the analysis. It corresponds to the value read in the graph Peak throughput=f(C/I) (Rev0) for the
DL
specified required throughput, TP req .

DL
For 1xEV-DO Rev. A users, the required throughput ( TP req ) is obtained by using a certain downlink transmission format (i.e.
a 1xEV-DO radio bearer ( Index DL Bearer ) with a certain number of timeslots ( n TS )). It is calculated as follows:
DL
DL
TP req = -----------------------------------------------------------n TS
C
---
I req is the value defined in the 1xEV-DO Radio Bearer Selection (Downlink) table for this downlink transmission format
(radio bearer Index, mobility and number of timeslots). It corresponds to the C/I required to obtain the defined required
DL
throughput, TP req .
Effective C/I
Ec
Let ----- ic BS b pilot be the effective C/I at the receiver on icBS.
Nt
398

AT332_TRR_E0
For the best cell (BS,icBS) of the receiver active set, we have:
Ec
1
----- ic BS b pilo t = -------------------------------
1
Nt
--------------------- 1
Q pilot
resulting
Where
pilot
DL
Q resulting ic BS = G macro diversity Q pilot ic BS

BS
Obtained throughput
For 1xEV-DO Rev. 0 users, the obtained throughput, TP

defined for the mobility type selected in the analysis. TP
Ec
the calculated effective C/I, ----- ic BS b pilot .
Nt
For 1xEV-DO Rev. A users, the obtained throughput ( TP
DL
DL
DL
, is determined from the graph Peak throughput=f(C/I) (Rev0)

is the value read in the graph Peak throughput=f(C/I) (Rev0) for
) on downlink depends on the downlink transmission format, i.e the
radio bearer index ( Index DL Bearer ) with the number of timeslots ( n TS ). For the defined mobility type, Atoll selects the
Ec
C
downlink transmission format where ----- ic BS b pilot --- . Then, it determines the downlink obtained throughput as
Nt
I req
follows:
TP
DL
DL
= -----------------------------------------------------------n TS
The traffic data channel in downlink is available if TP
DL
DL
TP req .
Bearer Consumption
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, Atoll calculates the 1xEV-DO bearer consumption.
TPD min DL
C DL Bearer = -----------------------------------------------------------DL
TPP R LC Index DL Bearer
Where TPD min DL corresponds to the minimum bit rate required by the service in the downlink.
1xEV-DO Rev. B Service Users
For single-carrier and multi-carrier 1xEV-DO Rev. B users, Atoll displays the following results:
Required throughput
DL
The required throughput, TP req , is the downlink throughput selected for the analysis.
Obtained throughput
The obtained throughput corresponds to the sum of the obtained throughputs on each carrier.
TP
DL
TP
DL
ic
ic
The traffic data channel on downlink is available if TP
DL
DL
TP req .
For each sub-active set, Atoll indicates the effective C/I and the obtained throughput:
Ec
Let ----- ic b pilot be the effective C/I at the receiver on ic, the carrier associated with the sub-active set.
Nt
For the best cell (BS,ic) of the receiver sub-active set, we have:
pilot
E
Q resulting ic
----c- ic b pilot = ----------------------------------------pilot
Nt
Q resulting ic
Where
399

pilot
DL
Q resulting ic = G macro diversity Q pilot ic

BS
DL
The obtained throughput ( TP ic ) on downlink depends on the downlink transmission format, i.e the radio bearer index
( Index DL Bearer ) with the number of timeslots ( n TS ). For the defined mobility type, Atoll selects the downlink transmission
E
C
format where ----c- ic b pilot ---
and whose modulation scheme is supported by the terminal.
I req
Nt
C
---
I req is the value defined in the 1xEV-DO Radio Bearer Selection (Downlink) table for this downlink transmission format
(radio bearer Index, mobility and number of timeslots). It corresponds to the C/I required to obtain the defined required
DL
throughput, R req .
The downlink obtained throughput is determined as follows:
DL
TP P RLC Index DL Bearer

DL
TP ic = -------------------------------------------------------------n TS
5.5.1.3 Uplink Sub-Menu

Outputs calculated by Atoll depend on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO).
5.5.1.3.1
CDMA2000 1xRTT
For each cell (i,icBS) in the receiver active set, Atoll calculates the uplink traffic channel quality on FCH and SCH from the
receiver. No power control is performed as in simulations. Here, Atoll determines the uplink traffic channel quality on FCH at
the cell for the maximum terminal power allowed on FCH. In the same way, it evaluates the uplink traffic channel quality on
SCH at the cell for the maximum terminal power allowed on SCH. Then, total uplink traffic channel qualities on FCH and SCH
are evaluated with respect to the receiver handover status. From these values, Atoll deduces required terminal powers on
FCH and SCH, calculates the total terminal power required and compares this value with the maximum terminal power
allowed.
Max terminal power on FCH and SCH

max
The Max terminal power parameter ( P term ) is user-defined for each terminal. It corresponds to the maximum terminal power
allowed. On uplink, the terminal power is shared between pilot, FCH and SCH channels. So, we may write:
max
max
max
max
P term = P term pilot + P term FCH + P term SCH

We have:
max
max
P term pilot = p P term

Where p is the percentage of the terminal power dedicated to pilot. This parameter is user-defined in the terminal properties.
And
UL
FCH
UL
max
Q req FCH TP P UL AF FCH
P term FCH
- -------------------------------------------------------------- = ---------------------UL
SCH
max
Q req SCH
TP P UL
P term SCH
Therefore,
max
1 p P term
max
P term FCH = ----------------------------------------------------------------------------UL
SCH
Q req SCH TP P UL
1 + -------------------------------------------------------------------UL
FCH
UL
And
max
1 p P term
max
P term SCH = ----------------------------------------------------------------------------UL
FCH
UL
1 + -------------------------------------------------------------------UL
SCH
Q req SCH TP P UL
Required terminal power on FCH and SCH

req
req
The required terminal powers on FCH and SCH, respectively P term FCH and P term SCH , are calculated as follows:
400

AT332_TRR_E0
UL
1st step: Evaluation of uplink traffic channel qualities on FCH and SCH, Q max ic BS
i
UL
FCH
and Q max ic BS
i
SCH
, for each cell
of active set.
For each cell (i,icBS), we have:
UL FCH
term P b max i ic BS
UL
FCH
- G UL
Q max i ic BS FCH = --------------------------------------------------------p
UL
N tot i ic BS
And
UL SCH
UL
SCH
- G UL
Q max i ic BS SCH = --------------------------------------------------------p
UL
N tot i ic BS
max
max
P term FCH
P term SCH
UL FCH
UL SCH
With P b max i ic BS = ------------------------ and P b max i ic BS = -----------------------LT
LT
i

i

UL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
UL
N tot i ic BS is the total noise at the transmitter on the best carrier of the best server. This value is deduced from the cell
UL
uplink load factor X i ic BS .

tx
N0
UL
N tot i ic BS = ----------------------------------UL
1 X i ic BS
tx

UL
UL
2nd step: Calculation of FCH and SCH total traffic channel qualities at the transmitter on icBS, Q MAX FCH and Q max SCH ,
based on the receiver handover status.
If there is no handoff, we have:
UL
UL
UL
UL
Q MAX ic BS FCH = Q max i ic BS FCH and Q MAX ic BS SCH = Q max i ic BS SCH

For soft handover, we have:
UL
UL
UL
Q MAX ic BS FCH = G macro diversity 2 links max Q max i ic BS FCH

i
And
UL
UL
UL
Q MAX ic BS SCH = G macro diversity 2 links max Q max i ic BS SCH

i
UL
UL
UL
max Q max i ic BS corresponds to the highest Q max i ic BS value.

i
For soft-soft handover, we have:

UL
UL
UL

i
And
UL
UL
UL

i
UL
G macro diversity 3 links is the uplink macro-diversity gain.This parameter is determined from the fixed cell edge coverage
For softer and softer-softer handovers, we have:
401

UL
UL
Qmax i icBS FCH

UL
And
UL
Q MAX ic BS SCH
UL
f rake efficiency
Qmax i icBS SCH

UL
For softer-soft handover, there are two possibilities. If the MRC option is selected (option available in Global parameters), we
have:
UL
UL
UL
Q MAX ic BS FCH = G macro diversity 2 links max f rake efficiency
UL
UL
Q max i ic BS FCH Q max
i ic BS
i on the other site
i on the same site
FCH
And
UL
UL
UL
Q MAX ic BS SCH = G macro diversity 2 links max f rake efficiency
UL
UL
Q max i ic BS SCH Q max
i ic BS
i on the other site
i on the same site
SCH
otherwise,
UL
UL
UL

i
And
UL
UL
UL

i
req
req
3rd step: Calculation of P term FCH and P term SCH

UL
UL
Q req FCH
Q req SCH
req
req
- P max
- P max
P term FCH = --------------------------------------term FCH and P term SCH = --------------------------------------term SCH
UL
UL
Q MAX ic BS FCH
Q MAX ic BS SCH
Where
UL
Q req FCH is the user-defined uplink data traffic quality target on FCH for a given service and a terminal. This parameter is
available in the Services table.
UL
Q req SCH is the user-defined uplink data traffic quality target on SCH for a given service, terminal and SCH throughput. This
parameter is available in the Services table.
req
Then, from the required terminal power on FCH and SCH, Atoll determines the total terminal power required ( P term ).
req
req
req
req
P term = P term FCH + P term SCH + P term pilot

req
req
As P term pilot = p P term , we have:

req
req
P term FCH + P term SCH

req
P term = -------------------------------------------------------1p
req
max
Therefore, the service on the uplink data traffic channel is available if P term P term .
Eb/Nt max on FCH for each cell in active set

UL FCH
UL
FCH
- G UL
Q max i ic BS FCH = --------------------------------------------------------p
UL
N tot i ic BS
max
P term FCH
UL FCH
With P b max i ic BS = -----------------------LT
i

i

UL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
402

AT332_TRR_E0
UL
UL

tx
max
req
N0
P FCH P FCH
UL
N tot i ic BS = -----------------------------------,0)
- + 1 term max (--------------------------UL
LT
1 X i ic BS
i
tx
Eb/Nt max on SCH for each cell in active set

UL SCH
UL
SCH
- G UL
Q max i ic BS SCH = --------------------------------------------------------p
UL
N tot i ic BS
max
P term SCH
UL SCH
With P b max i ic BS = -----------------------LT
i

i

UL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
UL
UL

tx
max
req
N0
P SCH P SCH
UL
- + 1 term max (--------------------------N tot i ic BS = -----------------------------------,0)
UL
LT
1 X i ic BS
i
tx
Eb/Nt max on FCH and SCH
UL
UL
Q MAX ic BS FCH and Q MAX ic BS SCH are respectively the traffic channel qualities on FCH and SCH at the transmitter on icBS
after signal combination of all the transmitters of the active set.
UL
UL
UL
UL
Q MAX ic BS FCH = Q max i ic BS FCH and Q MAX ic BS SCH = Q max i ic BS SCH

UL
UL
UL

i
And
UL
UL
UL

i
UL
UL
UL

i

UL
UL
UL

i
And
UL
UL
UL

i
403

UL
UL
UL
Qmax i icBS FCH

UL
And
UL
Q MAX ic BS SCH
UL
f rake efficiency
Qmax i icBS SCH

UL
have:
UL
UL
UL
Q MAX ic BS FCH = G macro diversity 2 links max f rake efficiency
UL
UL
Q max i ic BS FCH Q max
i ic BS
i on the other site
i on the same site
FCH
And
UL
UL
UL
Q MAX ic BS SCH = G macro diversity 2 links max f rake efficiency
UL
UL
Q max i ic BS SCH Q max
i ic BS
i on the other site
i on the same site
SCH
otherwise,
UL
UL
UL

i
And
UL
UL
UL

i
Effective Eb/Nt on FCH and SCH
UL
Q eff FCH is the uplink effective traffic channel quality on FCH at the receiver on icBS.
UL
Q eff SCH is the uplink effective traffic channel quality on SCH at the receiver on icBS.
UL
UL
UL
UL
UL
UL
Q eff FCH = min Q MAX FCH Q req FCH and Q eff SCH = min Q MAX SCH Q req SCH
Uplink soft handover gain FCH and SCH
UL
G SHO FCH corresponds to the UL soft handover gain on FCH.

UL
G SHO SCH corresponds to the UL soft handover gain on SCH.

UL
UL
Q MAX ic BS FCH
Q MAX ic BS SCH
UL
UL
G SHO FCH = ------------------------------------------------------------ and G SHO SCH = -----------------------------------------------------------UL
UL
max Q max i ic BS FCH
max Q max i ic BS SCH
I
UL
UL

I
5.5.1.3.2
CDMA2000 1xEV-DO
1xEV-DO Rev.0 and 1xEV-DO Rev. A Service Users
For each cell (l,icBS) in the receiver active set, Atoll calculates the uplink quality level from the receiver. No power control is
performed as in simulations. Here, Atoll determines the uplink quality level at the cell for the maximum terminal power
allowed. Then, the total uplink quality level is evaluated with respect to the receiver handover status. From this value, Atoll
calculates the required terminal power and compares it with the maximum terminal power allowed.
Max terminal power

max
allowed.
Required terminal power with ACK

req
The required terminal power ( P term ) calculation may be divided into four steps:
404

AT332_TRR_E0
UL
1st step: Evaluation of the uplink quality, Q max i ic BS , for each cell of active set
UL
UL
Q max i
- G UL
ic BS = ----------------------------------------------------p
UL
N tot i ic BS
max
P term
UL
With P b max i ic BS = -----------LT
i

i

UL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
UL
UL

tx
max
req
N0
P term P term
UL
- + 1 term max (-----------------------------N tot i ic BS = -----------------------------------,0)
UL
LT
1 X i ic BS
i
tx

UL
2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status.
UL
UL
Q MAX ic BS = Q max i ic BS
UL
UL
UL
Q MAX ic BS = G macro diversity 2 links max Q max i ic BS

i
UL
UL
UL

i

UL
UL
UL

i
UL
UL
UL
Qmax i icBS
UL
have:
UL
UL
UL
Qmax i icBS Qmaxi on the other site i icBS

UL
UL
i on the same site
otherwise,
UL
UL
UL

i
UL
3rd step: Evaluation of the required quality level on uplink, Q req
405

In case of a 1xEV-DO Rev. 0 capable terminal, we have:

E UL
UL
UL
Q req = ----c-
N t min
Where
UL
Ec
---is the minimum pilot quality level on uplink. This parameter is available in the Mobility types table.
N t min
G ACK , G DRC and G TCH are respectively acknowledgement, data rate control and traffic data gains relative to the pilot. They
are defined in the terminal properties (1xEV-DO Rev. 0 tab).
In case of a 1xEV-DO Rev. A capable terminal, we have:
E UL
UL
UL
Q req = ----c-
G p 1 + G ACK + G DRC + G TCH + G RRI + G Auxiliary pilot
N t min
Where
Ec
UL
--- N t min is the minimum pilot quality level required on uplink to obtain the defined throughput, TP req . The required
UL
UL
throughput, TPreq (i.e. the uplink throughput selected for the analysis) is obtained by using a certain uplink transmission
format (i.e. 1xEV-DO radio bearer ( Index UL Bearer ) with a certain number of subframes ( n SF )) and calculated as follows:
UL
TP P R LC Index UL Bearer
UL
TP req = -----------------------------------------------------------n SF
UL
Ec
---is the value defined in the 1xEV-DO Radio Bearer Selection (Uplink) table for this uplink transmission format (radio
N t min
bearer Index, mobility and number of subframe). Two values are available for this parameter, one when the service uplink
mode is "Low Latency" and another one for high capacity services.
G ACK , G DRC , G TCH , G RRI and G Auxiliary pilot are respectively acknowledgement, data rate control, traffic data channel,
reverse rate indicator and auxiliary pilot channel gains relative to the pilot. They are defined in the terminal properties (1xEVDO Rev. A tab). Two values of G TCH are available, one when the service uplink mode is "Low Latency" and another one for
high capacity services.
req
4th step: Calculation of P term

UL
Q req
req
- P max
P term = -------------------------term
UL
Q MAX ic BS
req
max
Therefore, the service on the uplink traffic data channel is available if P term P term .
Required terminal power without ACK
Atoll also calculates the required terminal power without taking into account the ACK channel contribution. Calculations are
quite similar to those detailed in the previous paragraph, only the evaluation of the required quality on uplink is different.
E
UL
UL
Q req withoutACK = ----c-
G p 1 + G DRC + G TCH for 1xEV-DO Rev. 0 capable terminals
N t min
UL
And
E UL
UL
UL
Q req withoutACK = ----c-
G p 1 + G DRC + G TCH + G RRI + G Auxiliary pilot for 1xEV-DO Rev. A capable terminals
N t min
And then,
UL
Q req withoutACK
req
- P max
P term withoutACK = -------------------------------------term
UL
Q MAX ic BS
UL SHO gain
UL
1st step: Evaluation of the uplink quality, Q max i ic BS , for each cell of active set.
406

AT332_TRR_E0

UL
UL
Q max i
- G UL
ic BS = ----------------------------------------------------p
UL
N tot i ic BS
max
P term
UL
With P b max i ic BS = -----------LT
i

i

UL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
UL
UL

tx
max
req
N0
P term P term
UL
N tot i ic BS = -----------------------------------,0)
- + 1 term max (-----------------------------UL
LT
1 X i ic BS
i
tx

UL
2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status.
UL
Q MAX ic BS is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters of the active
set.
UL
UL
Q MAX ic BS = Q max i ic BS
UL
UL
UL

i
UL
UL
UL

i

UL
UL
UL

i
UL
UL
UL
Qmax i icBS
UL
have:
UL
UL
UL
Qmax i icBS Qmaxi on the other site i icBS

UL
UL
i on the same site
otherwise,
UL
UL
UL

i
3rd step: Calculation of the UL SHO gain
407

UL
G SHO corresponds to the uplink soft handover gain.

UL
Q MAX ic BS
UL
G SHO = ----------------------------------------------UL
max Q max i ic BS
i
Bearer Consumption
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, Atoll calculates the 1xEV-DO bearer consumption.
TPD min UL
C UL Bearer = -------------------------------------------------------------UL
TP P RLC Index UL Bearer
Where TPD min UL corresponds to the minimum bit rate required by the service in the uplink.
1xEV-DO Rev. B Service Users
For multi-carrier 1xEV-DO Rev. B users, Atoll models load balancing between carriers. Atoll shares the available terminal
power between the carriers and determines the uplink 1xEV-DO radio bearer obtained on each carrier. It starts allocating the
maximum terminal power to the best carrier and selects the highest 1xEV-DO radio bearer. If it remains terminal power after
serving the first carrier, then Atoll continues allocating the available terminal power to the second carrier, and so on for the
other carriers of the active set as long as the remaining terminal power is sufficient to obtain the lowest bearer.
The following results are displayed:
For each carrier used in the selected configuration, Atoll indicates the UL SHO Gain, the obtained throughput and the
required power.
The calculations can be divided into four steps:

UL
1st step: Evaluation of the uplink quality, Q max i ic , for each cell of the sub-active set
For each cell (i,ic), we have:
UL
term P b max i ic
UL
- G UL
Q max i ic = -----------------------------------------------p
UL
N tot i ic
max
P term ic
UL
With P b max i ic = --------------------LT
i
max
P term ic is the terminal power available for the carrier (ic).

i

UL
L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx G term
UL
UL
N tot i ic is the total noise at the transmitter on the carrier ic. This value is deduced from the cell uplink load factor X i ic .
tx
max
req
N0
P term ic P term
UL
N tot i ic = -----------------------------,0)
- + 1 term max (---------------------------------------UL
LT
1 X i ic
i
tx

UL
2nd step: Calculation of the total quality at the transmitter on ic ( Q MAX ) based on the receiver handover status.
UL
UL
Q MAX ic = Q max i ic
UL
UL
UL
Q MAX ic = G macro diversity 2 links max Q max i ic

i
408

AT332_TRR_E0
UL
UL
UL
max Q max i ic corresponds to the highest Q max i ic value.

i

UL
UL
UL

i
UL
UL
UL
Q MAX ic = f rake efficiency
Qmax i ic
UL
have:
UL
UL
UL
Q MAX ic = G macro diversity 2 links max f rake efficiency
Qmax i ic Qmaxi on the other site i ic

UL
UL
i on the same site
otherwise,
UL
UL
UL

i
UL
3rd step: Calculation of the UL SHO gain ( G SHO )

UL
Q MAX ic
UL
G SHO = ------------------------------------------UL
max Q max i ic
i
4th step: Selection of the uplink 1xEV-DO radio bearer

UL
req
Atoll evaluates of the required quality level in the uplink ( Q req ) and the required terminal power ( P term ic ) for each 1xEVDO radio bearer.
E UL
UL
UL
Q req = ----c-
N t min
Where
UL
Ec
---is the minimum pilot quality level required in the uplink to obtain the 1xEV-DO radio bearer. The values are defined
N t min
in the 1xEV-DO Radio Bearer Selection (Uplink) table for each uplink transmission format (radio bearer Index, mobility and
number of subframe). Two values are available, one when the service uplink mode is "Low Latency" and another one for high
capacity services.
And
UL
Q req
req
- P max
P term ic = --------------------term ic
UL
Q MAX ic
Then, Atoll selects the best 1xEV-DO radio bearer. This is the 1xEV-DO radio bearer ( Index UL Bearer ) with the highest
UL

UL
obtained throughput ( TP ic = -------------------------------------------------------------) where:
n SF Index UL Bearer
req
max
P term ic P term ic ,
409

And the required modulation scheme is supported by the terminal.
n SF is the number of subframes associated with the 1xEV-DO radio bearer ( Index UL Bearer ).
Max terminal power

max
allowed.
Required throughput
UL
The required throughput, R req , is the uplink throughput selected for the analysis.
Obtained throughput
Atoll calculates the throughput for all combinations of carriers.

m
UL
TP total =
TP
UL
ic where m corresponds to the number of carriers in the combination.
ic = 1
The obtained throughput ( TP
UL
) corresponds to the best configuration among all combinations of carriers, i.e., the
UL
combination which provides the highest throughput, Max TP total .

The traffic data channel is available in uplink if TP
UL
UL
TP req .
Required terminal power

m
req
P term
Pterm ic
req
ic = 1

Atoll calculates CDMA-specific coverage studies on each pixel where the pilot signal level exceeds the minimum RSCP
threshold. Let us assume each pixel of the map corresponds to a probe receiver with a terminal, a mobility type and a service.
This receiver does not create any interference. You can make the coverage prediction for a specific carrier or for the best
1xRTT or 1xEV-DO carrier. The type of carrier and the carriers you can select depend on the service and on the frequency
band(s) supported by the terminal. Coverage studies are based on the uplink load percentage and the downlink total power
of cells. These parameters can either be either simulation results, or average values calculated from a group of simulations,
or user-defined cell inputs.
5.5.2.1 Pilot Quality Analysis

For further details on calculation formulas, see "Definitions and Formulas" on page 338. For further details on calculations,
see "Bar Graph and Pilot Sub-Menu" on page 390
The carrier that can be used by transmitters is fixed. In this case, for each transmitter i containing the receiver in its calculation
area and using the selected carrier, Atoll calculates pilot quality at the receiver on this carrier icgiven. Then, it determines the
best serving transmitter BS using the carrier icgiven ( Q pilot ic given ) and deduces the best pilot quality received with a fixed
BS
cell edge coverage probability,
Resulting
Q pilot
ic given
2nd case: Analysis based on the best carrier of all frequency bands
Atoll proceeds as in point analysis. It determines the best carrier of each transmitter i containing the receiver in its calculation
area and using a frequency band supported by the receivers terminal. The best carrier selection depends on the option
selected for the site equipment (UL minimum noise, DL minimum power, random, sequential) and is based on the UL load
percentage and the downlink total power of cells (simulation results or cell properties). Atoll calculates the pilot quality at the
receiver from these transmitters on their best carrier and determines the best serving transmitter BS on its best carrier icBS
Resulting
( Q pilot ic BS ). Then, it deduces the best pilot quality received with a fixed cell edge coverage probability, Q pilot
BS
ic BS .
3rd case: Analysis based on the best carrier of any frequency band (for multi-band terminals with priority defined on frequency
bands only)
410

AT332_TRR_E0
The frequency band that can be used is fixed. Atoll determines the best carrier of each transmitter i containing the receiver
in its calculation area and using the selected frequency band. The best carrier selection depends on the option selected for
the site equipment (UL minimum noise, DL minimum power, random, sequential) and is based on the UL load percentage and
the downlink total power of cells (simulation results or cell properties). Then, Atoll calculates the pilot quality at the receiver
from these transmitters on their best carrier and determines the best serving transmitter BS on its best carrier icBS
Resulting
( Q pilot ic BS ). Then, it calculates the best pilot quality received with a fixed cell edge coverage probability, Q pilot
BS
ic BS .
Single colour
Resulting
req
ic Q pilot . Coverage consists of a single layer with a unique colour.
ic = ic BS or ic given

Resulting
req
ic Q pilot ( ic = ic BS or ic given ). Coverage consists of several layers with associated
colours. There is a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to the best
serving transmitter BS.
Colour per mobility
In this case, the receiver is not completely defined and no mobility assigned. Coverage consists of several layers with a layer
per user-defined mobility type defined in the Mobility Types sub-folder. For each layer, area is covered if
Resulting
Q pilot
req
ic Q pilot ( ic = ic BS or ic given ). Each layer is assigned a colour and displayed with intersections between layers.
and the All option in the Condition tab of prediction properties).
Coverage consists of several layers with a layer per user-defined probability level defined in the Display tab (Prediction
Resulting
req
ic Q pilot ( ic = ic BS or ic given ) in the required number of
simulations. Each layer is assigned a colour and displayed with intersections between layers.
Coverage consists of several layers with a layer per user-defined cell edge coverage probability, p, defined in the Display tab
Resulting
(Prediction properties). For each layer, area is covered if Q pilot
req
ic p Q pilot ( ic = ic BS or ic given ). Each layer is assigned
a colour and displayed with intersections between layers.
Colour per quality level (Ec/I0)
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
Resulting
ic Q pilot threshold ( ic = ic BS or ic given ). Each layer is assigned a
colour and displayed with intersections between layers.
Colour per quality margin (Ec/I0 margin)
Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction
Resulting
req
ic Q pilot Q pilot m arg in ( ic = ic BS or ic given ). Each layer is
assigned a colour and displayed with intersections between layers.

Colour per pilot signal level (Ec)
Coverage consists of several layers with a layer per user-defined pilot signal level defined in the Display tab (Prediction
Resulting
ic Q pilot threshold ( ic = ic BS or ic given ). Each layer is assigned a
5.5.2.2 Downlink Service Area Analysis

The downlink service area analysis depends on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO). Several
display options are available when calculating this study, some of which are dedicated to CDMA2000 1xRTT networks while
others are relevant when analysing CDMA2000 1xEV-DO systems only.
5.5.2.2.1
CDMA2000 1xRTT
As in point analysis, Atoll calculates downlink quality on FCH at the receiver for each cell (k,ic) (with ic=icBS or icgiven) (these
cells are the first f cells in the receivers active set and f is the number of fingers defined for the terminal). No power control
411

is performed as in simulations. Here, Atoll determines the downlink quality on FCH at the receiver for a maximum traffic
channel power per transmitter allowed on the fundamental channel (FCH). Then, the total downlink quality on FCH
DL
( Q MAX ic FCH ) is evaluated after recombination.

Best server and active set determination is performed as in point prediction.
Atoll displays total traffic channel quality at the receiver on the carrier ic ( ic BS or ic given ).
For further details on formulas, see "Definitions and Formulas" on page 338. For further details on calculation, see "Downlink
Sub-Menu" on page 393.
You may choose following display options:
Single colour
DL
DL
DL
Atoll displays a coverage with a unique colour if Q MAX ic FCH Q req FCH . Q req FCH is the downlink traffic data quality
target on the fundamental channel (FCH). This parameter is user-defined for a given service and a terminal in the Services subfolder.

DL
DL
Atoll displays a coverage if Q MAX ic FCH Q req FCH . Coverage consists of several layers with associated colours. There is
a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to best serving transmitter.
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned. Coverage consists of several layers with a layer per
DL
DL
user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if Q MAX ic FCH Q req FCH . Each layer
is assigned a colour and displayed with intersections between layers.
Colour per service
In this case, receiver is not completely defined and no service is assigned. Coverage consists of several layers with a layer per
DL
DL
user-defined service defined in Services sub-folder. For each layer, area is covered if Q MAX ic FCH Q req FCH . Each layer
and the All option in the Condition tab of prediction properties).
Coverage consists of several layers with a layer per user-defined probability level defined in the Display tab (Prediction
DL
DL
properties). For each layer, area is covered if Q MAX ic FCH Q req FCH in the required number of simulations. Each layer is
DL
DL
(Prediction properties). For each layer, area is covered if Q MAX ic p FCH Q req FCH . Each layer is assigned a colour and
displayed with intersections between layers.
Colour per maximum quality level (max Eb/Nt)
DL
properties). For each layer, area is covered if Q MAX ic FCH Threshold . Each layer is assigned a colour and displayed with
intersections between layers.
DL
properties). For each layer, area is covered if Q eff ic FCH Threshold . Each layer is assigned a colour and displayed with
412

AT332_TRR_E0
DL
DL
properties). For each layer, area is covered if Q MAX ic FCH Q req FCH M arg in . Each layer is assigned a colour and

req
Atoll calculates the downlink power required on FCH, P FCH ic , as follows:

DL
Q req FCH
req
- P max
P FCH ic = ---------------------FCH
DL
Q MAX ic
max
Where P FCH is a user-defined input for a given service and terminal. It corresponds to the maximum traffic data power
allowed on FCH for a transmitter.
Coverage consists of several layers with a layer per user-defined required power threshold defined in the Display tab
req
(Prediction properties). For each layer, area is covered if P FCH ic Threshold . Each layer is assigned a colour and displayed
with intersections between layers.
Coverage consists of several layers with a layer per user-defined power margin defined in the Display tab (Prediction
max
req
properties). For each layer, area is covered if P FCH P FCH ic M arg in . Each layer is assigned a colour and displayed with
Colour per throughput
This display option is relevant for CDMA2000 1xRTT data services only. For each possible throughput, TP
FCH
DL
FCH
DL
FCH
DL
FCH
DL
FCH
DL
( TP P DL AF FCH ,
DL
TP P DL AF FCH + 2 , TP P DL AF FCH + 4 , TPP DL AF FCH + 8 , TP P DL AF FCH + 16 ), Atoll calculates traffic

channel quality at the receiver for each cell (k,ic) (with ic=icBS or icgiven). Downlink traffic channel quality at the receiver is
evaluated from a maximum traffic channel power per transmitter allowed for the corresponding throughput. Then, the total
DL
DL
downlink traffic channel quality ( Q MAX ic TP ) is calculated after recombination.

Coverage consists of several layers with a layer per possible throughput, TP
DL
Q MAX ic
DL
DL
TP
DL
DL
Q req TP
DL
. For each layer, area is covered if
. Each layer is assigned a colour and displayed with intersections between layers.
DL
Q req TP is the downlink traffic data quality target for the throughput, TP
DL
. This parameter is user-defined for a given
service, terminal and throughput in the Services sub-folder.
5.5.2.2.2
CDMA2000 1xEV-DO
E
As in point analysis, Atoll calculates the effective pilot quality level at the receiver from the best server cell, ----c- ic b pilot . Best
Nt
server and active set determination is performed as in point prediction (AS analysis). Then, from this value, it determines the
effective downlink throughput received, TP
DL
For further details on formulas, see "Definitions and Formulas" on page 338. For further details on calculations, see "Downlink
1xEV-DO Rev. 0 Users
For 1xEV-DO Rev. 0 users (users with EV-DO Rev. 0-capable terminals and EV-DO Rev. 0 services), the obtained throughput
( TP
DL
) on downlink is determined from the graph Peak throughput=f(C/I) (Rev0) defined for the mobility type selected in
DL
the Condition tab (Prediction properties). TP is the value read in the graph Peak throughput=f(C/I) (Rev0) for the
E
calculated effective pilot quality level, ----c- ic BS b pilot .
Nt
1xEV-DO Rev. A Users
For 1xEV-DO Rev. A users (users with EV-DO Rev. A-capable terminals and EV-DO Rev. A services), the obtained throughput
( TP
DL
) on downlink depends on the downlink transmission format, i.e the radio bearer index ( Index DL Bearer ) with the
413

E
C
number of timeslots ( n TS ). Atoll selects the downlink transmission format where ----c- ic BS b pilot --- . Then, it
I req
Nt
determines the downlink obtained throughput as follows:
TP
DL
DL

= -------------------------------------------------------------.
n TS
The obtained throughput corresponds to the guaranteed throughput after a certain number of retransmissions (i.e. the
number of timeslots, n TS ).
DL
throughput is an average throughput ( TP av ) calculated as follows:

DL

DL
TP av = --------------------------------------------------------------------DL
n Rtx (Index DL Bearer,n TS) av
DL
The average number of retransmissions ( n Rtx av ) is determined from early termination probabilities defined for the selected
downlink transmission format. The Early Termination Probability graph shows the probability of early termination ( p ) as a
DL
DL
function of the number of retransmissions ( n Rtx ). Atoll calculates the average number of retransmissions ( n Rtx av ) as
follows:
n DL
Rtx max
DL
DL
DL
DL
DL
p n Rtx p n Rtx 1 n Rtx
=1
Rtx
n Rtx av = -------------------------------------------------------------------------------------------DL
p n Rtx max
1xEV-DO Rev. B Users

Single-carrier EV-DO Rev. B service users are managed as 1xEV-DO Rev. A service users.
For multi-carrier EV-DO Rev. B service users, the obtained throughput ( TP
obtained throughputs on each carrier.
DL
) in the downlink corresponds to the sum of the
DL
The obtained throughput ( TP ic ) on a carrier depends on the downlink transmission format, i.e the radio bearer index
( Index DL Bearer ) with the number of timeslots ( n TS ). Atoll selects the downlink transmission format where
E
C
----c- ic b pilot ---
and whose modulation scheme is supported by the terminal.
I req
Nt
The downlink obtained throughput corresponds to the guaranteed throughput after a certain number of retransmissions (i.e.
the number of timeslots, n TS ). It is determined as follows:
DL

DL
TP ic = -------------------------------------------------------------n TS
DL
throughput on a carrier is an average throughput ( TP av ic ) calculated as follows:

DL

DL
TP av ic = --------------------------------------------------------------------DL
n Rtx (Index DL Bearer,n TS) av
DL
downlink transmission format. The Early Termination Probability graph shows the probability of early termination ( p ) as a
DL
DL
function of the number of retransmissions ( n Rtx ). Atoll calculates the average number of retransmissions ( n Rtx av ) as
follows:
414

AT332_TRR_E0
n DL
Rtx max
DL
DL
DL
DL
DL
=1
Rtx
n Rtx av = -------------------------------------------------------------------------------------------DL
p n Rtx max
DL
The average throughput ( TP av ) provided on downlink corresponds to the sum of the average throughputs obtained on each
carrier.
Display Options
You may choose the following display options:
Colour per C/I
Coverage consists of several layers with a layer per quality threshold defined in the Display tab (Prediction properties). For
E
each layer, area is covered if ----c- ic b pilot Threshold . Each layer is assigned a colour and displayed with intersections
Nt
between layers.
Coverage consists of several layers with a layer per possible throughput ( TP

throughput, TP
DL
DL
). For each layer, area is covered if the
, can be obtained. Each layer is assigned a colour and displayed with intersections between layers.
Colour per average throughput
This display option is available for 1xEV-DO Rev. A and 1xEV-DO Rev. B users only. It enables you to view the obtained downlink
DL
throughput when HARQ is used. Coverage consists of several layers with a layer per possible average throughput ( TP av ). For
DL
each layer, area is covered if the average throughput, TP av , can be obtained. Each layer is assigned a colour and displayed
5.5.2.3 Uplink Service Area Analysis

The results displayed when calculating the uplink service area analysis depend on the studied network (CDMA2000 1xRTT or
CDMA2000 1xEV-DO).
5.5.2.3.1
CDMA2000 1xRTT
As in point analysis, Atoll calculates uplink quality on FCH from receiver for each cell (l,ic) (with ic=icBS or icgiven) in receiver
active set. No power control simulation is performed. Atoll determines uplink quality on FCH at the transmitter for the
UL
maximum terminal power. Then, the total uplink traffic channel quality ( Q MAX ic FCH ) is evaluated with respect to the
receiver handover status.
Best server and active set determination is performed as in point prediction (AS analysis).
Atoll displays uplink quality on FCH at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the receiver.
For further details on formulas, see "Definitions and Formulas" on page 338. For further details on calculations, see "Uplink
Single colour
UL
UL
UL
Atoll displays a coverage if Q MAX ic FCH Q req FCH . Coverage colour is unique. Q req FCH is the uplink data traffic quality
target on the fundamental channel (FCH). This parameter is user-defined for a given service and a terminal in the Services subfolder.

UL
UL
Atoll displays a coverage if Q MAX ic FCH Q req FCH . Coverage consists of several layers with associated colours. There is
a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to best server transmitter.
Colour per mobility
415

UL
UL
user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if Q MAX ic FCH Q req FCH . Each layer
Colour per service
UL
UL
user-defined service defined in Services sub-folder. For each layer, area is covered if Q MAX ic FCH Q req FCH . Each layer
and the All option in the Condition tab of prediction properties). Coverage consists of several layers with a layer per userdefined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if
UL
UL
Q MAX ic FCH Q req FCH in the required number of simulations. Each layer is assigned a colour and displayed with
UL
UL
(Prediction properties). For each layer, area is covered if Q MAX ic p FCH Q req FCH . Each layer is assigned a colour and
UL
properties). For each layer, area is covered if Q MAX ic FCH Threshold . Each layer is assigned a colour and displayed with
UL
properties). For each layer, area is covered if Q effective ic FCH Threshold . Each layer is assigned a colour and displayed
UL
UL
properties). For each layer, area is covered if Q MAX ic FCH Q req FCH M arg in . Each layer is assigned a colour and

FCH re q
The required terminal power, P term
, is calculated as described in the Point analysis AS analysis tab Uplink sub-menu
part. Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction
FCH re q
properties). For each layer, area is covered if P term
ic Threshold . Each layer is assigned a colour and displayed with
max
FCH re q
properties). For each layer, area is covered if P term P term
ic M arg in . Each layer is assigned a colour and displayed
This display option is relevant for CDMA2000 1xRTT data services only. For each possible throughput, TP
FCH
UL
FCH
UL
FCH
UL
FCH
UL
FCH
UL
( TP P UL AF FCH ,
UL
TP P UL AF FCH + 2 , TP P UL AF FCH + 4 , TP P UL AF FCH + 8 , TP P UL AF FCH + 16 ), Atoll calculates the total

UL
UL
uplink traffic channel quality ( Q MAX ic TP ). Coverage consists of several layers with a layer per possible throughput,
TP
UL
UL
UL
UL
UL
. For each layer, area is covered if Q MAX ic TP Q req TP . Each layer is assigned a colour and displayed with
UL
UL
intersections between layers. Q req TP is the uplink traffic data quality target for the throughput, TP
user-defined for the service, a given terminal and throughput in the service properties.
416
UL
. This parameter is

AT332_TRR_E0
5.5.2.3.2
CDMA2000 1xEV-DO
As in point analysis, Atoll calculates the uplink quality from receiver for each cell (l,ic) (with ic=icBS or icgiven) in receiver active
set. No power control simulation is performed. For 1xEV-DO Rev. 0 users, Atoll determines the uplink quality at the
transmitter for the maximum terminal power allowed and an uplink channel throughput of 9.6 kbps. For 1xEV-DO Rev. A and
1xEV-DO Rev. B users, Atoll determines the uplink quality at the transmitter for the maximum terminal power allowed and an
UL
uplink channel throughput of 4.8 kbps. Then, the total uplink quality ( Q MAX ic ) is evaluated with respect to the receiver
handover status.
Best server and active set determination is performed as in point prediction (AS analysis).
Atoll displays the uplink quality at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the receiver. For
multi-carrier EV-DO users, Atoll considers the best sub-active set.
For further details on formulas, see "Definitions and Formulas" on page 338. For further details on calculations, see "Uplink
Single colour
UL
UL
UL
Atoll displays a coverage if Q MAX ic Q req . Coverage colour is unique. For 1xEV-DO Rev. 0 users, Q req is the quality required
UL
on uplink for a 9.6 kbps channel throughput. For 1xEV-DO Rev. A and 1xEV-DO Rev. B users, Q req is the quality required on
uplink for a 4.8 kbps channel throughput. This parameter is calculated from the minimum uplink pilot quality and gains on the
different uplink channels.
We have:
E UL
UL
UL
Q req = ----c-
G p 1 + G ACK + G DRC + G TCH for 1xEV-DO Rev. 0 terminals,
N t min
And
E UL
UL
UL
Q req = ----c-
G p 1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary Pilot for 1xEV-DO Rev. A and 1xEV-DO Rev. B terminals.
N t min

UL
UL
Atoll displays a coverage if Q MAX ic Q req . Coverage consists of several layers with associated colours. There is a layer per
transmitter with no intersection between layers. Layer colour is the colour assigned to best server transmitter.
Colour per mobility
UL
UL
user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if Q MAX ic Q req . Each layer is assigned
Colour per service
UL
UL
user-defined service defined in Services sub-folder. For each layer, area is covered if Q MAX ic Q req . Each layer is assigned
and the All option in the Condition tab of prediction properties). Coverage consists of several layers with a layer per userUL
UL
defined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if Q MAX ic Q req
in the required number of simulations. Each layer is assigned a colour and displayed with intersections between layers.
UL
UL
(Prediction properties). For each layer, area is covered if Q MAX ic p Q req . Each layer is assigned a colour and displayed
417


UL
Here, Atoll calculates the total uplink traffic channel quality ( Q MAX ic TCH ).
UL
Q req TCH
UL
- P max
Q MAX ic TCH = ---------------------term
req
P term
With
E UL
UL
UL
Q req TCH = ----c-
G p G TCH
N t min
UL
properties). For each layer, area is covered if Q MAX ic TCH Threshold . Each layer is assigned a colour and displayed with
UL
properties). For each layer, area is covered if Q effective ic TCH Threshold . Each layer is assigned a colour and displayed
UL
UL
UL
Q effective ic TCH = min Q MAX ic TCH Q req TCH
UL
UL
properties). For each layer, area is covered if Q MAX ic Q req M arg in . Each layer is assigned a colour and displayed with
1xEV-DO Rev. 0, 1xEV-DO Rev.A and single-carrier 1xEV-DO Rev. B service users
Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction
TCH re q
properties). For each layer, area is covered if P term
ic Threshold . Each layer is assigned a colour and displayed with

TCH re q
The required terminal power on traffic data channel, P term
, is calculated as described in the Point analysis AS analysis
tab Uplink sub-menu part.

TCH re q
P term
req
P term
- G TCH for 1xEV-DO Rev. 0 terminals,
= -------------------------------------------------------1 + G ACK + G DRC + G TCH
And
TCH re q
P term
req
P term
- G TCH for 1xEV-DO Rev. A terminals.
= ------------------------------------------------------------------------------------------------------------------1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary Pilot
Multi-carrier 1xEV-DO Rev. B service users

For multi-carrier EV-DO users, the coverage consists of several layers with a layer per user-defined power threshold defined
TCH re q
in the Display tab (Prediction properties). For each layer, area is covered if P term
Threshold . Each layer is assigned a

TCH re q
For the selected configuration (i.e., the combination of carriers which provides the highest throughput), P term
corresponds
to the sum of the terminal powers required on each carrier of the configuration.
max
req
properties). For each layer, area is covered if P term P term ic M arg in . Each layer is assigned a colour and displayed with
1xEV-DO Rev. 0 service users
418

AT332_TRR_E0
For each possible throughput, TP
UL
UL
UL
, Atoll calculates the total uplink quality ( Q MAX ic TP ). Coverage consists of several
UL
UL
UL
UL
layers with a layer per possible throughput. For each layer, area is covered if Q MAX ic TP Q req TP . Each layer is
UL
UL
Q req TP is the uplink quality required to obtain the throughput, TP

The possible throughputs on uplink, TP
UL
UL
, are: 9.6, 19.2, 38.4, 76.8 and 153.6 kbps
E UL
UL
UL
Q req = ----c-
N t min
Where
Ec
--- N t min is the minimum pilot quality level on uplink. This parameter is available in the Mobility types table.
UL
G ACK , G DRC and G TCH are respectively acknowledgement, data rate control and traffic data gains relative to the pilot. They
are defined in the terminal properties (1xEV-DO Rev. 0 tab).
1xEV-DO Rev. A and single-carrier 1xEV-DO Rev. B service users
For each possible throughput, TP
UL
UL
UL
, Atoll calculates the total uplink quality ( Q MAX ic TP ). Coverage consists of several
UL
UL
UL
UL
layers with a layer per possible throughput. For each layer, area is covered if Q MAX ic TP Q req R v . Each layer is
UL
UL
Q req TP is the uplink quality required to obtain the throughput, TP

The throughput, TP
UL
UL
is obtained when a certain uplink transmission format (i.e. 1xEV-DO radio bearer ( Index UL Bearer )
with a certain number of subframes ( n SF )) is used. It is calculated as follows:

UL

UL
TP req = -------------------------------------------------------------n SF
E
UL
UL
Q req = ----c-
N t min
UL
Where
UL
Ec
UL
---is the minimum pilot quality level required on uplink to obtain the throughput, TP . The value is defined in the
N t min
1xEV-DO Radio Bearer Selection (Uplink) table for the uplink transmission format (radio bearer Index, mobility and number of
subframe). Two values are available for this parameter, one when the service uplink mode is "Low Latency" and another one
for high capacity services.
For multi-carrier 1xEV-DO Rev. B users, Atoll models load balancing between carriers. Atoll allocates the available terminal
power to carriers sequentially and determines the uplink 1xEV-DO radio bearer obtained on each carrier. Then, it selects the
best configuration among all combinations of carriers, i.e., the combination which provides the highest throughput.
Coverage consists of several layers with a layer per possible throughput. For each layer, area is covered if TP
UL
UL
TP req . Each
layer is assigned a colour and displayed with intersections between layers.

UL
TP req is the uplink throughput associated with the layer.

TP
UL
corresponds to the throughput of the best configuration, i.e., the combination which provides the highest throughput.
Colour per average throughput
419

This display option is available for 1xEV-DO Rev. A and 1xEV-DO Rev. B users only. When HARQ (Hybrid Automatic Repeat
Request) is used, the required average number of retransmissions is smaller and the throughput is an average throughput
UL
( TP av ) calculated as follows:
UL

UL
TP av = ----------------------------------------------------------------------UL
n Rtx Index UL Bearer n SF av
UL
uplink transmission format (i.e. the radio bearer index ( Index UL Bearer ) with the number of subframes ( n SF )). The Early
Termination Probability graph shows the probability of early termination ( p ) as a function of the number of retransmissions
UL
UL
( n Rtx ). Atoll calculates the average number of retransmissions ( n Rtx av ) as follows:

n UL
Rtx max
UL
UL
UL
UL
UL
=1
Rtx
n Rtx av = -------------------------------------------------------------------------------------------UL
p n Rtx max
1xEV-DO Rev. A and single-carrier 1xEV-DO Rev. B service users

UL
UL
UL
For each possible average throughput, TP av , Atoll calculates the total uplink quality ( Q MAX ic TP av ). Coverage consists of
UL
UL
UL
UL
several layers with a layer per possible average throughput. For each layer, area is covered if Q MAX ic TP av Q req TP av .
Each layer is assigned a colour and displayed with intersections between layers.
UL
UL
UL
Q req TP av is the uplink quality required to obtain the average throughput, TP av .

For multi-carrier 1xEV-DO Rev. B users, the coverage consists of several layers with a layer per possible throughput. For each
UL
UL
layer, area is covered if TP av TP req . Each layer is assigned a colour and displayed with intersections between layers.
UL
TP req is the uplink throughput associated with the layer.

UL
For the selected configuration (i.e., the combination of carriers which provides the highest throughput), TP av corresponds to
the sum of the average throughputs obtained on each carrier of the configuration.
5.5.2.4 Downlink Total Noise Analysis

Atoll determines downlink total noise generated by cells.
For CDMA2000 1xRTT systems, we have:
Ptot icadj
DL
DL
N tot ic =
txj j
+ N0
Ptot ic + -----------------------------------RF ic ic adj
DL
term
txj j
For CDMA2000 1xEV-DO systems, we have:
Ptot icadj bpilot

DL
DL
N tot ic =
txj j
- + N0
Ptot ic bpilot + --------------------------------------------------RF ic ic adj
DL
term
txj j
term
N0
DL
-
Downlink noise rise, NR DL ic , is calculated from the downlink total noise, N tot , as: NR DL ic = 10 log ----------- N DL
tot
5.5.2.4.1
Analysis on the Best Carrier

If the best carrier is selected, Atoll determines DL total noise for the best carrier. Then, allows the user to choose different
displays.
420
Colour per minimum noise level

AT332_TRR_E0
Coverage consists of several layers with a layer per user-defined noise level defined in the Display tab (Prediction properties).
DL
For each layer, area is covered if min NR tot ic Threshold . Each layer is assigned a colour and displayed with
ic
Colour per maximum noise level
DL
For each layer, area is covered if max NR tot ic Threshold . Each layer is assigned a colour and displayed with intersections
ic
between layers.
Colour per average noise level
DL
For each layer, area is covered if average NRtot ic Threshold . Each layer is assigned a colour and displayed with
ic
Colour per minimum noise rise
Atoll displays bins where min NR DL ic Threshold . Coverage consists of several areas with an area per user-defined noise
ic
rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
Colour per maximum noise rise
Atoll displays bins where max NR DL ic Threshold . Coverage consists of several areas with an area per user-defined noise
ic
rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
Colour per average noise rise
Atoll displays bins where average NRDL ic Threshold . Coverage consists of several areas with an area per user-defined
ic
noise rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
5.5.2.4.2
Analysis on a Specific Carrier

When only one carrier is analysed, Atoll determines DL total noise or DL noise rise on this carrier. In this case, the displayed
coverage is the same for any selected display per noise level (average, minimum or maximum) or any display per noise rise
(average, minimum or maximum).
Colour per noise level
DL
For each layer, area is covered if N tot ic Threshold . Each layer is assigned a colour and displayed with intersections
between layers.
Colour per noise rise
Atoll displays bins where NRDL ic Threshold . Coverage consists of several areas with an area per user-defined noise rise
threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.

Atoll permits the automatic allocation of intra-technology neighbours in the current network. Two allocation algorithms are
available, one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours.
The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all the
cells of TBC transmitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfill the following conditions:
They are active,

421

If no focus zone exists in the ATL document, Atoll takes into account the computation
zone.
"Neighbour Allocation for all Transmitters" on page 422.

5.6.1 Neighbour Allocation for all Transmitters

We assume that we have a reference cell A and a candidate neighbour, cell B. When automatic allocation starts, Atoll checks
The distance between both cells must be less than the user-definable maximum inter-site distance. If the distance
discarded.
"Calculation of the Inter-Transmitter Distance" on page 428. Otherwise, it takes the real distance.
[Neighbours]

Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may
choose one or more carriers. Atoll will allocate neighbours to cells using the selected carriers.
Force co-site cells as neighbours: This option enables you to force cells located on the reference cell site in the
candidate neighbour list. This constraints can be weighted among the others and ranks the neighbours through
the importance field (see after).
Force adjacent cells as neighbours (only for intra-carrier neighbours): This option enables you to force cells
geographically adjacent to the reference cell in the candidate neighbour list.This constraints can be weighted
among the others and ranks the neighbours through the importance field (see after).
reference cell is a candidate neighbour of another cell, this one will be considered as candidate neighbour of the
reference cell.
[Neighbours]
422

AT332_TRR_E0
Let CellB be a candidate neighbour cell of CellA. CellB is considered adjacent to CellA if
there exists at least one pixel in the CellA Best Server coverage area where CellB is Best
Server (if several cells have the same best server value) or CellB is the second best server
that enters the Active Set (respecting the T_Drop of the allocation).
When the Force adjacent cells as neighbours check box is selected, adjacent cells are
sorted and listed from the most adjacent to the least, depending on the above criterion.
Adjacence is relative to the number of pixels satisfying the criterion.
Intra-carrier neighbours: intra-carrier handover is a soft handover.
The reference cell A and the candidate cell B are located inside a continuous layer of cells with carrier c1 (c1 is the selected
carrier on which you run the allocation).
SA is the area where the cell A is the best serving cell. It means that the cell A is the first one in the active set.
The pilot signal received from the cell A is greater than the minimum pilot signal level.
The pilot quality from A exceeds Min. Ec/I0.
The pilot quality from A is the best.
SB is the area where the cell B can enter the active set.
The pilot signal received from the cell B is greater than the minimum pilot signal level.
The pilot quality from B is greater than T_Drop.
Inter-carrier neighbours: inter-frequency handover is a hard handover. It is needed in a multi-carrier (1xRTT and 1xEVDO carriers) CDMA network:
To balance loading between carriers and layers (1st case),

To make a coverage reason handover from micro cell frequency to macro cells (2nd case).
1st case: the reference cell A is located inside a continuous layer of cells with carrier c1 (c1 is the selected carrier on which you
run the allocation) and the candidate cell B belongs to a layer of cells with carrier c2.
SA is the area where:
The pilot signal from A is not the highest one. It is strictly lower than the best pilot signal received and higher than
the best pilot signal minus the margin.
SB is the area where:
The pilot signal from B is the highest one.
Figure 5.4: Overlapping Zones - 1st Case
423

2nd case: the reference cell A is located on the border of a layer with carrier c1 (c1 is the selected carrier on which you run the
allocation) and the candidate cell B belongs to a layer of cells with carrier c2.
SA is the area where:
The pilot signal from A is the highest one
The pilot signal from A is lower than the minimum pilot signal level plus the margin.
SB is the area where:
The pilot signal from B is the highest one.
Figure 5.5: Overlapping Zones - 2nd Case

Two ways enable you to determine the I0 value:
Global Value: A percentage of the cell maximum power is considered. If the % of

maximum power is too low, i.e. if % Pmax P pilot , Atoll takes into account the
pilot power of the cell. Then, I0 represents the sum of values calculated for each cell.
Defined per Cell: Atoll takes into account the total downlink power defined per cell.
I0 represents the sum of total transmitted powers.
SA SB
SA
The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see after).
allocated to each transmitter is exceeded. If we consider the case for which there are 15 candidate neighbours and the
maximum number of neighbours to be allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8
(having the highest importance values) will be allocated to the reference cell. Note that specific maximum numbers of
neighbours (maximum number of intra-carrier neighbours, maximum number of inter-carrier neighbours) can be defined at
the cell level (property dialog box or cell table). If defined there, this value is taken into account instead of the default one
available in the Neighbour Allocation dialog box.
neighbour may be marked as exceptional pair, co-site, adjacent, coverage or symmetric. For neighbours accepted for co-site,
adjacency and coverage reasons, Atoll displays the percentage of area meeting the coverage conditions and the
corresponding surface area (km2), the percentage of area meeting the adjacency conditions and the corresponding surface
area (km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing.
424

AT332_TRR_E0
No simulation or prediction study is needed to perform an automatic neighbour

allocation. When starting an automatic neighbour allocation, Atoll automatically
calculates the path loss matrices if not found.
Even if no specific terminal, mobility or service is selected in the automatic allocation,
it is interesting to know that the algorithm works such as finding the maximum
number of neighbours by selection the multi-service traffic data as follows:
Service: selection of the one with the lowest body loss.
Mobility: no impact on the allocation, no specific selection.
Terminal: selection of the one with the greatest (Gain - Loss) value, and, if equal, the
one with the lowest noise figure.
The neighbour lists may be optionally used in the power control simulations to
determine the mobile's active set.
The force neighbour symmetry option enables the users to consider the reciprocity of
a neighbourhood link. This reciprocity is allowed only if the neighbour list is not
already full. Thus, if the cell B is a neighbour of the cell A while the cell A is not a
neighbour of the cell B, two cases are possible:
1st case: There is space in the cell B neighbour list: the cell A will be added to the list.
It will be the last one.
2nd case: The cell B neighbour list is full: Atoll will not include cell A in the list and will
cancel the link by deleting cell B from the cell A neighbour list.
When the options Force exceptional pairs and Force symmetry are selected, Atoll
considers the constraints between exceptional pairs in both directions so as to
respect symmetry condition. On the other hand, if neighbourhood relationship is
forced in one direction and forbidden in the other one, symmetry cannot be
respected. In this case, Atoll displays a warning in the Event viewer.

Transmitter
TBA cells,
Neighbours of TBA cells marked as exceptional pair, adjacent and symmetric,
Neighbours of TBA cells that satisfy coverage conditions.
Automatic neighbour allocation parameters are described in "Neighbour Allocation for all Transmitters" on page 422.

Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason, and to
quantify the neighbour importance.
5.6.3.1 Importance of Intra-carrier Neighbours

Neighbourhood cause
When
Importance value
Existing neighbour

Existing importance
Exceptional pair
100 %
Co-site cell

selected
Adjacent cell

selected
425

Neighbourhood cause
When
Importance value

relationship
transmitter.
d Di = 1 ---------d max
on page 428.
[Neighbours]

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
30%
Min(A)
30%
Max(A)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

Neighbourhood cause
Importance Function

No
10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
No
Yes
30%+30%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Yes
Yes
60%+40%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Co-site
Adjacent
No
Where:
426

AT332_TRR_E0
overlapping.
causes.
5.6.3.2 Importance of Inter-carrier Neighbours

Neighbourhood cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
Co-site cell

relationship
If the Force neighbour symmetry option is selected
transmitter.
d
Di = 1 ----------d max
on page 428.
[Neighbours]

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%
427

cause
IF
Resulting IF using the default

values from the table above
No
10%+50%{10%(Di)+90%(O)}
Yes
60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.
5.6.4 Appendices
5.6.4.1 Calculation of the Inter-Transmitter Distance

5.7 PN Offset Allocation

PN offset is used to identify a cell. It is a time offset used by a cell to shift a Pseudo Noise sequence. Mobile processes the
strongest received PN sequence and reads its phase that identifies the cell. There are a maximum of 512 PN offsets numbered
from 0 to 511.
The cells to which Atoll allocates PN offsets are referred to as the TBA cells (cells to be allocated). TBA cells fulfil following
conditions:
428
They are active,

They belong to the folder on which allocation has been executed. This folder can be either the Transmitters folder
or a group of transmitters or a single transmitter.

AT332_TRR_E0
zone.

5.7.1.1 Options and Constraints
The PN offset allocation algorithm can take into account following constraints and options:
PILOT_INC parameter,
Atoll uses this parameter to determine the pool of possible PN offsets (512 divided by PILOT_INC value). The first PN
offset is PILOT_INC and other ones are multiples of this value.
For example: When PILOT_INC is set to 4, the pool of possible PN offsets consists of PN offsets from 4 to 508 with a
separation interval of 4 (i.e. [4,8,12,16,...508]).
Neighbourhood between cells,
You may consider:
In the context of the PN offset allocation, the term "neighbours" refers to intra-carrier
neighbours.
In 3GPP2 multi-RAT documents, Atoll also tries to allocate different PN offsets to
CDMA cells that are neighbours of a common LTE cell.
Cells fulfilling a criterion on Ec/I0 (option Additional Ec/I0 conditions),
Atoll reuses the intra-carrier neighbour allocation algorithm to determine the list of cells which cannot be allocated the same
scrambling code, and to calculate their importance.
For a reference cell A, Atoll considers all the cells B that can enter the active set on the area where the reference cell is
the best server (area where (Ec/I0)A exceeds Min. Ec/I0 and is the highest one and (Ec/I0)B exceeds T_Drop).
Atoll considers either a percentage of the cell maximum powers or the total downlink
power used by the cells in order to evaluate I0. In this case, I0 equals the sum of total
transmitted powers. When this parameter is not specified in the cell properties, Atoll
uses 50% of the maximum power.
Co-PN Reuse distance,
Reuse distance is a constraint on the allocation of PN offsets. A PN offset cannot be reused at a site that is not at least as far
away as the reuse distance from the site allocated with the particular PN offset.
PN offset reuse distance can be defined at cell level. If this value is not defined, then Atoll
will use the default reuse distance defined in the PN offset Automatic Allocation dialog
box.
PN-cluster size. Within the context of PN offset allocation, the term "PN-cluster" refers to a sub-group of PN offsets.
Exceptional pairs,
Domains of PN offsets,
When no domain is assigned to cells, Atoll considers the PILOT_INC parameter only to
determine available PN offsets (e.g., If PILOT_INC is set to 4, all PN offsets from 4 to 508
with a separation interval of 4 can be allocated).
429

The carrier on which the allocation is run: It can be a given carrier or all of them. In this case, either Atoll independently
plans PN offsets for the different carriers, or it allocates the same PN offset to each carrier of a transmitter if the option
"Allocate carriers identically" is selected.
The possibility to use a maximum of PN offsets (option "Use a Maximum of PN Offsets"): Atoll will try to spread the
PN offset spectrum the most.
The "Delete All Codes" option: When selecting this option, Atoll deletes all the current PN offsets and carries out a
new PN offset allocation. If not selected, the existing PN offsets are kept.
In addition, it depends on the selected allocation strategy. Allocation strategies can be:
PN offset per cell: The purpose of this strategy is to reduce the spectrum of allocated PN offsets the maximum
possible. Atoll will allocate the first possible PN offsets in the domain.
Adjacent PN-Clusters per site: This strategy consists of allocating one cluster of adjacent PN offsets to each site, then,
one PN offset of the cluster to each cell of each transmitter according to its azimuth. When all the clusters have been
allocated and there are still sites remaining to be allocated, Atoll reuses the clusters at another site.
Distributed PN-clusters per site: This strategy consists of allocating one cluster of PN offsets to each site in the
network, then, one PN offset of the cluster to each cell of each transmitter according to its azimuth. With this strategy,
the cluster is made of PN offsets separated as much as possible. When all the clusters have been allocated and there
are still sites remaining to be allocated, Atoll reuses the clusters at another site.
In the Results table, Atoll only displays PN offsets allocated to TBA cells.

TBA cell may be:
"First Order"),
The cells that fulfil Ec/I0 condition (option Additional Ec/I0 conditions),
One additional constraint is considered in 3GPP2 multi-RAT documents:
The cell and its near cells are neighbours of the same LTE cell.
and the cost of the PN offset plan. During the allocation, Atoll tries to assign different PN offsets to the TBA cell and its near
cells. If it respects all the constraints, the cost of the PN offset plan is 0. When a cell has too many constraints and there are
not anymore PN offsets available, Atoll breaks the constraint with the lowest cost so as to generate the PN offset plan with
the lowest cost. For information on the cost generated by each constraint, see "Cell Priority" on page 431.
5.7.1.2.1

Strategy: PN offset per cell
Atoll processes TBA cells according to their priority. It allocates PN offsets starting with the highest priority cell and its near
cells, and continuing with the lowest priority cells not allocated yet and their near cells. For information on calculating cell
priority, see "Cell Priority" on page 431.
Strategy: Adjacent PN-Clusters per site
Atoll assigns a PN-cluster of adjacent PN offsets to each site, starting with the highest priority site and its near sites, and
continuing with the lowest priority sites not allocated yet and their near sites. When all the clusters have been allocated and
there are still sites remaining to be allocated, Atoll reuses the clusters at another site. When the Co-PN Reuse Distance option
is selected, the algorithm reuses the clusters as soon as the Co-PN reuse distance is exceeded. Otherwise, when the option is
not selected, the algorithm tries to assign reused clusters as spaced out as possible.
Then, Atoll allocates a PN offset from the cluster to each cell of each transmitter located on the sites according to the
transmitter azimuth. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not
allocated yet and their near cells.
430

AT332_TRR_E0
Strategy: Distributed PN-Clusters per site

Atoll assigns one cluster to each site, starting with the highest priority site and its near sites, and continuing with the lowest
priority sites not allocated yet and their near sites. When all the clusters have been allocated and there are still sites remaining
to be allocated, Atoll reuses the clusters at another site. When the Co-PN Reuse Distance option is selected, the algorithm
reuses the clusters as soon as the Co-PN reuse distance is exceeded. Otherwise, when the option is not selected, the algorithm
tries to assign reused clusters as spaced out as possible.
Then, Atoll assigns a PN offset from the cluster to each cell of each transmitter located on the sites according to the
transmitter azimuth. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not
allocated yet and their near cells.
5.7.1.2.2
In case you have a multi-carrier network and you run the PN offset allocation on all the carriers, the allocation process
depends on wether the option "Allocate Carriers Identically" is selected or not.
When the option is not selected, algorithm works for each strategy, as explained above. On the other hand, when the option
is selected, allocation order changes. It is no longer based on the cell priority but depends on the transmitter priority. All
transmitters which have constraints with the studied transmitter will be referred to as near transmitters.
In case of a "Per cell" strategy (PN offset per cell), Atoll starts PN offset allocation with the highest priority transmitter and its
near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. The same
PN offset is assigned to each cell of the transmitter.
In case of a "Per site" strategy (Adjacent and Distributed PN-clusters per site strategies), Atoll assigns a cluster to each site
and then, allocates a PN offset to each transmitter. It starts with the highest priority transmitter and its near transmitters and
continues with the lowest priority transmitters not allocated yet and their near transmitters. The same PN offset is assigned
to each cell of the transmitter.
For information on calculating cell priority, see "Cell Priority" on page 431. For information on calculating transmitter priority,
see "Transmitter Priority" on page 433.
alphanumeric order.
5.7.1.2.3
Difference between Adjacent and Distributed PN-Clusters

The following example explains the difference between "Adjacent PN-clusters" and "Distributed PN-clusters". The PILOT_INC
has been set to 4 and the PN-cluster size to 3. There are:
128 PN offsets that can be allocated: they are all PN offsets from 4 to 508 with a separation interval of 4.
Each PN-cluster consists of three PN offsets. So, there are 42 PN-clusters available.
If you select "Adjacent PN-cluster per site" as allocation strategy, Atoll will consider PN-clusters consisted of adjacent PN
offsets (e.g. {4,8,12}, {16,20,24}, ...,{496,500,504}).
If you select "Distributed PN-cluster per site" as allocation strategy, Atoll will consider PN-clusters consisted of PN offsets
separated as much as possible (e.g. {4,172,340}, {8,176,344}, ...,{168,336,504}).

5.7.1.3.1
Cell Priority
PN offset allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities assigned to
cells depend upon how much constrained each cell is and the cost defined for each constraint. A cell without any constraint
has a default cost, C , equal to 0. The higher the cost on a cell, the higher the priority it has for the PN offset allocation process.
There are five criteria employed to determine the cell priority:
PN Offset Domain Criterion
The cost due to the domain constraint, C i Dom , depends on the number of PN offsets available for the allocation. The
domain constraint is mandatory and cannot be broken.
When no domain is assigned to cells, 512 PN offsets are available and we have:
C i Dom = 0
431

When domains of PN offsets are assigned to cells, each unavailable PN offset generates a cost. The higher the number of codes
available in the domain, the less will be the cost due to this criterion. The cost is given as:
C i Dom = 512 Number of PN Offsets in the domain
Distance Criterion
Cj Dist i
C i Dist =
Where
C i EP =
cEP i j
j
Where

Ci N =
j

C j N1 i = I j c N1
Where
Because two first order neighbours must not have the same PN offset, Atoll considers the cost created by two first order
neighbours to be each other.
432

AT332_TRR_E0
C j N1 i + C j N1 i
C j j N1 i = ----------------------------------------------------2
Where
Because two second order neighbours must not have the same PN offset, Atoll considers the cost created by two second order
C k N2 i + C k N2 i
C k k N2 i = ------------------------------------------------------2
C l N3 i = Max j
c N3
Where
Because two third order neighbours must not have the same PN offset, Atoll considers the cost created by two third order
C l N3 i + C l N3 i
C l l N3 i = ----------------------------------------------------2
.
And
LTE Neighbour Criterion
This criterion is considered in 3GPP2 multi-RAT documents. If the cell i is neighbour of an LTE cell, the cell constraint level
depends on how many cells j are neighbours of the same LTE cell. The total cost due to LTE neighbour constraint is given as:
C i N LTE =
cNLTE j TxLTE
j
Where
cN
LTE
is the cost of the LTE neighbour constraint. This value can be defined in the Constraint Cost dialog box.
Therefore, the total cost due to constraints on any cell i is defined as:
With
C i U = C i Dist + C i EP + C i N + C i N 2G
5.7.1.3.2
In case you have a multi-carrier network and you run PN offset allocation on "all" the carriers with the option "allocate carriers
identically", algorithm in Atoll allots priorities to transmitters. Priorities assigned to transmitters depend on how much
constrained each transmitter is and the cost defined for each constraint. The higher the cost on a transmitter, the higher the
priority it has for the PN offset allocation process.
433

Max C U and C Dom = 512 Number of PN offsets in the domain
Tx
i Tx i
With C Tx U =
5.7.1.3.3
Site Priority
In case of "Per Site" allocation strategies (Adjacent PN-clusters per site and Distributed PN-clusters per site), algorithm in Atoll
allots priorities to sites. Priorities assigned to sites depend on how much constrained each site is. The higher the constraint on
a site, the higher the priority it has for the PN offset allocation process.
Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. The site constraint is given as:
With, C S U =
CTx U , and CS Dom
= 512 Number of PN offsets in the domain
Tx
Here, the domain considered for the site is the intersection of domains available for transmitters of the site.
5.7.2 Allocation Examples

In order to understand the differences between the different allocation strategies and the behaviour of the algorithm when
using a maximum of PN offsets or not, let us consider the following sample scenario:
Figure 5.8: PN Offset Allocation

Let Site0, Site1, Site2 and Site3 be four sites with 3 cells using carrier 0 whom PN offsets have to be allocated. The PILOT_INC
parameter has been set to 4 and the PN Cluster Size is 3. Therefore, all PN offsets from 4 to 508 with a separation interval of
4 can be allocated. The reuse distance is supposed to be lower than the inter-site distance. Only co-site neighbours exist and
all of them have the same importance.
The following section lists the results of each combination of options with explanation where necessary.
5.7.2.1 Strategy: PN Offset per Cell

Since the restrictions of neighbourhood only apply to co-sites with the same importance and inter-site distances are greater
than reuse distances, every cell has the same priority. Then, the PN offset allocation to cells is performed in an alphanumeric
order.
434

AT332_TRR_E0
Without Use a Maximum of PN Offsets
With Use a Maximum of PN Offsets
Atoll allocates the first three PN offsets in the domain (4, 8

and 12) to the Site0s cells. Under given constraints of
neighbourhood and reuse distance, same PN offsets can be
allocated to each sites cells.
Atoll allocates the first three PN offsets in the domain (4, 8

and 12) to the Site0s cells. As it is allowed to use a maximum
of PN offsets, Atoll allocates different PN offsets to each
sites cells so that there is least repetition.
5.7.2.2 Strategy: Adjacent PN-Clusters Per Site

order.
Atoll allocates a PN cluster of adjacent PN offsets to Site0 and As it is possible to use a maximum of PN offsets, Atoll
then, one PN offset of the PN cluster to each cell. Under given allocates different PN clusters of adjacent PN offsets to sites
constraints of neighbourhood and reuse distance, the same so that there is least repetition of PN offsets.
PN cluster can be allocated to each site and same PN offsets
to each sites cells.
5.7.2.3 Strategy: Distributed PN-Clusters Per Site

order.
435

Atoll allocates a PN cluster of distributed PN offsets to Site0 As it is possible to use a maximum of PN offsets, Atoll
and then, one PN offset of the PN cluster to each cell. Under allocates different PN clusters of distributed PN offsets to
given constraints of neighbourhood and reuse distance, the sites so that there is least repetition of PN offsets.
same PN cluster can be allocated to each site and same PN
offsets to each sites cells.
5.8 Automatic GSM-CDMA Neighbour Allocation

5.8.1 Overview
You can automatically calculate and allocate neighbours between GSM/TDMA and CDMA2000 networks. In Atoll, it is called
inter-technology neighbour allocation.
When the CDMA coverage is not continuous. In this case, the CDMA coverage is extended by CDMA-GSM handover
into the GSM network,
And in order to balance traffic and service distribution between both networks.
Note that the automatic inter-technology neighbour allocation algorithm takes into account both cases.
An .atl document containing the GSM/TDMA network, GSM.atl, and another one containing the CDMA2000 network,
CDMA.atl,
An existing link on the Transmitters folder of GSM.atl into CDMA.atl.
transmitters of GSM.atl are potential neighbours. The cells to be allocated will be called TBA cells which, being cells of
CDMA.atl, fulfill following conditions:
They are active,

They satisfy the filter criteria applied to Transmitters folder,
a group of transmitters subfolder.
Only CDMA TBA cells may be assigned neighbours.

The inter-transmitter distance,

The maximum number of neighbours fixed,
Allocation options,
The selected allocation strategy,
We assume we have a CDMA reference cell, A, and a GSM candidate neighbour, transmitter B.
436

AT332_TRR_E0

When the automatic allocation starts, Atoll checks the following conditions:
The distance between the CDMA reference cell and the GSM neighbour must be less than the user-definable
maximum inter-site distance. If the distance between the CDMA reference cell and the GSM neighbour is greater than
this value, then the candidate neighbour is discarded.
azimuths of antennas. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or
more carriers. Atoll will allocate neighbours to cells using the selected carriers.
reference CDMA cell in the candidate neighbour list. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference CDMA cell.
allocated to each cell is exceeded. If we consider the case for which there are 15 candidate neighbours and the maximum
number of neighbours to be allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8 (having the
highest importance values) will be allocated to the reference cell. Note that the maximum number of inter-technology
neighbours can be defined at the cell level (property dialog box or cell table). If defined there, this value is taken into account
Neighbourhood cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
Co-site transmitter
100 %

d1 ---------d max
Where d is the effective distance between the CDMA reference cell and the GSM neighbour and d max is the maximum intersite distance. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter Distance" on
page 428.
as existing.

The distance between the CDMA reference cell and the GSM neighbour must be less than the user-definable
maximum inter-site distance. If the distance between the CDMA reference cell and the GSM neighbour is greater than
this value, then the candidate neighbour is discarded.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or
more carriers. Atoll will allocate neighbours to cells using the selected carriers.
reference CDMA cell in the candidate neighbour list. This option is automatically selected.
437

force/forbid a GSM transmitter to be candidate neighbour of the reference CDMA cell.
There must be an overlapping zone ( S A S B ) with a given cell edge coverage probability where:
Four different cases may be considered for SA:
1st case: SA is the area where the cell A is the best serving cell of the CDMA network.
The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0) and is the highest one.
In this case, the Ec/I0 margin must be equal to 0dB and the max Ec/I0 option disabled.
2nd case: SA represents the area where the pilot quality from the cell A strats decreasing but the cell A is still the
best serving cell of the CDMA network.
The Ec/I0 margin must be equal to 0dB, the max Ec/I0 option selected and a maximum Ec/I0 user-defined.
The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/I0.
The pilot quality from A is the highest one.
3rd case: SA represents the area where the cell A is not the best serving cell but can enter the active set.
Here, the Ec/I0 margin has to be different from 0dB and the max Ec/I0 option disabled.
The pilot quality from A is within a margin from the best Ec/I0, where the best Ec/I0 exceeds the minimum Ec/
I0.
4th case: SA represents the area where:
The pilot quality from A is within a margin from the best Ec/I0 (where the best Ec/I0 exceeds the minimum Ec/
I0) and lower than the maximum Ec/I0.
In this case, the margin must be different from 0dB, the max Ec/I0 option selected and a maximum Ec/I0 userdefined.
1st case: SB is the area where the cell B is the best serving cell of the GSM network.
In this case, the margin must be set to 0dB.
highest one.
nd
2 case: The margin is different from 0dB and SB is the area where:
SA SB
SA
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of covered
area.
When the automatic allocation is based on coverage overlapping, we recommend you to
perform two successive automatic allocations:
A first allocation in order to find handovers due to non-continuous CDMA coverage.

In this case, you have to select the max Ec/I0 option and define a high enough value.
A second allocation in order to complete the previous list with handovers motivated
for reasons of traffic and service distribution. Here, the max Ec/I0 option must be
disabled.
Importance values are used by the allocation algorithm to rank the neighbours according to the distance and the allocation
reason. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list
438

AT332_TRR_E0
if the maximum number of neighbours to be allocated to each cell is exceeded. If we consider the case for which there are 15
candidate neighbours, only 8 (having the highest importance values) will be allocated to the reference cell. Note that the
maximum number of inter-technology neighbours can be defined at the cell level (property dialog box or cell table). If defined
there, this value is taken into account instead of the default one available in the Neighbour Allocation dialog box.
As indicated in the table below, the neighbour importance depends on the neighbourhood cause; this value varies between
0 to 100%.
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
Co-site transmitter
IF

IF
transmitter.
d Di = 1 ---------d max
d is the distance between the CDMA reference cell and the GSM neighbour.
d max is the maximum inter-site distance.

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

cause
IF

No
10%+50%{10%(Di)+90%(O)}
Yes
60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.
439


5.8.2.3 Delete Existing Neighbours Option

checked. We assume that we have an existing allocation of inter-technology neighbours.
A new TBA cell i is created in CDMA.atl. Therefore, if you start a new allocation without selecting the Delete existing
neighbours option, Atoll determines the neighbour list of the cell i,
If you change some allocation criteria (e.g. increase the maximum number of neighbours or create a new GSM TBC
transmitter) and start a new allocation without selecting the Delete existing neighbours option, it examines the neighbour list
of TBA cells and checks allocation criteria if there is space in their neighbour lists. A new GSM TBC transmitter can enter the
TBA cell neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list.
440
Chapter 6
LTE Networks
"Calculation Quick Reference" on page 449
"Available Calculations" on page 470
"Calculation Details" on page 485
"Automatic Planning Algorithms" on page 566

442

Chapter 6: LTE Networks
AT332_TRR_E0
6 LTE Networks
This chapter describes all the calculations performed in Atoll LTE documents. The first part of this chapter lists all the input
parameters in the LTE documents, their significance, location in the Atoll GUI, and their usage. It also contains the lists of the
formulas used for the calculations.
The second part describes all the calculation processes, i.e., signal level coverage predictions, point analysis calculations,
signal quality coverage predictions, calculations on subscriber lists, and Monte Carlo simulations. The calculation algorithms
used by these calculation processes are available in the next part.
The third part describes all the calculation algorithms used in all the calculations. These algorithms include the calculation of
signal levels, noise, and interference for downlink and uplink considering power control, MIMO, smart antennas, and the radio
resource management algorithms used by the different available schedulers.
If you are new to LTE, you can also see the Glossary of LTE Terms in the User Manual for information on LTE terms and
concepts, especially in the context of their user in Atoll.
A cell refers to a transmitter-carrier (TX-c) pair. The cell being studied during a
calculation is referred to as TXi(ic) in this chapter.
All the calculation algorithms in this section are described for two types of cells:
A studied cell (represented by the subscript "i") comprising the studied transmitter
TXi and its carrier ic. It is the cell which is currently the focus of the calculation. For
example, a victim cell when calculating the interference it is receiving from other
cells.
Other cells (represented by the subscript "j") comprising the other transmitter TXj
and its carrier jc. The other cells in the network can be interfering cells (downlink)
or the serving cells of interfering mobiles (uplink).
All the calculation algorithms in this section are described for two types of receivers:
Mi: A pixel (coverage predictions), subscriber (calculations on subscriber lists), or

mobile (Monte Carlo simulations) covered/served by the studied cell TXi(ic).
Mj: A mobile (Monte Carlo simulations) covered/served by any other cell TXj(jc).
6.1 Definitions
This table lists the input to calculations, coverage predictions, and simulations.
Name
Value
Unit
Description
D Frame
3GPP parameter (Fixed to 10 ms in Atoll)
ms
Frame duration
W FB
3GPP parameter (Fixed to 180 kHz in Atoll)
kHz
Width of a resource/frequency block
3GPP parameter (Fixed to 15 kHz in Atoll)
kHz
Subcarrier width
N FB SS PBCH
3GPP parameter (Fixed to 6 in Atoll)
None
Number of frequency blocks for SS

and PBCH transmission
N SF Frame
None
Number of subframes per frame
N Slots SF
None
Number of slots per subframe
1.38 x 10-23
J/K
Boltzmanns constant
290
Ambient temperature
n0
Calculation result ( 10 Log K T 1000 = 174 dBm/Hz )
dBm/Hz
Power spectral density of thermal

noise
D CP
Frame configuration or, otherwise, global parameter
None
Cyclic prefix duration
N SD PDCCH
SD
Number of PDCCH symbol durations

per subframe
N FB PUCCH
RB
Average number of PUCCH

frequency blocks per frame
443

Name
Value
Unit
Description
M PC
Global parameter
dB
Uplink power control adjustment

margin
CNR Min
Global parametera
dB
Minimum signal to thermal noise

threshold (interferer cutoff)
W Channel
MHz
Channel bandwidth
First
None
First channel number of the

frequency band
N Channel
Last
None
Last channel number of the

frequency band
F Start TDD
MHz
Start frequency of the TDD frequency

band
F Start FDD DL
MHz
DL start frequency of the FDD

frequency band
F Start FDD UL
MHz
UL start frequency of the FDD

frequency band
F Sampling
MHz
Sampling frequency
f ACS
dB
Adjacent channel suppression factor
N FB
None
Number of frequency blocks per

channel bandwidth
ICS Band
MHz
Inter-channel spacing
CN Band
None
Channel number step
N SCa Total
F Sampling
-)
Calculation result ( N SCa Total = -------------------F
None
Total number of subcarriers
N SCa Used
N FB W FB
-)
Calculation result ( N SCa Used = -----------------------F
None
Number of used subcarriers
N SCa DC
Hard-coded parameter ( N SCa DC = 1 )
None
Number of DC subcarriers
N SCa Guard
Calculation result
( N SCa Guard = N SCa Total N SCa Used N SCa DC )
None
Number of guard subcarriers
i Layer
Layer parameter
None
Layer index
p Layer
Layer parameter
None
Layer priority
V Layer
Max
Layer parameter
km/h
Maximum mobile speed supported

by a layer
CE
Frame configuration parameter
dB
Cell-edge power boost
N FB CE0
None
Number of cell-edge frequency

blocks for PSS ID 0
N FB CE1
None

blocks for PSS ID 1
N FB CE2
None

blocks for PSS ID 2
Bearer parameter
None
Bearer index
Mod B
Bearer parameter
None
Modulation used by the bearer
CR B
Bearer parameter
None
Coding rate of the bearer
Bearer parameter
bits/
symbol
Bearer efficiency
TB
Bearer parameter
dB
Bearer selection threshold
Site
Site parameter
kbps
Maximum S1 interface site downlink

throughput
N Channel
TP S1 DL
444

AT332_TRR_E0
Name
Value
Unit
Description
Site
Site parameter
kbps
Maximum S1 interface site uplink

throughput

dB
Transmitter noise figure
N Ant TX
None
Number of antenna ports used for

transmission
N Ant RX
None
Number of antenna ports used for

reception
TX
Transmitter antenna parameter
dB
Antenna gain
TX

dB
Transmitter loss
TX
Smart antenna parameter
None
Array
dB
Array gain offset
Combining
dB
Power combining gain offset
G SA
dB
Diversity gain (cross-polarisation)
N Channel
Cell parameter
None
Cells channel number
ID
Cell parameter
None
Cells physical ID
ID SSS
ID
Cell parameter: Floor --------
3
None
Cells SSS ID (one of 168 pseudorandom sequences)
ID PSS
Cell parameter: ID Mod 3
None
Cells PSS ID (one of 3 cyclic shifts of

the sequence given by the SSS ID)
Shift
Cell parameter: ID Mod 6
None
Cells v shift (also known as the

reference signal hopping index)
P Max
Cell parameter
dBm
Maximum cell transmission power
EPRE DLRS
Cell parameter
dBm
Energy per resource element for the

downlink reference signals
(User-defined or calculated)
EPRE SS
Cell parameter
dB
Energy per resource element offset

for the SS with respect to the
downlink reference signal EPRE
EPRE PBCH
Cell parameter
dB

for the PBCH with respect to the
EPRE PDCCH
Cell parameter
dB

for the PDCCH with respect to the
EPRE PDSCH
Cell parameter
dB

for the PDSCH with respect to the
T RSRP
Cell parameter
dB
Minimum Required RSRP
TX i ic
Cell parameter
dB
Cell selection threshold
TX ic
i
Cell parameter
dB
Cell individual offset
Cell parameter
dB
Handover margin
TL DL
Cell parameter
Downlink traffic load
r DL CE
Cell parameter
Downlink cell-edge traffic ratio
TP S1 UL
nf
G
L
TX
E SA
G SA
G SA
Div
T Selection
O Individual
TX i ic
M HO
445

Name
Value
Unit
Description
TL UL
Cell parameter
Uplink traffic load
TL DL Max
Cell parameter
Maximum downlink traffic load
TL UL Max
Cell parameter
Maximum uplink traffic load
NR UL
Cell parameter
dB
Uplink noise rise
NRUL ICIC
Cell parameter
dB
ICIC uplink noise rise
NR UL Max
Cell parameter
dB
Maximum uplink noise rise
N Users Max
Cell parameter
None
Maximum number of users per cell
N Users DL
Cell parameter
None
Number of users connected to the

cell in downlink
N Users UL
Cell parameter
None

cell in uplink
L Path
Cell parameter
dB
Delta path loss threshold
E DLRS
Cell parameter
dB
Delta RSRP threshold
N SF DL
Cell parameter
None
Number of downlink subframes per

frame
N SF UL
Cell parameter
None
Number of uplink subframes per

frame
N TDD SSF
TX i ic
Cell parameter
None
Number of TDD special subframes

per frame
D Reuse
Cell parameter
Channel and physical cell ID reuse

distance
G MU MIMO DL
Cell parameter
None
Average number of co-scheduled

MU-MIMO users in downlink
G MU MIMO UL
Cell parameter
None
Average number of co-scheduled

MU-MIMO users in uplink
FPC
Cell parameter
None
Fractional power control factor
CINR PUSCH Max
Cell parameter
dB
Maximum PUSCH C/(I+N)
Inter Tech
Cell parameter
dB
Inter-technology downlink noise rise
Inter Tech
Cell parameter
dB
AU DL
Cell parameter
Downlink AAS usage ratio
TX i ic
Proportional Fair scheduler parameter
None
Downlink multi-user diversity gain

(MUG)
TX ic
i
None
Uplink multi-user diversity gain

(MUG)
CINR MUG
dB
Maximum C/(I+N) above which no

MUG gain is applied
T SU MIMO UL
Cell reception equipment parameter
dB
Uplink SU-MIMO threshold

Empty = -200 dB
T MU MIMO UL
dB
Uplink MU-MIMO threshold

Empty = -200 dB
G SU MIMO UL
None
Maximum uplink SU-MIMO gain
G Div UL
dB
Receive, SU-MIMO, or MU-MIMO

diversity gain
UL
dB
Uplink secondary cell activation

threshold
NR DL
NR UL
G MUG DL
G MUG UL
Max
Max
T SCell
446

AT332_TRR_E0
Name
Value
Unit
Description
Max DL
CoMP set parameter
None
Transmission set size: Maximum

number of CoMP cells that can be
coordinated in the downlink
Max UL
CoMP set parameter
None
Reception set size: Maximum

number of CoMP cells that can be
coordinated in the uplink
DL
CoMP set parameter
None
Collision probability for downlink

coordinated scheduling CoMP as a
function of downlink traffic load
P CoMP
UL
CoMP set parameter
None
Collision probability for uplink

coordinated scheduling CoMP as a
function of downlink traffic load
QCI
Service parameter
None
QoS class identifier (QCI) of the

service
p QCI
Service parameter (automatically determined from the QCI)
None
Services QCI priority
p Service
Service parameter
None
User-defined service priority
B DL Highest
Service parameter
None
Highest bearer used by a service in

the downlink
B UL Highest
Service parameter
None

the uplink
B DL Lowest
Service parameter
None
Lowest bearer used by a service in

the downlink
B UL Lowest
Service parameter
None

the uplink
f Act
UL
Service parameter
Uplink activity factor
f Act
DL
Service parameter
Downlink activity factor
TPD Min UL
Service parameter
kbps
Minimum throughput demand in the

uplink (Guaranteed Bit Rate, GBR)
TPD Min DL
Service parameter
kbps

downlink (Guaranteed Bit Rate, GBR)
TPD Max UL
Service parameter
kbps
Maximum throughput demand in the

uplink (Maximum Bit Rate, MBR)
TPD Max DL
Service parameter
kbps

downlink (Maximum Bit Rate, MBR)
UL
Service parameter
kbps
Average requested throughput in the

uplink
TP Average
DL
Service parameter
kbps

downlink
TP Offset
Service parameter
kbps
Throughput offset
f TP Scaling
Service parameter
Scaling factor
L Body
Service parameter
dB
Body loss
N FB UL
Min
Service parameter
None
Minimum number of frequency

blocks
P Min
Terminal parameter
dBm
Minimum terminal power
P Max
Terminal parameter
dBm
Maximum terminal power
nf
Terminal parameter
dB
Terminal noise figure
Terminal parameter
dB
Terminal antenna gain
Terminal parameter
dB
Terminal loss
N CoMP
N CoMP
P CoMP
TP Average
447

Name
Value
Unit
Description
N Ant TX
Terminal parameter
None
Number of antenna ports for

transmission
N Ant RX
Terminal parameter
None
Number of antenna ports for

reception
Max DL
Terminal parameter
None
Maximum number of downlink

secondary cells
Max UL
Terminal parameter
None
Maximum number of uplink

secondary cells
Max DL
UE category parameter
Bits
Maximum number of transport block

bits per TTI (subframe) in downlink
Max UL
Bits
Maximum number of transport block

bits per TTI (subframe) in uplink
None
Highest modulation supported in

uplink
N Ant UE
None
Maximum number of reception

antenna ports supported in downlink
T SU MIMO DL
Terminal reception equipment parameter
dB
Downlink SU-MIMO threshold

Empty = -200 dB
G SU MIMO DL
None
Maximum downlink SU-MIMO gain
T MU MIMO DL
dB
Downlink MU-MIMO threshold

Empty = -200 dB
G Div PBCH
dB
PBCH diversity gain
G Div PDCCH
dB
PDCCH diversity gain
G Div DL
dB
Transmit, SU-MIMO, or MU-MIMO

diversity gain
T SCell
DL
dB
Downlink secondary cell activation

threshold
DL
dB
Downlink AAS threshold

Empty = +200 dB
UL
Clutter parameter
dB
Additional uplink diversity gain
G Div
DL
Clutter parameter
dB
Additional downlink diversity gain
f SU MIMO
Clutter parameter
None
SU-MIMO gain factor
L Indoor
Clutter parameter
dB
Indoor loss
L Path
dB
Path loss
N SCell
N SCell
N TBB TTI
N TBB TTI
Max UL
Mod UE
Max DL
Max
T AAS
G Div
ICPDL
Network parameter
None
Inter-technology downlink channel

protection ratio for a frequency
offset F between the interfered
and interfering frequency channels
M Shadowing Model
Monte Carlo simulations: Random result calculated from model

standard deviation
Coverage Predictions: Result calculated from cell edge coverage
probability and model standard deviation
dB
Model shadowing margin
M Shadowing C I

probability and C/I standard deviation
dB
C/I shadowing margin
Coverage Predictions: Gain due to the increased probability of

finding a server in downlink dynamic point selection and coherent
joint transmission CoMP
dB
Downlink macro-diversity gain
MD
G CoMP
a.
448
Any interfering cell whose signal to thermal noise ratio is less than CNR Min will be discarded.

AT332_TRR_E0
6.2 Calculation Quick Reference

The following tables list the formulas used in calculations.
6.2.1 Downlink Transmission Powers Calculation

Name
Value
Unit
Description
N Sym SRB
N SCa FB N SD Slot N Slot SF
None
Number of symbols per scheduler

resource block
N Sym SSF
DwPTS
N SCa FB N SD SSF
None
Number of DwPTS modulation

symbols per scheduler resource
block in the TDD special subframes
N SCa FB
W FB
--------F
None
Number of subcarriers per

frequency block
None
Total number of symbols in

downlink
None
Number of symbols reserved for

downlink reference signals in one
scheduler resource block
See "Downlink Transmission Power Calculation" on page 485
None

downlink reference signals in
DwPTS of one TDD special
subframe
TX i ic
None

frame
None
Number of symbols for downlink

reference signals in one scheduler
resource block
None

reference signals in DwPTS of one
TDD special subframe
TX i ic
None

reference signals in one frame
Where N Sym PSS = 2 N FB SS PBCH N SCa FB = 144
None
Number of symbols for the PSS and

the SSS
None
Number of symbols for the PBCH
TX i ic
N Sym DL
DwPTS
TX i ic
N FB
TX i ic
16
24
TX ic
i
N Res SRB
TX i ic
N Res DwPTS
TX i ic
N Sym Res
TX i ic
N SF DL N FB
N DLRS DwPTS
TX i ic
N Sym DLRS
TX i ic
DwPTS
N TDD SSF N Sym SSF
TX ic
i
if N Ant TX = 1
TX ic
i
if N Ant TX = 2
TX ic
i
if N Ant TX = 4 or 8
TX i ic
TX i ic
TX i ic
N Res SRB + N TDD SSF N FB
TX ic
i
N DLRS SRB
TX i ic
TX i ic
N Sym SRB N SF DL + N FB
TX i ic
N Res DwPTS
TX ic
i
if N Ant TX = 1
TX ic
i
if N Ant TX = 2
TX ic
i
See "Downlink Transmission Power Calculation" on page 485

TX i ic
TX i ic
N SF DL N FB
TX i ic
TX i ic
TX i ic
N DLRS SRB + N TDD SSF N FB
N DLRS DwPTS
N Sym PSS + N Sym SSS = 288

N Sym SS
N Sym SSS = 2 N FB SS PBCH N SCa FB = 144

TX i ic
N Sym PBCH
Extended CP: 216

Normal CP: 240
449

Name
Value
Unit
Description
None
Number of symbols for the PDCCH
TX ic
i
if N SD PDCCH = 0 : 0
TX ic
TX ic
i
i
if N SD PDCCH = 1 AND N Ant TX 2 :
TX i ic
N Sym PDCCH
TX ic
TX ic
TX i ic
TX i ic
TX i ic
i
N i
N SCa FB 4 N FB
SD PDCCH
N SF DL
+ NSD PDCCH N SCa FB 4 N FB
TX i ic
N TDD SSF
Otherwise:
TX i ic
TX ic
TX ic
i
i
N
N SCa FB 2 Min 4 N Ant TX N FB
SD PDCCH
TX ic
TX ic
TX ic
i
i
i
+ Min 2 N SD PDCCH N SCa FB 2 Min 4 N Ant TX N FB
TX i ic
N Sym PDSCH
TX i ic
TX i ic
TX i ic
TX i ic
N SF DL
TX i ic
N TDD SSF
TX i ic
None
N Sym DL N Sym Res N Sym SS N Sym PBCH N Sym PDCCH
Number of symbols for the PDSCH
TX ic
P i
Max
-------------------
TX i ic
TX i ic
DwPTS
10
10 Log 10
N SD Slot N Slot SF N SF DL + N TDD SSF N SD SSF
TX ic
TX i ic
EPRE DLRS
TX ic
i
i
EPRE SS
EPRE PBCH
TX ic
--------------------------------------------------------------------i
10
10
dBm/Sym
+ N Sym PBCH 10
10 L og N Sym DLRS + N Sym SS 10
+ N Sym PDCCH 10
TX i ic
EPRE PDCCH
-----------------------------------10
+ N Sym PDSCH 10
TX i ic
EPRE PDSCH
------------------------------------
10
Energy per resource element for 1

modulation symbol (dBm/Sym) of
the downlink reference signals
With reference signal EPRE
calculation method is set to
Calculated (equal distribution of
unused EPRE)
TX ic
P i
Max
-------------------
TX i ic
TX i ic
DwPTS
10
10 Log 10
TX ic
TX i ic
EPRE DLRS
+ N Sym PDCCH 10
TX i ic
EPRE PDCCH
-----------------------------------10
+ N Sym PDSCH 10
TX i ic
dBm/Sym

the PBCH
TX i ic
TX i ic
dBm/Sym

the PDCCH
TX i ic
TX i ic
dBm/Sym

the PDSCH
TX i ic
TX i ic
TX i ic
TX i ic
EPRE PDSCH
450

the SS
TX ic
i
EPRE DLRS
EPRE PDCCH
TX i ic
EPRE PDSCH
------------------------------------
10
With reference signal EPRE

calculation method is set to
Calculated (with boost) or
Calculated (without boost)
dBm/Sym
TX ic
i
EPRE SS
EPRE PBCH
TX ic
i
i
EPRE SS
EPRE PBCH
TX ic
--------------------------------------------------------------------i
10
10
dBm/Sym
+ N Sym PBCH 10
10 L og N Sym Res + N Sym SS 10

the downlink reference signals
TX ic
i
EPRE SS
EPRE DLRS + EPRE PBCH
EPRE DLRS + EPRE PDCCH
EPRE DLRS + EPRE PDSCH

AT332_TRR_E0
Name
Value
Unit
Description
dbm/Sym
"Boosted" energy per resource

element for 1 modulation symbol
(dBm/Sym) of downlink reference
signals when the reference signal
EPRE calculation method is set to
Calculated (with boost)
dBm
Instantaneous transmission power

of the downlink reference signals
+ 10 Log N SCa FB N FB SS PBCH
dBm

of the SS
EPRE PBCH + 10 Log N SCa FB N FB SS PBCH
dBm

of the PBCH
TX i ic
N SymPDCCH
EPRE PDCCH + 10 Log ---------------------------------------------------------------------------------------------------------------------------------------dBm
TX
ic
TX
ic
TX
ic
TX
ic
i
i
i
i
N SD PDCCH N SF DL + Min 2 N SD PDCCH N TDD SSF
Average transmission power of the

PDCCH
TX ic
i
TX ic
i
EPRE DLRS + 10
EPRE DLRS
TX ic
TX i ic
TX i ic
TX i ic
P SS
EPRE SS
TX i ic
TX i ic
P PDSCH
TX i ic
P PBCH
P PDCCH
TX ic
i
i
EPRE DLRS + 10 Log 2 N FB
P DLRS
TX i ic
N TXi ic
Sym Res
Log ------------------------ TX i ic
N Sym DLRS
TX i ic
TX i ic
TX ic
N
i
Sym
PDSCH
EPRE PDSCH + 10 Log --------------------------------------------------------------------------------------------------------------------
TX i ic
TX i ic
N
N Slot SF N SD PDCCH N SF DL
SD Slot
TX i ic
TX i ic
DwPTS
+ N SD SSF Min 2 N SD PDCCH N TDD SSF
dBm
Average transmission power of the

PDSCH
6.2.2 Co- and Adjacent Channel Overlaps Calculation

Name
Value
Unit
Description
TX i ic
N TXi ic N First TXi ic

TX i ic
TX i ic
Channel
Channel
-
F Start Band + W Channel + ICS Band ------------------------------------------------------TX i ic
CN Band
MHz
Start frequency for the channel

number assigned to a cell
MHz
End frequency for the channel

MHz
Co-channel overlap bandwidth
None
Co-channel overlap ratio
MHz
Bandwidth of the lower-frequency

adjacent channel overlap
None
Lower-frequency adjacent channel

overlap ratio
MHz
Bandwidth of the higher-frequency

None
Higher-frequency adjacent channel

overlap ratio
F Start
TX i ic
TX i ic
TX i ic
TX ic TX jc
i
j
W CCO
TX jc
TX i ic TX j jc
L
TX ic TX jc
i
j
H
TX i ic TX j jc
r ACO
TX jc
TX ic
TX i ic TX j jc
W CCO
----------------------------------TX j jc
W Channel
TX jc
TX ic
TX jc
TX ic
TX ic
j
i
j
i
i
Min F End F Start Max F Start F Start W Channel
TX i ic TX j jc
W ACO
L
---------------------------------TX j jc
W Channel
TX ic TX jc
i
j
r ACO
L
W ACO
TX ic
j
i
j
i
Min F End F End Max F Start F Start
TX ic TX jc
i
j
r CCO
W ACO
TX i ic
F Start + W Channel
F End
TX j jc
TX i ic
Min F End F End
TX ic
TX jc
TX ic
i
j
i
+ W Channel Max F Start F End
TX i ic TX j jc
W ACO
H
---------------------------------TX j jc
W Channel
451

Name
Value
TX ic TX jc
i
j
TX ic TX jc
i
j
r ACO
r ACO
TX ic TX jc
i
j
rO
TX ic TX jc
i
j
r CCO
TX ic TX jc
i
j
+ r ACO
TX ic TX jc
i
j
+ r ACO
10
TX ic
i
f
ACS
----------------------10
Unit
Description
None
Adjacent channel overlap ratio
None
Total overlap ratio
6.2.3 Signal Level Calculation (DL)

The received signal levels (dBm) from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi as follows:
Name
TX i ic
C Max
Value
TX i ic
EIRP Max L Path M Shadowing Model L Indoor + G

L
Mi
Mi
TX i ic
TX i ic
C DLRS
TX i
TX
i
G Ant
TX
TX i ic
Mi
TX ic
i
TX i
TX i
TX i ic
EIRP1 SS
L
Mi
TX i
L Path M Shadowing Model L Indoor + G
Mi
C PBCH
TX i
+ G Ant L
TX
i
G Ant
TX
TX ic
i
Mi
TX i ic
C PDCCH
TX i
TX i
Mi
TX i ic
TX i
Received PDCCH signal level
dBm
PDCCH EIRP
TX i
With smart antennas:

TX i
P PDCCH + G Ant L
452
dBm
Mi
Mi
TX i ic
TX i ic
PBCH EIRP
L Ant L Body + f CP
Without smart antennas: P PDCCH + G Ant L

EIRP1 PDCCH
dBm
TX
i
Combining
Div
+ 10 Log E SA + G SA
+ G SA
TX i ic
Mi
Received PBCH signal level
TX i
EIRP1 PDCCH L Path M Shadowing Model L Indoor + G

L
dBm

TX i
P PBCH + G Ant L
TX i ic
Mi
TX i ic
TX i ic
SS EIRP
L Ant L Body + f CP
Without smart antennas: P PBCH + G Ant L

EIRP1 PBCH
dBm
TX
i
Combining
Div
+ G SA
EIRP1 PBCH L Path M Shadowing Model L Indoor + G

Mi
Received SS signal level
TX i
dBm
Mi
Mi
TX i ic
TX i ic
RS EIRP
L Ant L Body + f CP
TX i ic
TX ic
i
P SS
dBm
TX
i
Combining
Without smart antennas: P SS

EIRP1 SS
Received downlink reference signal

level
TX i

TX i ic
P DLRS + G Ant L
C SS
dBm
L Ant L Body + f CP
TX i ic
TX i ic
Downlink max EIRP
Mi
Mi
Without smart antennas: P DLRS + G Ant L

EIRP1 DLRS
dbm
TX
i
Combining
Div
+ G SA
EIRP1 DLRS L Path M Shadowing Model L Indoor + G

L
Received max cell power
TX i
Mi
dBm
Mi
Mi
TX i ic
TX ic
i
P Max
Description
L Ant L Body + f CP
Without smart antennas: P Max + G Ant L

EIRP Max
Unit
TX i
TX
i
Combining
Div
+ G SA

AT332_TRR_E0
Name
Value
TX ic
i
EIRP1 PDSCH L Path M Shadowing Model L Indoor + G
TX ic
i
C PDSCH
dBm
Received PDSCH signal level
dBm
PDSCH EIRP
dBm/Sym

energy per resource element
(RSRP)
dBm/Sym
RS EIRP
dBm/Sym
Received SS energy per resource

element
TX i ic
TX i ic
Description
L Ant L Body + f CP
TX i
Without smart antennas: P PDSCH + G Ant L

EIRP1 PDSCH
Unit
TX i

TX i ic
TX i
Array
P PDSCH + G SA + G SA
Combining
+ G SA
Div
+ G SA L
TX i
TX i ic
EIRP2 DLRS L Path M Shadowing Model
TX i ic
E DLRS
L Indoor + G
Mi
Mi
Mi
Mi
L Ant L Body + f CP
TX i ic
TX i
Without smart antennas: EPRE DLRS + G Ant L

TX i ic
TX i
EIRP2 DLRS
TX i ic
TX i
EPRE DLRS + G Ant L

TX i ic
EIRP2 SS
TX i ic
E SS
TX i
TX i
+ 10 Log E SA +
Combining
G SA
L Path M Shadowing Model
L Indoor + G
Mi
Mi
Mi
Mi
L Ant L Body + f CP
TX i ic
Without smart antennas: EPRE SS
TX i
+ G Ant L
TX i

TX ic
i
EIRP2 SS
Ser.:
TX ic
i
EPRE SS
TX
TX ic
i
Int.: EPRE SS
TX i ic
E PBCH
+ G Ant L
TX
TX
TX
i
Combining
Div
+ G SA dBm/Sym
+ G Ant L
TX
TX
i
Combining
TX ic
i
EIRP2 PBCH L Path M Shadowing Model

L Indoor + G
Mi
SS EIRP
Mi
Mi
Mi
dBm/Sym
Received PBCH energy per resource

element
dBm/Sym
PBCH EIRP
dBm/Sym
Received PDCCH energy per resource

element
dBm/Sym
PDCCH EIRP
dBm/Sym
Received PDSCH energy per resource

element
L Ant L Body + f CP
TX i ic
TX i
Without smart antennas: EPRE PBCH + G Ant L
TX i

TX i ic
EIRP2 PBCH
TX i ic
TX i
Ser.: EPRE PBCH + G Ant L

TX i ic
TX i
TX i
TX i
Int.: EPRE PBCH + G Ant L

TX i ic
E PDCCH
TX i
Combining
TX i
Div
+ G SA
Combining
TX i ic
EIRP2 PDCCH L Path M Shadowing Model

L Indoor + G
Mi
Mi
Mi
Mi
L Ant L Body + f CP
TX i ic
TX i
Without smart antennas: EPRE PDCCH + G Ant L
TX i

TX i ic
EIRP2 PDCCH
TX i ic
TX i
Ser.: EPRE PDCCH + G Ant L

TX i ic
TX i
TX i
Int.: EPRE PDCCH + G Ant L

TX i ic
E PDSCH
TX
i
Combining
Div
+ G SA
TX i
TX
i
Combining
TX i ic
EIRP2 PDSCH L Path M Shadowing Model

L Indoor + G
Mi
Mi
Mi
Mi
L Ant L Body + f CP
453

Name
Value
TX ic
i
TX
Without smart antennas: EPRE PDSCH + G Ant L

TX ic
i
EIRP2 PDSCH
TX
Array
+ G SA + G SA
TX i ic
TX i
Combining
+ G SA
dBm/Sym
PDSCH EIRP
dB
Path loss
dB
Total losses
dB
Cyclic prefix factor, i.e., the ratio of

the useful symbol energy to the total
symbol energy
Array
Div
+ G SA L
Combining
+ G SA
TX i
L Path
L Model + L Ant
L Path + L
+L
Mi
TX i
+ L Indoor + M Shadowing Model G
Mi
Mi
TX i
TX i
Mi
TX i ic
= Normal
TX i ic
10 Log 6 7,5 If D CP
0
TX
+ L Ant + L Body
f CP
Description

TX ic
i
Ser.: EPRE PDSCH
Int.: EPRE PDSCH + G SA + G SA
L Total
TX
Unit
If
= Extended
TX i ic is an interferer
TX ic
CoMP JTC
C Max
C i
Max -
------------------- 10 10
10 Log
TXi ic
Received max signal level from all

the downlink coherent joint
transmission servers
TX ic
CoMP JTC
C DLRS
C i
DLRS -
------------------- 10 10
10 Log
TXi ic

level from all the downlink coherent
joint transmission servers
TX ic
CoMP JTC
C SS
C i
SS
--------------------
10
10 Log
10
TXi ic
Received SS signal level from all the

downlink coherent joint transmission
servers
TX ic
CoMP JTC
C PBCH
C i
PBCH -
------------------- 10 10
10 Log
TXi ic
Received PBCH signal level from all

TX ic
CoMP JTC
C PDCCH
C i
PDCCH-
-------------------- 10 10
10 Log
TXi ic
Received PDCCH signal level from all

TX ic
C PDSCH
C i
PDSCH-
------------------- 10 10
10 Log
TXi ic
CoMP JTC
E DLRS
E i
DLRS
-------------------10
10 Log
10
TXi ic
CoMP JTC
TX ic
454
Received PDSCH signal level from all


energy per resource element from all

AT332_TRR_E0
Name
Value
Unit
Description
TX ic
CoMP JTC
E SS
E i
SS
-------------------
10 10
10 Log
TXi ic
Received SS energy per resource

element from all the downlink
coherent joint transmission servers
TX ic
CoMP JTC
E PBCH
E i
PBCH
------------------- 10 10
10 Log
TXi ic
Received PBCH energy per resource

TX ic
CoMP JTC
E PDCCH
E i
PDCCH
------------------- 10 10
10 Log
TXi ic
Received PDCCH energy per resource

TX ic
CoMP JTC
E PDSCH
E i
PDSCH
-------------------10
10 Log
10
TXi ic
Received PDSCH energy per resource

6.2.4 Noise Calculation (DL)

Name
Value
Unit
Description
TX i ic
n 0 + 10 Log F
dBm
Thermal noise for one resource

element
dBm
Downlink noise for one resource

element
n 0 Sym
TX ic
i
TX ic
i
n 0 Sym + nf
n Sym
6.2.5 Interference Calculation (DL)

Name
Value
Unit
Description
dBm/Sym
Interfering energy per resource

element (dBm/Sym) received
over downlink reference signals
(Method 1: synchronised
transmission and reception)
Case: RS of the interfered cell
collide only with RS of the
interfering cell
j
TX jc
E DLRS
N j
N TX i ic N TXj jc
---------------------
TX i ic TX j jc
10
Ant TX
Ant TX
Ant TX
10 Log ------------------ 10
+ fO
+ 10 Log --------------------------------------------TX i ic
TXi ic
N Ant TX
N Ant TX
dBm/Sym
TX j jc
TX i ic TX j jc
TX j jc TX i ic TX j jc
E PDCCH + f PDCCH
E PDSCH + f PDSCH
---------------------------------------------------------------------------------------------------------------------------------------------
10
10
+ 3 10
10
-----------------------------------------------------------------------------------------------------------------------------------
4

collide with RS, PDCCH, and
PDSCH of the interfering cell
With 1 or 2 antenna ports
TX jc
TX j jc
DLRS
j
TX ic
E DLRS
N i
---------------------
TX i ic TX j jc
10
Ant TX
10 Log ------------------ 10
+
f
O
TX jc
j
N Ant TX
TX jc
TX j jc
DLRS
455

Name
Value
Unit
TX jc
TX j jc
DLRS
j
TX jc
E
DLRS
N j
N TXi ic N TX j jc
---------------------
TX ic TX jc
i
j
10
Ant TX
Ant TX
Ant TX
10 Log ------------------ 10
+ fO
+ 10 Log --------------------------------------------TX ic
TXi ic
i
N Ant TX
N Ant TX
dBm/Sym
TX j jc
TX i ic TX j jc
E PDCCH + f PDCCH
E PDSCH + f PDSCH
---------------------------------------------------------------------------------------------------------------------------------------------
10
10
+
5
10
10
------------------------------------------------------------------------------------------------------------------------------------
6
TX jc
TX j jc
DLRS
j
TX jc
E DLRS
N TXi ic N TX j jc
N j
---------------------
TX i ic TX j jc
10
Ant TX
Ant TX
Ant TX
------------------+ 10 Log --------------------------------------------

10 Log TX ic 10
TX i ic
+ fO
i
N Ant TX
N Ant TX
dBm/Sym
TX jc
TX ic TX jc
TX jc TX ic TX jc
j
i
j
j
i
j
E PDCCH + f PDCCH
E PDSCH + f PDSCH
---------------------------------------------------------------------------------------------------------------------------------------------
10
10
10
+
2
10
------------------------------------------------------------------------------------------------------------------------------------
3
TX jc
TX jc
j
DLRS
TX ic TX jc
TX jc
TX ic TX jc
TX i ic TX j jc
TX j jc
TX i ic TX j jc
j
j
i
j
i
j
EPDCCH
E
+f
+f
PDSCH PDSCH
PDCCH
-----------------------------------------------------------------------
----------------------------------------------------------------------10
10
10
+ 3 10
10 Log ------------------------------------------------------------------------------------------------------------------------------------
4
TX i ic TX j jc
+ fO
TX j jc
TX j jc
DLRS
EPDCCH + fPDCCH
E
+f
PDSCH PDSCH
-----------------------------------------------------------------------
----------------------------------------------------------------------10
10
10
+ 5 10
10 Log ------------------------------------------------------------------------------------------------------------------------------------
6
TX i ic TX j jc
TX i ic TX j jc
TX j jc
TX i ic TX j jc
EPDCCH + fPDCCH
E PDSCH + f PDSCH
-----------------------------------------------------------------------
----------------------------------------------------------------------10
10
10
+ 2 10
10 Log ------------------------------------------------------------------------------------------------------------------------------------
3

dBm/Sym
collide with PDCCH and PDSCH of
the interfering cell
With 4 or 8 antenna ports and
TX i ic
N SD PDCCH 1
TX jc
j
TX jc
j
E PBCH
ESS
--------------------TX j jc
-------------------
10
10
10
N Sym SS + 10
N Sym PBCH
-
10 Log ------------------------------------------------------------------------------------------------------------TX j jc
N
+
N
Sym SS
Sym PBCH
TX ic TX jc
i
j
+ fO
456
TX i ic
N SD PDCCH 1
TX i ic
TX ic TX jc
i
j
TX jc
j


dBm/Sym
+ fO
SS PBCH
TX i ic
N SD PDCCH = 1
N SD PDCCH = 1
TX j jc
TX j jc


dBm/Sym
With 1 or 2 antenna ports
+ fO
DLRS
Description
TX jc
j
+ f MIMO
dBm/Sym

over the SS and the PBCH

AT332_TRR_E0
Name
Value
Unit
Description
TX jc
TX j jc
PDCCH
E
TX jc
DLRS
j
-------------------TX ic TX jc
N Sym DLRS in PDCCH
1
i
j
10
- 10
10 Log ------------------+ fO
----------------------------------------TX ic
TXj jc
i
N Ant TX
N Sym PDCCH
TX jc
TX ic TX jc
j
i
j
TX ic
E
+f
TX j jc
PDCCH PDCCH
N i
-----------------------------------------------------------------------
N
10
Sym PDCCH
Sym DLRS in PDCCH
- 10
+ 10 L og ----------------------------------------------------------------------------TX i ic
N Sym PDCCH

over the PDCCH
dBm/Sym
Case: PDCCH of the interfered cell
collides with PDCCH and all the RS
of the interfering cell
TX jc
TX j jc
PDCCH
TX j jc
PDCCH
j
TX ic
E
TX j jc
TX i ic
DLRS
N i
---------------------
N
N
10
Ant TX
Sym DLRS in PDCCH
Sym DLRS in PDCCH
- 10
10 Log ------------------ -----------------------------------------------------------------------------------------TX i ic
TXj jc
N Ant TX
N Sym PDCCH
TX jc
TX ic TX jc
j
i
j
TX jc
E PDCCH + f PDCCH
N j
-----------------------------------------------------------------------
TX ic TX j jc
10
Sym PDCCH
+f i
+ 10 L og ---------------------------- 10
O
TX i ic
N Sym PDCCH
TX j jc
TX i ic TX j jc
E PDCCH + f PDCCH

over the PDCCH
dBm/Sym
collides with PDCCH and some RS
over the PDCCH
dBm/Sym
collides only with PDCCH of the
interfering cell
TX i ic TX j jc
+ fO
TX jc
TX jc
j
PDSCH
E DLRS
TX j jc
---------------------
TX ic TX jc
N
1
j
Sym DLRS in PDSCH
- 10 10 + f O i
----------------------------------------10 Log ------------------TX j jc
TX i ic
N Sym PDSCH
N Ant TX
TX jc
TX ic TX jc
j
i
j
TX ic
E PDSCH + f PDSCH
TX j jc
N i
-----------------------------------------------------------------------
N
10
Sym PDSCH
Sym DLRS in PDSCH
+ 10 L og ---------------------------------------------------------------------------- 10
TX i ic
N Sym PDSCH

over the PDSCH
dBm/Sym
Case: PDSCH of the interfered cell
collides with PDSCH and all the RS
TX jc
TX j jc
PDSCH
TX j jc
PDSCH
j
TX ic
E DLRS
TX j jc
TX i ic
N i
--------------------
N Sym DLRS in PDSCH N Sym DLRS in PDSCH
Ant TX
- 10 10
----------------------------------------------------------------------------------------10 Log ------------------TX j jc
TX i ic
N Ant TX
N Sym PDSCH
TX jc
TX ic TX jc
j
i
j
TX jc
E PDSCH + f PDSCH
N j
-----------------------------------------------------------------------
TX ic TX jc
10
j
Sym PDSCH
+f i
- 10
+ 10 L og ---------------------------O
TX i ic
N Sym PDSCH
TX j jc
TX i ic TX j jc
E PDSCH + f PDSCH
TX i ic TX j jc
+ fO

over the PDSCH
dBm/Sym
collides with PDSCH and some RS
over the PDSCH
dBm/Sym
collides only with PDSCH of the
interfering cell
457

Name
Value
Unit
Description
dBm/Sym

(Method 2: non-synchronised
+ f MIMO
E PBCH + f MIMO
ESS
----------------------------------------------TX j jc
---------------------------------------------
10
10
TX ic TX j jc
N Sym SS + 10
N Sym PBCH
10
- + f O i
10 Log --------------------------------------------------------------------------------------------------------------------------------------------------TX j jc
N Sym SS + N Sym PBCH

dBm/Sym

over the SS and the PBCH
TX jc
TX j jc
DLRS
TX j jc
TX j jc
TX i ic TX j jc
TX j jc
TX j jc
TX jc
TX j jc
PDCCH
TX ic TX jc
E
+f
TX j jc
PDSCH PDSCH
----------------------------------------------------------------------TX ic TX j jc
N
10
Sym PDSCH
- + fO i
+ 10
---------------------------TX j jc
N Sym DL
TX j jc
SS PBCH
TX jc
j
i
j
E j
E
+f
TX jc
TX jc
DLRS
PDCCH PDCCH
j
j
-----------------------------------------------------------------------------------------N Sym DLRS
N Sym PDCCH
10
10
- + 10
----------------------------10 Log 10
------------------------
TX jc
TX jc
j
j
N Sym DL
N Sym DL
TX jc
TX j jc
TX ic TX jc
j
i
j
E j
E PDCCH + f PDCCH
TX jc
TX jc
DLRS
j
j
-----------------------------------------------------------------------------------------N Sym DLRS
N Sym PDCCH
10
10
----------------------------------------------------10 Log 10
TX jc + 10
TX jc
j
j
N Sym DL
N Sym DL
+ 10
TX j jc TX ic TX jc
i
j
E
+f
PDSCH PDSCH
----------------------------------------------------------------------10
dBm/Sym

over the PDCCH
dBm/Sym

over the PDSCH
TX jc
j
N Sym PDSCH
----------------------------TX jc
j
N Sym DL
TX i ic TX j jc
+ fO
TX jc
TX j jc
PDSCH
TX j jc
TX j jc
TX i ic TX j jc
f PDCCH
458
TX ic TX jc
TX i ic TX j jc
E
+f
TX jc
PDSCH PDSCH
j
----------------------------------------------------------------------TX ic TX j jc
N Sym PDSCH
10
- + fO i
+ 10
---------------------------TX j jc
N Sym DL
TX j jc
RSSI
TX jc
j
i
j
E j
E
+f
TX jc
TX jc
DLRS
PDCCH PDCCH
j
j
-----------------------------------------------------------------------------------------N
N Sym PDCCH
10
10
Sym DLRS
- + 10
----------------------------10 Log 10
------------------------
TX jc
TX jc
j
j
N Sym DL
N Sym DL
TX i ic TX j jc
EPDSCH + fPDSCH
TX j jc
---------------------------------------------------------------------10
N Sym PDSCH
10
10 Log -------------------------------------------------------------------------------------------- 10
TX j jc
TX j jc
+
N
N
Sym PDSCH
Sym PDCCH
Interfering energy per frequency

block (dBm/RB) received over 1
frequency block during an OFDM
symbol carrying reference signals
dBm/RB
E DLRS
TX j jc
For number of antenna ports > 1,

--------------------TX j jc
TX ic TX j jc
N Sym PDCCH
10
instead of encircled 10
- 10 + 10 10 2 Min 2 N Ant TX + f O i 8 is used
+ --------------------------------------------------------------------------------------------TX j jc
TX j jc
N Sym PDSCH + N Sym PDCCH
TX j jc
TX i ic TX j jc
E PDCCH + f PDCCH
----------------------------------------------------------------------10
TX j jc
TX jc TX jc TX ic TX jc TX ic TX jc
DL
j
j
i
j
i
j
f MIMO + f TL
+ f ICIC DL
+ f ABS DL
+ f CoMP
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------
TX j jc
10
1 AU
10
DL
10 Log
dB
TX
jc
TX
ic
TX
jc
TX
ic
TX
jc
j
i
j
i
j
DL
f
+
f
+
f
+
f
TL
ICIC DL
ABS DL
CoMP
-----------------------------------------------------------------------------------------------------------------------------------------------TX j jc
10
10
+ AU DL
PDCCH interference weighting

factor

AT332_TRR_E0
Name
Value
TX jc
TX ic TX jc
i
j
f PDSCH
Unit
TX jc
TX ic TX jc
TX ic TX jc
j
j
i
j
i
j
DL
f
+f
+f
+f
+f
MIMO
TL
ICIC DL
ABS DL
CoMP
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
TX jc
j
10
1 AU DL 10
10 Log
dB
TX
TX
TX
ic
TX
jc
TX
ic
TX
jc
j
j
i
j
i
j
DL
G G + f
+f
+f
SA
ICIC DL
ABS DL
CoMP
SA
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
TX jc
10
+ AU DLj 10
TX i ic TX j jc
TX i ic TX j jc
10 Log r O
TX i ic TX j jc
10 Log p Collision
Description
PDSCH interference weighting

factor
dB
Interference reduction factor due

to channel overlap
dB

to static downlink ICIC using
fractional frequency reuse
dB

to the downlink traffic load
dB
Interference reduction due to the

downlink coordinated scheduling
CoMP
f MIMO
j
10 Log N Ant TX
dB
Interference increment due to

more than one transmission
antenna port
Inter Tech
I DL
TX k
P DL Rec
--------------------------------------
F
TX
ic
TX
i
k
TX k ICP DL
interference
fO
f ICIC DL
TX j jc
f TL
TX i ic TX j jc
TX j jc
10 Log TLDL
DL
NCoMP
TX kc
TL DL
DL
k
for k downlink CoMP servers
10 Log P CoMP = f ----------------------------------DL
N CoMP
0
for other interferers
DL
f CoMP
TX j jc
TX jc
6.2.6 C/N Calculation (DL)

Name
TX i ic
CNR DLRS
TX i ic
CNR SS
TX i ic
CNR PBCH
TX i ic
CNR PDCCH
TX i ic
CNR PDSCH
Value
TX i ic
Unit
Description
TX i ic
dB
Downlink reference signals C/N
TX i ic
dB
SS C/N
dB
PBCH C/N
dB
PDCCH C/N
dB
PDSCH C/N
E DLRS n Sym
TX i ic
E SS
TX i ic
n Sym
TX i ic
E PBCH n Sym
TX ic
i
TX ic
i
DL
With MIMO: CNR PBCH = CNR PBCH + G Div PBCH + G Div

TX i ic
TX i ic
E PDCCH n Sym
TX i ic
TX i ic
Mi
DL
With MIMO: CNR PDCCH = CNR PDCCH + G Div PDCCH + G Div

TX i ic
TX i ic
E PDSCH n Sym
TX i ic
TX i ic
Mi
DL
With MIMO: CNR PDSCH = CNR PDSCH + G Div DL + G Div
459

6.2.7 C/(I+N) Calculation (DL)

Name
Value
Unit
TX jc
TX i ic
CINR DLRS
TX ic
j
n
DLRS
Sym
------------------
---------------------
TX i ic
Inter Tech
Inter Tech
10
10
dB
E DLRS 10 Log
10
+I
+ 10
+ NR DL
DL
All TXj jc
TX jc
TX i ic
CINR SS
TX i ic
Downlink reference signals C/(I+N)
TX ic
n Sym
SS PBCH
------------------------
---------------------
TX i ic
Tech
10 10 + I Inter Tech + 10 10 + NR Inter dB
E SS
10 Log
DL
DL
All TXj jc
TX jc
CINR PBCH
Description
SS C/(I+N)
TX ic
n Sym
SS PBCH
---------------------------------------------
TX i ic
10 10 + I Inter Tech + 10 10 + NR Inter Tech
E PBCH 10 Log
DL
DL
dB
All TXj jc
TX i ic
TX i ic
Mi
PBCH C/(I+N)
DL
With MIMO: CINR PBCH = CINR PBCH + G Div PBCH + G Div

TX jc
TX i ic
CINR PDCCH
TX ic
j
n Sym
PDCCH-
----------------------------------------
TX i ic
E PDCCH 10 Log
DL
DL
dB
All TXj jc
TX i ic
TX i ic
Mi
PDCCH C/(I+N)
DL
With MIMO: CINR PDCCH = CINR PDCCH + G Div PDCCH + G Div

TX jc
TX i ic
CINR PDSCH
TX ic
n
PDSCH
Sym
-------------------
---------------------
TX i ic
Inter
Tech
Inter
Tech
10
10
+ NR
10
+I
E PDSCH 10 Log
+ 10
DL
DL
dB
All TXj jc
TX i ic
TX i ic
Mi
PDSCH C/(I+N)
DL
With MIMO: CINR PDSCH = CINR PDSCH + G Div DL + G Div

RSRQ
TX ic
TX i ic
i
10 Log N FB
TX ic
+ E i RSSI
DLRS
TX jc
TX i ic
dB
Reference signal received quality

(RSRQ)
dBm
Received signal strength indicator

(RSSI)
dBm
Downlink reference signals total

noise (I+N)
TX ic
j
n Sym
RSSI -
TX ic
--------------------------------------10
10
i
Inter
Tech
10
+I
10 Log RSSI +
+ 10
12 +
DL
All TX j jc
RSSI
TX i ic
Inter Tech
NRDL
TX ic
i
+ 10 Log NFB
TX jc
TX ic
i
j
n Sym
DLRS
------------------
---------------------
Inter
Tech
10
10
10
+I
10 Log
+ 10
DL
All TXj jc
TX i ic
I + N DLRS
Inter Tech
+ NR DL
460
TX i ic
+ 10 Log 2 N FB

AT332_TRR_E0
Name
Value
TX jc
Unit
Description
dBm
SS and PBCH total noise (I+N)
dBm
PDCCH total noise (I+N)

dBm
PDCCH total noise (I+N)

dBm
PDSCH total noise (I+N)

dBm
PDSCH total noise (I+N)

Unit
Description
dBm
Nominal PUSCH power
dBm
Maximum allowed transmission

power of a user equipment
dBm
Received PUSCH and PUCCH signal

level
TX ic
n
SS PBCH
Sym
---------------------------------------------
Inter Tech
10
10
10 Log
10
+I
+ 10
DL
All TXj jc
TX i ic
I + N SS PBCH
Inter Tech
+ NR DL
TX jc
TX ic
j
n Sym
PDCCH-
----------------------------------------
10 10 + I Inter Tech + 10 10
10 Log
DL
All TXj jc
TX i ic
I + N PDCCH
TX i ic
N Sym PDCCH
Tech
- + NR Inter
+ 10 Log ---------------------------------------------DL
TX i ic
TX i ic
N SF DL + N TDD SSF
TX jc
TX ic
j
n Sym
PDCCH
----------------------------------------
10 10 + I Inter Tech + 10 10
10 Log
DL
All TXj jc
TX i ic
I + N PDCCH
TX i ic
N TXi ic
Tech
Sym PDSCH + N Sym PDCCH
- + NR Inter
+ 10 Log ------------------------------------------------------------------DL
TX ic
i
N SD Slot N Slot SF N SF DL
TX jc
TX ic
n
PDSCH
Sym
----------------------------------------
Inter
Tech
10
10
10
+I
10 Log
+ 10
DL
All TXj jc
I +
TX ic
i
N PDSCH
TX i ic
N Sym PDSCH
Tech
- + NR Inter
+ 10 Log ---------------------------------------------DL
TX i ic
TXi ic
N SF DL + N TDD SSF
TX jc
TX ic
i
j
n Sym
PDSCH
-------------------
---------------------
10 10 + I Inter Tech + 10 10
10 Log
DL
All TXj jc
TX i ic
I + N PDSCH
TX i ic
N TXi ic
Tech
- + NR Inter
+ 10 Log ------------------------------------------------------------------DL
TX i ic
6.2.8 Signal Level Calculation (UL)

Name
TX i ic
P O_PUSCH
Mi
P Allowed
Mi
C PUSCH PUCCH
Value
TX i ic
TX i ic
CINR PUSCH Max + NR UL
TX ic
TX ic
i
i
+ n PUSCH PUCCH 10 Log N FB
TX i ic
TX i ic
TX i ic
Mi
Min P Max 10 Log N FB + P O_PUSCH + FPC L Total
Mi
TX i
EIRP PUSCH PUCCH L Path M Shadowing Model L Indoor + G Ant

L
TX
i
L Ant
i
L Body
+ f CP
461

Name
Value
P
EIRP PUSCH PUCCH
With P
+G
Mi
dBm
PUSCH and PUCCH EIRP of a user

equipment
dB
Cyclic prefix factor, i.e., the ratio of

the useful symbol energy to the total
symbol energy
Unit
Description
dBm
PUSCH and PUCCH thermal noise
dBm
PUSCH and PUCCH noise
Unit
Description
dBm
Received PUSCH and PUCCH

interference
dB
Interference reduction factor due to

the co- and adjacent channel overlap
dB

the interfering mobiles uplink traffic
load
dB

static uplink ICIC using fractional
frequency reuse
dB
Interference reduction due to the

uplink coordinated scheduling CoMP
Mi
= P Eff after power control adjustment

TX i ic
TX i ic
f CP
Description
= P Allowed without power control adjustment and
Unit
= Normal
= Extended
If M i is an interferer
6.2.9 Noise Calculation (UL)

Name
Value
TX ic
TX i ic
i
n 0 + 10 Log N FB
n 0 PUSCH PUCCH
TX ic
i
W FB 1000
TX ic
i
n 0 PUSCH PUCCH + nf
n PUSCH PUCCH
TX ic
i
6.2.10 Interference Calculation (UL)

Name
Value
TX i ic TX j jc
Mj
Mj
I PUSCH PUCCH
TX i ic TX j jc
fO
Mj
f TL UL
TX i ic TX j jc
f ICIC UL
C PUSCH PUCCH + f O
TX i ic TX j jc
f ABS UL
TX i ic TX j jc
Mj
+ f TL UL + f ICIC UL
UL
f CoMP
TX i ic TX j jc
10 Log r O
j
10 Log TL UL
TX i ic TX j jc
10 Log p Collision
UL
NCoMP
TX k kc
TL UL
UL
k
for k uplink CoMP servers
10 Log P CoMP = f ----------------------------------UL
N CoMP
0
UL
f CoMP
462

AT332_TRR_E0
6.2.11 Noise Rise Calculation (UL)

Name
Value
Unit
Description
dB
Uplink noise rise for any mobile Mi in

cell centre of the interfered cell
TXi(ic)
dB
Uplink noise rise for any mobile Mi in

cell-edge of the interfered cell TXi(ic)
dBm
PUSCH and PUCCH total noise (I+N)
Unit
Description
dB
PUSCH and PUCCH C/N
Unit
Description
dB
PUSCH and PUCCH C/(I+N)
dBm
Effective transmission power of a

user equipment after power control
adjustment
TX i ic
NR UL
TX i ic
IPUSCH PUCCH
non-ICIC M i
n PUSCH PUCCH
-
-------------------------------------------
----------------------------------------------------------------------------10
10
10 Log
10
+ 10
All M j
All
TX
jc
Inter Tech
+ NR UL
TX i ic
n PUSCH PUCCH
M
TX i ic
IPUSCH PUCCH
n PUSCH PUCCH
ICIC M i
-
-------------------------------------------
-----------------------------------------------------------------10
10
10 Log
10
+ 10
All Mj
All TXj jc
TX i ic
NR UL ICIC
Inter Tech
+ NR UL
TX ic
i
n PUSCH PUCCH
For any mobile Mi in cell centre of the interfered cell TXi(ic):

I +
TX ic
i
N PUSCH PUCCH
TX i ic
NRUL
TX i ic
+ n PUSCH PUCCH
For any mobile Mi in cell-edge of the interfered cell TXi(ic):

TX i ic
TX i ic
NRUL ICIC + n PUSCH PUCCH
6.2.12 C/N Calculation (UL)

Name
Value
TX i ic
Mi
C PUSCH PUCCH n PUSCH PUCCH
Mi
CNR PUSCH PUCCH
With MIMO:
Mi
TX i ic
Mi
UL
CNR PUSCH PUCCH = CNR PUSCH PUCCH + G Div UL + G Div
6.2.13 C/(I+N) Calculation (UL)

Name
Value
For any mobile Mi in cell centre of the interfered cell TXi(ic):
TX i ic
Mi
CNR PUSCH PUCCH NR UL
For any mobile Mi in cell-edge of the interfered cell TXi(ic):
Mi
CINR PUSCH PUCCH
TX i ic
Mi
CNR PUSCH PUCCH NR ICIC UL

With MIMO:
Mi
TX i ic
Mi
UL
CINR PUSCH PUCCH = CINR PUSCH PUCCH + G Div UL + G Div

Mi
P Eff
Mi
Mi
TX i ic
Mi
Max P Allowed CINR PUSCH PUCCH T M + M PC P Min
B i
UL
463

6.2.14 Calculation of Downlink Cell Resources

Name
Value
Unit
Description
N Sym SRB
None
Number of modulation symbols per

N Sym SSF
DwPTS
N SCa FB N SD SSF
None
Number of DwPTS modulation

symbols per scheduler resource
block in the TDD special subframes
N SCa FB
W FB
--------F
None
Number of subcarriers per frequency

block
None
Total number of modulation symbols

in downlink
None
Number of modulation symbols in

DwPTS
None
Number of PDSCH modulation

symbols
None
Number of PDSCH modulation

symbols in the DwPTS
None
Downlink reference signals overhead
None
Downlink reference signals overhead

in the DwPTS
None

DwPTS
TX i ic
TX i ic
N Sym DL
N FB
TX i ic
TX i ic
N Sym DwPTS
TX ic
i
R DL
TX i ic
N FB
TX ic
i
TX i ic
N Sym SRB N SF DL + N Sym DwPTS

TX i ic
DwPTS
N TDD SSF N Sym SSF
TX ic
i
TX ic
i
TX ic
i
TX ic
i
N Sym DL O DLRS O PSS O SSS O PBCH O PDCCH O UERS
TX i ic
TX i ic
TX i ic
TX i ic
N Sym DwPTS O DLRS DwPTS O PDCCH DwPTS
R DwPTS
TX i ic
TX i ic
O DLRS
N FB
TX i ic
TX i ic
TX i ic
O DLRS DwPTS
N FB
16
24
TX i ic
N DLRS SRB
TX i ic
TX i ic
N DLRS SRB N SF DL + O DLRS DwPTS

TX i ic
TX i ic
N DLRS DwPTS N TDD SSF

TX ic
i
if N Ant TX = 1
TX ic
i
if N Ant TX = 2
TX ic
i
N DLRS DwPTS
See "Calculation of Downlink Cell Resources" on page 541
None

downlink reference signals in DwPTS
of one TDD special subframe
O PSS
2 NFB SS PBCH N SCa FB = 144
None
PSS overhead
O SSS
2 NFB SS PBCH N SCa FB = 144
None
SSS overhead
Extended CP: 216

Normal CP: 240
None
PBCH overhead
None
PDCCH overhead
TX i ic
TX i ic
O PBCH
TX ic
i
TX ic
TX i ic
O PDCCH
TX ic
i
i
TX ic
TX ic
i
N i
SD PDCCH N SCa FB 4 N FB
TX i ic
TX i ic
N SF DL + O PDCCH DwPTS
Otherwise:
TX i ic
TX ic
TX ic
i
i
N
SD PDCCH N SCa FB 2 Min 4 N Ant TX N FB
464
TX i ic
TX i ic

AT332_TRR_E0
Name
Value
Unit
Description
TX ic
i
TX ic
TX i ic
O PDCCH DwPTS
TX ic
i
i
TX ic
TX ic
i
N i
N SCa FB 4 N FB
SD PDCCH
Otherwise:
TX ic
PDCCH overhead in the DwPTS
TX i ic
N TDD SSF
TX ic
TX ic
i
i
Min 2 N i
TX ic
i
N TDD SSF
Without smart antennas and MIMO: 0

TX i ic
With smart antennas and without MIMO: 12 N FB

TX ic
i
O DMRS
TX ic
i
With smart antennas and with MIMO: 24 N FB
TX i ic
N SF DL
TX ic
i
N SF DL
None
UE-specific reference signals or

demodulation reference signal
overhead
Without smart antennas and with SU-MIMO or MU-MIMO and

TX i ic
TX i ic
N Ant TX 4 : 24 N FB
TX i ic
N SF DL
6.2.15 Calculation of Uplink Cell Resources

Name
Value
Unit
Description
N Sym SRB
None
Number of modulation symbols per

N SCa FB
W FB
--------F
None
Number of subcarriers per frequency

block
None
Total number of modulation symbols

in uplink
N Sym UL O ULSRS O ULDRS
None
Nnumber of PUSCH modulation

symbols
TX i ic
TX ic
N SCa FB
i
--------------------- N Sym
UL
N Sym SRB
None
Uplink sounding reference signal

overhead
TX ic
i
TX i ic
N SCa FB
- N Sym
2 -------------------- UL
N Sym SRB
None
Uplink demodulation reference

signal overhead
Unit
Description
bps
Maximum downlink throughput

capacity of a UE category
Unit
Description
bps
Maximum uplink throughput

capacity of a UE category
TX i ic
N Sym UL
TX i ic
R UL
O ULSRS
O ULDRS
TX ic
N i
FB
TX ic
TX ic
i
i
N FB PUCCH N Sym SRB N SF UL
TX i ic
TX i ic
TX i ic
6.2.16 Calculation of Downlink UE Capacity

Name
Value
TX ic
Max
TP UE DL
Max DL
N TBB TTI
TX ic
i
N i
+ N TDD SSF
SF DL
---------------------------------------------------D Frame
6.2.17 Calculation of Uplink UE Capacity

Name
Max
TP UE UL
Value
Max UL
N TBB TTI
TX i ic
N SF UL
----------------D Frame
465

6.2.18 Channel Throughput, Cell Capacity, Allocated Bandwidth

Throughput, and Per-user Throughput Calculation
Name
Value
TX ic
i
R DL
Unit
Description
kbps
Downlink peak RLC channel

throughput
kbps
Downlink effective RLC channel

throughput
kbps
Downlink application channel

throughput
kbps
Downlink peak RLC cell capacity
kbps
Downlink effective RLC cell capacity
kbps
Downlink application cell capacity
kbps
Downlink peak RLC throughput per

user
kbps
Downlink effective RLC throughput

per user
kbps
Downlink application throughput per

user
kbps
Uplink peak RLC channel throughput
Mi
B DL
--------------------------------D Frame
TX i ic
R DL
Mi
CTP P DL
Mi
TX ic
i
DL
G MUG DL
For proportional fair schedulers: --------------------------------D Frame
With SU-MIMO:
Mi
B DL
Max M
Mi
B DL
i
1 + f SU MIMO G SU MIMO DL 1
With MU-MIMO in throughput coverage predictions:

TX i ic
Mi
CTP P DL G MU MIMO DL
M
Mi
Mi
Mi
f TP Scaling
- TP Offset
CTP E DL -----------------------100
Mi
CTP A DL
Mi
TX i ic
Mi
CTP P DL TL DL Max
Cap P DL
M
i
i
CTP P DL 1 BLER B DL
CTP E DL
i
i
Cap P DL 1 BLER BDL
Cap E DL
Mi
Mi
Mi
f TP Scaling
- TP Offset
Cap E DL -----------------------100
Mi
Cap A DL
Mi
Cap P DL
----------------------TX i ic
N Users DL
PUTP P DL
Mi
Cap E DL
----------------------TX i ic
N Users DL
Mi
PUTP E DL
Mi
Mi
Mi
f TP Scaling
- TPOffset
PUTP E DL -----------------------100
Mi
PUTP A DL
TX i ic
R UL
Mi
UL
--------------------------------D Frame
TX i ic
R UL
Mi
CTP P UL
Mi
TX ic
i
UL
For proportional fair schedulers: --------------------------------- G MUG UL
D Frame
With SU-MIMO:
Mi
B UL
Max TX ic
Mi
B UL
i
1 + f SU MIMO G SU MIMO UL 1
With MU-MIMO in throughput coverage predictions:

Mi
TX i ic
CTP P UL G MU MIMO UL
466

AT332_TRR_E0
Name
Value
i
i
CTP P UL 1 BLER B UL
CTP E UL
M
CTP A UL
Mi
i
CTP E UL
TX i ic
Mi
CTP P UL TL UL Max
i
i
Cap P UL 1 BLER B UL
kbps
Uplink effective RLC channel

throughput
kbps
Uplink application channel

throughput
kbps
Uplink peak RLC cell capacity
kbps
Uplink effective RLC cell capacity
kbps
Uplink application cell capacity
kbps
Uplink peak RLC allocated bandwidth

throughput
kbps
Uplink effective RLC allocated

bandwidth throughput
kbps
Uplink application allocated

Mi
Mi
Cap A UL
M
M
f TP Scaling
i
i
- TP Offset
Cap E UL -----------------------100
Mi
i
CTP P UL
i
ABTP P UL
Description
M
f TP Scaling
i
- TP Offset
-----------------------100
Cap P UL
Cap E UL
Unit
ABTP E UL
Mi
ABTP A UL
N FB UL
----------------TX i ic
N FB
i
i
ABTP P UL 1 BLER B UL

Mi
i
ABTP E UL
M
f TP Scaling
i
- TP Offset
-----------------------100
Mi
Cap Mi
M
P UL
- ABTP P i UL
Min ---------------------- TX i ic
N Users UL
kbps
Uplink peak RLC throughput per user
Mi
Cap Mi
M
E UL
- ABTP E i UL
Min ----------------------TX i ic
N Users UL
kbps
Uplink effective RLC throughput per

user
kbps
Uplink application throughput per

user
PUTP P UL
PUTP E UL
Mi
PUTP A UL
Mi
Mi
f TP Scaling
PUTP E UL ------------------------- TP Offset
100
Mi
6.2.19 Scheduling and Radio Resource Management

Name
Value
Sel
Mi
R Min DL
TPD Min DL
---------------------------
Sel
Mi
R Min UL
TPD Min UL
---------------------------
Unit
Description
None
Resources allocated to a mobile to

satisfy its minimum throughput
demand in downlink
None

demand in uplink
R Min DL
None
Remaining downlink cell resources

after allocation for minimum
throughput demands
Sel
i
R Min UL
None
Remaining uplink cell resources after

allocation for minimum throughput
demands
kbps
Remaining throughput demand for a

mobile in downlink
Sel
i
Sel
M
i
CTP P DL
Sel
Mi
Sel
Mi
CTP P UL
TX i ic
R Rem DL
TX i ic
R Rem UL
Sel
i
TPD Rem DL
M
TX i ic
TL DL Max
Sel
Mi
Sel
Mi
TX ic
i
TL UL Max
Sel
Mi
Sel
Sel
Mi
Mi
Max
Min TPD Max DL TP UE DL TPD Min DL
467

Name
Value
Sel
i
TPD Rem UL
M
M
i
Max
i
Min TPD Max UL TP UE UL TPD Min UL
Sel
Mi
Sel
Sel
Mi
TX i ic
CTP P DL
Without MUG
G MUG DL
Sel
i
CTP P UL
Sel
i
CTP P UL
Without MUG
TX ic
i
G MUG UL
Sel
Mi
RD Rem DL
Description
kbps

mobile in uplink
kbps
Downlink peak channel throughput

with multi-user diversity gain
(Proportional Fair)
kbps
Uplink peak channel throughput with

multi-user diversity gain
(Proportional Fair)
None
Remaining resource demand for a

mobile in downlink
None

mobile in uplink
None

satisfy its maximum throughput
demand in downlink
None

demand in uplink
Sel
CTP P DL
Unit
Sel
Mi
TPD Rem DL
---------------------------Sel
Mi
CTP P DL
Sel
Mi
RD Rem UL
Sel
Mi
TPD Rem UL
---------------------------Sel
Mi
CTP P UL
TX ic
i
Sel
Mi
R Rem DL
-
Proportional Fair: Min RD Rem DL -------------------N
Sel
TX i ic
R Rem DL
Mi
-
Round Robin: Min RD Rem DL -------------------N
Sel
Mi
R Max DL
Sel
Mi
TX i ic
RD Rem DL
Proportional Demand: R Eff Rem DL ---------------------------------Sel
Mi
RD Rem DL
Sel
i
Sel
Mi
TPD Rem DL
Max C/I: --------------------------Sel
M
CTP P DL
Sel
TX i ic
Mi
R Rem UL
-
Proportional Fair: Min RD Rem UL -------------------N
Sel
TX i ic
Mi
R Rem DL
-
Round Robin: Min RD Rem DL -------------------N
Sel
Mi
R Max UL
Sel
Mi
TX i ic
RD Rem UL
Proportional Demand: R Eff Rem UL ---------------------------------Sel
Mi
RDRem UL
Sel
Mi
Sel
Mi
TPD Rem UL
Max C/I: --------------------------Sel
Mi
CTP P UL
TX i ic
Sel
Mi
TXi ic
Min R Rem DL
RD Rem DL
Sel
M
i
None
Effective remaining downlink

resources in a cell
(Proportional Demand)
TX i ic
Sel
Mi
TX i ic
Min R Rem UL
RD Rem UL
Sel
M
i
None
Effective remaining uplink resources

in a cell
R Eff Rem DL
R Eff Rem UL
468

AT332_TRR_E0
Name
Value
Unit
Description
Sel
Sel
M
Mi
i
R Max DL CTP E DL
Sel
M
Site
i
-
Max 1 ----------------------------------------------------------------------------------------------------Sel
Sel
Mi
Mi
Site
TP
CTP
Min DL
S1 DL
E DL
Sel
M i Site
None
Site backhaul overflow ratio in

downlink
Sel
Sel
Mi
Mi
CTP
Max UL
UL
Sel
M i Site
Max 1 ------------------------------------------------------------------------------------------------------
Sel
Sel
Mi
Mi
Site
R Min UL CTP E UL
TPS1 UL
Sel
M i Site
None
Site backhaul overflow ratio in uplink
None
Total resources assigned to a mobile

in downlink
(Downlink traffic load of the mobile)
None

in uplink
(Uplink traffic load of the mobile)
Unit
Description
kbps
Downlink peak RLC user throughput
kbps
Downlink effective RLC user

throughput
kbps
Downlink application user

throughput
kbps
Uplink peak RLC user throughput
kbps
Uplink effective RLC user throughput
kbps
Uplink application user throughput
Site
BHOF DL
Site
BHOF UL
Sel
Sel
Mi
TL DL
Sel
Mi
= R DL
Sel
Mi
Sel
Mi
Mi
Mi
Max DL CTP P DL
R Min DL CTP P DL + -----------------------------------------------Site

BHOF DL
-----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P DL
Sel
Sel
Mi
TL UL
Sel
Mi
= R UL
Sel
Mi
Sel
Mi
Mi
Mi
Max UL CTP P UL
R Min UL CTP P UL + -----------------------------------------------Site

BHOF UL
-----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P UL
6.2.20 User Throughput Calculation

Name
Sel
Mi
UTP P DL
Sel
i
UTP E DL
M
Sel
Mi
UTP A DL
Sel
Mi
UTP P UL
Sel
Mi
UTP E UL
Sel
Mi
UTP A UL
Value
Sel
Mi
R DL
Sel
Mi
CTP P DL
Sel
Sel
Mi
i
UTP P DL 1 BLER B DL
Sel
Mi
Sel
Mi
Sel
Mi
f TP Scaling
- TP Offset
UTP E DL -----------------------100
Sel
Mi
R UL
Sel
Mi
CTP P UL
Sel
Sel
Mi
Mi
UTP P UL 1 BLER B UL
Sel
Mi
Sel
Mi
Sel
Mi
f TP Scaling
- TP Offset
UTP E UL -----------------------100
469

6.3 Available Calculations

6.3.1.1 Profile View
The point analysis profile view displays the following calculation results for the selected transmitter based on the calculation
algorithm described in "Signal Level Calculation (DL)" on page 502.
TX ic
i
Downlink reference signal level C DLRS
Path loss L Path
Total losses L Total
,G
, L Ant , L Body , and f CP are not used in the calculations performed for the profile view.
6.3.1.2 Reception View

Analysis provided in the reception view is based on path loss matrices. So, you can display received downlink reference signal
levels from the cells for which calculated path loss matrices are available. For each cell, Atoll displays the received RSRP or
reference signal, SS, or PDSCH signal levels.
Reception level bar graphs show the RSRP or signal levels in decreasing order. The maximum number of bars in the graph
depends on the studied signal level of the best server. The bar graph displays cells whose received RSRP are higher than their
minimum RSRP thresholds and are within a 30 dB margin from the studied signal level of the best server.
You can use a value other than 30 dB for the margin from the studied signal level of the best server, for example a smaller
value for improving the calculation speed. For more information on defining a different value for this margin, see the
Administrator Manual.
6.3.1.3 Interference View

Analysis provided in the interference view is based on path loss matrices. So, you can display the received signal level from
the best server and interfering signal levels from other cells for which calculated path loss matrices are available. For each cell,
Atoll displays the best server RS, SS, or PDSCH signal level, and interference from other cells.
Interference level bar graphs show the interference levels on different channels in decreasing order. The maximum number
of bars in the graph depends on the highest interference level on the studied channel. The bar graph displays cells whose C/
N are higher than the minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the
highest interference level on the studied channel.
You can use a value other than 30 dB for the margin from the highest interference level on the studied channel, for example
a smaller value for improving the calculation speed. For more information on defining a different value for this margin, see
6.3.1.4 Details View

Analysis provided in the details view is based on path loss matrices. So, you can display the received signal level from the best
server and interfering signal levels from other cells for which calculated path loss matrices are available. For each cell, Atoll
displays the RSRP and RS, SS, PBCH, PDCCH, and PDSCH signal levels, as well as interference levels on these channels from
interfering cells.
The results for the best server (first row) are displayed using bold italic characters. Other cells are listed in the decreasing order
of RSRP. All the cells from which the received RSRP is higher than their minimum RSRP thresholds are listed in the table. As
well, interference values are listed for all the cells whose C/N are higher than the minimum interferer C/N threshold and
whose interference levels are within a 30 dB margin from the highest interference level on RS.
You can use a value other than 30 dB for the margin from the highest interference level on RS, for example a smaller value for
improving the calculation speed. For more information on defining a different value for this margin, see the Administrator
Manual.
470

AT332_TRR_E0
6.3.2 Coverage Predictions

6.3.2.1 Downlink Signal Level Coverage Predictions
The following coverage predictions are based on the received downlink reference signal levels:
Coverage by Transmitter
Coverage by Signal Level
Overlapping Zones
For these calculations, Atoll calculates the received downlink reference signal level. Then, Atoll determines the selected
display parameter on each pixel inside the cells calculation area. Each pixel within the calculation area of TXi(ic) is considered
a non-interfering receiver.
These coverage predictions do not depend on the traffic input. Therefore, these calculations are of special interest before and
during the deployment stage of the network to study the coverage footprint of the system.
L
Mi
, G
Mi
Mi
Mi
, L Ant , and L Body are not considered in the calculations performed for the downlink signal level based coverage
predictions.
For more information on downlink reference signal level calculations, see "Signal Level Calculation (DL)" on page 502.
For more information on coverage area determination and available display options, see:
"Coverage Area Determination" on page 471.

"Coverage Display Types" on page 472.
Coverage Area Determination

Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialog box to determine coverage
areas to display. There are three possibilities.
All Servers
The coverage area of each cell TXi(ic) corresponds to the pixels where.
TX ic
TX ic
TX ic
i
i
i
MinimumThreshold C DLRS or L Total or L Path MaximumThreshold

TX ic
TX ic
TX ic
i
i
i
AND
TX i ic
TX j jc
C DLRS Best C DLRS M
ji
Where M is the specified overlap margin (dB). The Best function considers the highest value from a list of values.
If M = 0 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the highest.
If M = 2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the
highest or within a 2 dB margin from the highest
If M = -2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher
than the received downlink reference signal levels from the cells which are 2nd best servers

TX ic
TX ic
TX ic
i
i
i
AND
471

TX ic
TX jc
nd
i
j
C DLRS 2 Best C DLRS M
ji
Where M is the specified overlap margin (dB). The 2nd Best function considers the second highest value from a list of
values.
If M = 0 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the second
highest.
If M = 2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the
second highest or within a 2 dB margin from the second highest.
If M = -2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher
than the received downlink reference signal levels from the cells which are 3rd best servers.
Coverage Display Types

A pixel of a coverage area is coloured if the calculated value of the selected display parameter is greater than or equal to the
defined thresholds values. Coverage consists of several independent layers that can be displayed and hidden on the map.
It is possible to display the coverage predictions with colours depending on any transmitter or cell attribute, and other criteria
such as:
Signal Level (dBm, dBV, dBV/m)

Best Signal Level (dBm, dBV, dBV/m): Where cell coverage areas overlap, Atoll keeps the highest value of the signal
level.
Path Loss (dB)
Total Losses (dB)
Best Server Path Loss (dB): Where cell coverage areas overlap, Atoll determines the best cell (i.e., the cell with the
highest downlink reference signal level) and evaluates the path loss from this cell.
Best Server Total Losses (dB): Where cell coverage areas overlap, Atoll determines the best cell (i.e., the cell with the
highest downlink reference signal level) and evaluates the total losses from this cell.
Number of Servers: Atoll evaluates the number of cells that cover a pixel (i.e., the pixel falls within the coverage areas
of these cells).
6.3.2.2 Effective Signal Analysis Coverage Predictions

The following coverage predictions are based on the received downlink reference signal, SS, PDSCH, and PUSCH and PUCCH
signal levels and noise, and take into account the receiver characteristics ( L
Mi
,G
Mi
Mi
Mi
, L Ant , and L Body ) when calculating the
required parameter:
Effective Signal Analysis (DL)

Effective Signal Analysis (UL)
For these calculations, Atoll calculates the received signal level or C/N level at each pixel for the channel type being studied,
i.e., RS, SS, PBCH, PDCCH, PDSCH, PUSCH and PUCCH. Each pixel within the calculation area of TXi(ic) is considered a noninterfering receiver. The properties of the non-interfering probe receiver are set by selecting a terminal, a mobility type, and
a service.
For more information on signal level calculations, see:
"Signal Level Calculation (DL)" on page 502.

"Signal Level Calculation (UL)" on page 525.
For more information on C/N level calculations, see:
"C/N Calculation (DL)" on page 518.

"C/N Calculation (UL)" on page 531.
472


AT332_TRR_E0

These coverage predictions are best server coverage predictions, i.e., the coverage area of each cell comprises the pixels
where the cell is the best server. Best server for each pixel is calculated as explained in "Best Server Determination" on
page 537.
A pixel of a coverage area is coloured if the calculated value of the selected display type parameter is greater than or equal to
the defined thresholds values. Coverage consists of several independent layers that can be displayed and hidden on the map.
It is possible to display the Effective Signal Analysis (DL) coverage prediction with colours depending on the following display
options:
RSRP (RS EPRE) Level (DL) (dBm)

RS Signal Level (DL) (dBm)
SS Signal Level (DL) (dBm)
PBCH Signal Level (DL) (dBm)
PDCCH Signal Level (DL) (dBm)
PDSCH Signal Level (DL) (dBm)
RS C/N Level (DL) (dB)
SS C/N Level (DL) (dB)
PBCH C/N Level (DL) (dB)
PDCCH C/N Level (DL) (dB)
PDSCH C/N Level (DL) (dB)
Cell Edge Margin (dB): Atoll calculates the cell edge margin as explained in "Best Server Determination" on page 537.
Pixels are coloured according to the thresholds defined in the coverage prediction.
Transmitter: Atoll determines the best serving transmitters as explained in "Best Server Determination" on page 537.
Pixels are coloured according to the colours of the transmitter symbols on the map.
Cell-edge Areas: Atoll determines the cell-edge areas as explained in "Best Server Determination" on page 537. Pixels
are coloured according to the colours of the transmitter symbols on the map.
CoMP Set (DL): Atoll determines the CoMP set to which the best serving transmitter belongs in the downlink, as
explained in "Best Server Determination" on page 537. Pixels are assigned colours according to the CoMP sets they
belong to.
Number of CoMP Servers (DL): Atoll determines the number of CoMP servers covering each pixel in the downlink.
Pixels are assigned colours according to the number of downlink CoMP servers.
It is possible to display the Effective Signal Analysis (UL) coverage prediction with colours depending on the following display
options:
PUSCH & PUCCH Signal Level (UL) (dBm)

PUSCH & PUCCH C/N Level (UL) (dB)
CoMP Set (UL): Atoll determines the CoMP set to which the best serving transmitter belongs in the uplink, as explained
in "Best Server Determination" on page 537. Pixels are assigned colours according to the CoMP sets they belong to.
Number of CoMP Servers (UL): Atoll determines the number of CoMP servers covering each pixel in the uplink. Pixels
are assigned colours according to the number of uplink CoMP servers.
6.3.2.3 C/(I+N)-based Coverage Predictions

The following coverage predictions are based on the received signal levels, total noise, and interference.
Coverage by C/(I+N) Level (DL)

Service Area Analysis (DL)
Coverage by Throughput (DL)
Coverage by Quality Indicator (DL)
Coverage by C/(I+N) Level (UL)
Service Area Analysis (UL)
Coverage by Throughput (UL)
Coverage by Quality Indicator (UL)
These coverage predictions take into account the receiver characteristics ( L
Mi
,G
Mi
Mi
Mi
required parameter. For these calculations, Atoll calculates the received signal level, noise, and interference at each pixel.
Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. The properties of the non-interfering
probe receiver are set by selecting a terminal, a mobility type, and a service.
The downlink coverage predictions are based on the downlink traffic loads of the cells, and the uplink coverage predictions
are based on the uplink noise rise values. These parameters can either be calculated by Atoll during the Monte Carlo
simulations, or set manually by the user for all the cells.
473

For more information on RSRQ, RSSI, C/(I+N), (I+N), and bearer calculations, see:
"C/(I+N) and Bearer Calculation (DL)" on page 520.

"C/(I+N) and Bearer Calculation (UL)" on page 534.
For more information on thoughput calculations, see:
"Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation" on
page 551.


These coverage predictions are all best server coverage predictions, i.e., the coverage area of each cell comprises the pixels
page 537.
It is possible to display the Coverage by C/(I+N) Level (DL) coverage prediction with colours depending on the following display
options:
RSRQ Level (DL) (dB)

RSSI Level (DL) (dBm)
RS C/(I+N) Level (DL) (dB)
SS C/(I+N) Level (DL) (dB)
PBCH C/(I+N) Level (DL) (dB)
PDCCH C/(I+N) Level (DL) (dB)
SS & PBCH Total Noise (I+N) (DL) (dBm)
PDSCH C/(I+N) Level (DL) (dB)
PDSCH & PDCCH Total Noise (I+N) (DL) (dBm)
It is possible to display the Service Area Analysis (DL) coverage prediction with colours depending on the following display
options:
Bearer (DL)
Modulation (DL): Modulation used by the bearer
Service
It is possible to display the Coverage by Throughput (DL) coverage prediction with colours depending on the following display
options:
Peak RLC Channel Throughput (DL) (kbps)

Effective RLC Channel Throughput (DL) (kbps)
Application Channel Throughput (DL) (kbps)
Peak RLC Cell Capacity (DL) (kbps)
Effective RLC Cell Capacity (DL) (kbps)
Application Cell Capacity (DL) (kbps)
Peak RLC Throughput per User (DL) (kbps)
Effective RLC Throughput per User (DL) (kbps)
Application Throughput per User (DL) (kbps)
It is possible to display the Coverage by Quality Indicator (DL) coverage prediction with colours depending on the following
display options:
Quality indicators available in the document (Quality Indicators table): Atoll calculates the PDSCH C/(I+N) levels
received from the best serving cells at each pixel of their coverage areas. From the C/(I+N), Atoll determines the best
bearer available on each pixel. Then, for the calculated C/(I+N) and bearer, it determines the value of the selected
quality indicator from the quality graphs defined in the reception equipment of the selected terminal.
It is possible to display the Coverage by C/(I+N) Level (UL) coverage prediction with colours depending on the following display
options:
474
PUSCH & PUCCH C/(I+N) Level (UL) (dB)

AT332_TRR_E0
PUSCH & PUCCH Total Noise (I+N) (UL) (dBm)

Allocated Bandwidth (UL) (No. of Frequency Blocks)
PUSCH & PUCCH C/(I+N) Level for 1 Frequency Block (UL) (dB): PUSCH & PUCCH C/(I+N) level with N FB UL = 1
Transmission Power (UL) (dBm)
It is possible to display the Service Area Analysis (UL) coverage prediction with colours depending on the following display
options:
Bearer (UL)
Modulation (UL): Modulation used by the bearer
Service
It is possible to display the Coverage by Throughput (UL) coverage prediction with colours depending on the following display
options:
Peak RLC Channel Throughput (UL) (kbps)

Effective RLC Channel Throughput (UL) (kbps)
Application Channel Throughput (UL) (kbps)
Peak RLC Cell Capacity (UL) (kbps)
Effective RLC Cell Capacity (UL) (kbps)
Application Cell Capacity (UL) (kbps)
Peak RLC Allocated Bandwidth Throughput (UL) (kbps)
Effective RLC Allocated Bandwidth Throughput (UL) (kbps)
Application Allocated Bandwidth Throughput (UL) (kbps)
Peak RLC Throughput per User (UL) (kbps)
Effective RLC Throughput per User (UL) (kbps)
Application Throughput per User (UL) (kbps)
It is possible to display the Coverage by Quality Indicator (UL) coverage prediction with colours depending on the following
display options:
Quality indicators available in the document (Quality Indicators table): Atoll calculates the PUSCH and PUCCH C/(I+N)
levels received at the best serving cells from each pixel of their coverage areas. From the C/(I+N), Atoll determines the
best bearer available on each pixel. Then, for the calculated C/(I+N) and bearer, it determines the value of the selected
quality indicator from the quality graphs defined in the reception equipment of the best serving cell.
6.3.2.4 Cell Identifier Collision Zones Coverage Prediction

The Cell Identifier Collision Zones coverage prediction is based on the received downlink reference signal levels. Atoll
calculates the received downlink reference signal level then Atoll determines the selected display parameter on each pixel
inside the cells calculation area. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. L
G
Mi
Mi
Mi
Mi
, L Ant , and L Body are not considered in the calculations.
For more information on downlink reference signal level calculations, see "Signal Level Calculation (DL)" on page 502.


areas to display. It is possible to determine the coverage area based on the best signal level. The coverage area of each cell
TXi(ic) corresponds to the pixels where:
TX ic
TX ic
TX ic
TX ic
TX jc
i
i
i
i
j
MinimumThreshold C DLRS or L Total or L Path MaximumThreshold AND C DLRS Best C DLRS M
ji
If M = 0 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the highest.
If M = 2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the highest
or within a 2 dB margin from the highest
475

If M = -2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher than
the received downlink reference signal levels from the cells which are 2nd best servers

It is possible to display the coverage predictions with colours per cell or:
Number of interferers
Number of interferers per cell
6.3.3 Calculations on Subscriber Lists

When calculations are performed on a list of subscribers by running the Automatic Server Allocation, Atoll calculates the path
loss again for the subscriber locations and heights because the subscriber heights can be different from the default receiver
height used for calculating the path loss matrices.
Atoll calculates the following parameters for each subscriber in the list whose Lock Status is set to None.
Serving Base Station and Reference Cell as described in "Best Server Determination" on page 537.
Atoll calculates the following parameters for each subscriber in the list that has a serving base station assigned and whose
Lock Status is set to None or Server.
Azimuth ( ): Angle with respect to the north for pointing the subscriber terminal antenna towards its serving base
station.
Mechanical Downtilt ( ): Angle with respect to the horizontal for pointing the subscriber terminal antenna towards
its serving base station.
Atoll calculates the remaining parameters for each subscriber in the list that has a serving base station assigned, using the
properties of the default terminal and service. For more information, see:

"Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation" on
page 551.
6.3.4 Monte Carlo Simulations

The simulation process is divided into two steps.
Generating a realistic user distribution as explained in "User Distribution" on page 476.

Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data. The resulting user
distribution complies with the traffic database and maps selected when creating simulations.
Scheduling and Radio Resource Management as explained under "Simulation Process" on page 479.
6.3.4.1 User Distribution

During each simulation, Atoll performs two random trials. The first random trial generates the number of users and their
activity status as explained in the following sections depending on the type of traffic input.
"Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 477.
"Simulations Based on Sector Traffic Maps" on page 478.
Once all the user characteristics have been determined, a second random trial is performed to obtain their geographical
locations weighted according to the clutter classes, and whether they are indoor or outdoor according to the percentage of
indoor users per clutter class.
[Simulation]
RandomTotalUsers=0
476

AT332_TRR_E0
6.3.4.1.1
Simulations Based on User Profile Traffic Maps and Subscriber Lists

user profiles with an associated mobility type and a given density, i.e., number of users of a user profile per km.
User profile traffic maps: Each polygon or line of the map is assigned a density of users with a given user profile and mobility
type. If the map is composed of points, each point is assigned a number of users with given user profile and mobility type.
Fixed subscribers listed in subscriber lists have a user profile assigned to each of them.
User profiles model the behaviour of the different user categories. Each user profile contains a list of services and parameters
describing how these services are accessed by the user.
The number of users of each user profile is calculated from the surface area (SEnv) of each environment class map (or each
polygon) and the user profile density (DUP).
N Users = S Env D UP
In case of user profile traffic maps composed of lines, the number of users of each
user profile is calculated from the line length (L) and the user profile density (DUP)
(users per km): N Users = L D UP
The number of users is a direct input when a user profile traffic map is composed of
points.
Atoll calculates the probability for a user being active at a given instant in the uplink and in the downlink according to the
service usage characteristics described in the user profiles, i.e., the number of voice calls or data sessions, the average
duration of each voice call, or the volume of the data transfer in the uplink and the downlink in each data session.
Voice Service (v)
User profile parameters for voice type services are:
The user terminal equipment used for the service (from the Terminals table).
The average number of calls per hour N Call .
The average duration of a call (seconds) D Call .
N Call D Call
Calculation of the service usage duration per hour ( p 0 : probability of an active call): p 0 = ---------------------------3600
Calculation of the number of users trying to access the service v ( n v ): n v = N Users p 0
The activity status of each user depends on the activity periods during the call, i.e., the uplink and downlink activity
UL
DL
factors defined for the voice type service v, f Act and f Act .
UL
DL
Probability of being inactive: p Inactive = 1 f Act 1 f Act

UL
UL
DL
Probability of being active in the uplink: p Active = f Act 1 f Act

DL
DL
UL
Probability of being active in the downlink: p Active = f Act 1 f Act

UL + DL
UL
DL
Probability of being active in the uplink and downlink both: p Active = f Act f Act
Number of inactive users: n v Inactive = n v p Inactive
UL
UL
Number of users active in the uplink: n v Active = n v p Active

DL
DL
Number of users active in the downlink: n v Active = n v p Active

UL + DL
UL + DL
Number of users active in the uplink and downlink both: n v Active = n v p Active
Therefore, a connected user can be either active on both links, inactive on both links, active on UL only, or active on
DL only.
477

Data Service (d)

User profile parameters for data type services are:
The average number of data sessions per hour N Session .
The average data volume (in kBytes) transferred in the downlink V
The average throughputs in the downlink
Calculation of activity probabilities: f
UL
DL
TP Average
DL
and the uplink
and the uplink V
UL
TP Average
UL
during a session.
for the service d.
UL
DL
N Session V 8
N Session V 8
DL
= ------------------------------------------ and f = -----------------------------------------UL
DL
TP Average 3600
TP Average 3600
UL
DL
Probability of being inactive: p Inactive = 1 f 1 f

UL
Probability of being active in the uplink: p Active = f

DL
UL
DL
1 f
Probability of being active in the downlink: p Active = f
DL
UL
1 f
UL + DL
Probability of being active in the uplink and downlink both: p Active = f
UL
DL
Calculation of number of users:

Number of inactive users: n d Inactive = N Users p Inactive
UL
UL
Number of users active in the uplink: n d Active = N Users p Active

DL
DL
Number of users active in the downlink: n d Active = N Users p Active

UL + DL
UL + DL
Number of users active in the uplink and downlink both: n d Active = N Users p Active
Calculation of the number of active users trying to access the service d (nd):
UL
UL + DL
DL
n d = n d Active + n d Active + n d Active

are average distributions. The service and the activity status of each user are randomly
drawn in each simulation. Therefore, if you calculate several simulations at once, the
average number of users per service and average numbers of inactive, active on UL,
active on DL and active on UL and DL users, respectively, will correspond to calculated
distributions. But if you check each simulation, the user distribution between services as
well as the activity status distribution between users can be different in each of them.
6.3.4.1.2

Sector traffic maps are also referred to as live traffic maps. Live traffic data from the O&M is spread over the best server
coverage areas of the transmitters included in the traffic map. Either throughput demands per service or the number of active
users per service are assigned to the coverage areas of each transmitter.
For each transmitter TXi and each service s,
Sector Traffic Maps (Throughputs)

Atoll calculates the number of active users of each service s on UL and DL in the coverage area of TXi as follows:
N
UL
UL
DL
TP Cell
TP Cell
DL
= ---------------------- and N = ---------------------UL
DL
TP Average
TP Average
UL
Where TP Cell is the total uplink throughput demand defined in the map for any service s for the coverage area of the
DL
transmitter, TP Cell is the total downlink throughput demand defined in the map for any service s for the coverage
UL
DL
area of the transmitter, TP Average is the average uplink requested throughput of the service s, and TP Average is the
average downlink requested throughput of the service s.
478
Sector Traffic Maps (# Active Users)

AT332_TRR_E0
UL
Atoll directly uses the defined N and N

coverage area using the service s.
DL
values, i.e., the number of active users on UL and DL in the transmitter
At any given instant, Atoll calculates the probability for a user being active in the uplink and in the downlink as follows:
Users active in the uplink and downlink both are included in the N
UL
UL
accurately determine the number of active users in the uplink ( n Active
and N
DL
values. Therefore, it is necessary to

UL + DL
DL
), in the downlink ( n Active ), and both ( n Active ).
As for the other types of traffic maps, Atoll considers both active and inactive users.
UL
DL
factors defined for the service, f Act and f Act .

UL
DL

UL
UL
DL

DL
DL
UL

UL + DL
UL
DL
Calculation of the number of active users trying to access the service:
We have: N
UL
UL + DL
UL
= p Active + p Active n and N
DL
UL + DL
DL
= p Active + p Active n
Where, n is the total number of active users in the transmitter coverage area using the service.
UL
UL + DL
DL
UL + DL
N p Active
N p Active
UL + DL
Number of users active in the uplink and downlink both: n Active = Min -------------------------------------- -------------------------------------- or
UL
UL + DL
DL
+ DL
p Active + p Active p Active + p UL
Active
UL + DL
simply, n Active = Min N
UL
DL
f Act N
DL
UL
f Act
UL
Number of users active in the uplink: n Active = N

DL
UL
Number of users active in the downlink: n Active = N

UL
DL
UL + DL
n Active
DL
UL + DL
n Active
UL + DL
And, n = n Active + n Active + n Active
Calculation of the number of inactive users attempting to access the service:

nv
- p Inactive
Number of inactive users: n Inactive = --------------------------1 p Inactive
simulation, the activity status of each user is randomly drawn. Therefore, if you calculate
simulation, the activity status distribution between users can be different in each of
them.
6.3.4.2 Simulation Process

LTE cells include intelligent schedulers and radio resource management features for regulating network traffic loads,
optimising spectral efficiency, and satisfying the QoS demands of the users. Each Monte Carlo simulation in the Atoll LTE
module is a snap-shot of the network with resource allocation carried out over a duration of 1 second (100 frames). The steps
of this algorithm are listed below.
The simulation process can be summed up into the following iterative steps.
479


1. Generates mobiles according to the input traffic data as explained in "User Distribution" on page 476.
2. Sets initial values for the following parameters:
Cell transmission powers and EPREs are calculated from the maximum power and EPRE offset values defined by
the user as explained in "Downlink Transmission Power Calculation" on page 485.
Mobile transmission power is set to the maximum mobile power ( P Max ).
Cell loads ( TL DL
Mi
TX i ic
TX i ic
, TL UL
TX i ic
, NR UL
TX i ic
TX i ic
TX i ic
, NRUL ICIC , r DL CE , and AU DL
) are set to their current values in the
Cells table.
3. Determines the best servers for all the mobiles generated for the simulation, and determines whether they are in the
cell centre or cell-edge, as explained in "Best Server Determination" on page 537.
4. Sets the maximum PUSCH C/(I+N) of each cell to a value high enough to ensure that it will not cause any power
constraints for cell-edge mobiles.
TX i ic
For all the mobiles Mi served by any cell TXi(ic) in the uplink, Atoll calculates CINR PUSCH Max as follows to ensure
access to the highest bearer using all the frequency blocks.
From fractional power control (see "Signal Level Calculation (UL)" on page 525), we know that:
Mi
P Allowed = CINR PUSCH Max + NRUL + n PUSCH PUCCH + FPC L Total
(1)
Where CINR PUSCH Max is the maximum PUSCH C/(I+N), NRUL is the noise rise, n PUSCH PUCCH is the uplink thermal
noise, FPC is the fractional power control factor, and L Total are the total losses.
Mi
Transmitting P Allowed , a mobile Mi can access the highest bearer if:

Mi
Mi
P Allowed NR UL n PUSCH PUCCH L Total = T B
(2)
Mi
Where T B is the bearer selection thresholds of the highest bearer defined in the reception equipment used by the
cell TXi(ic).
Mi
Combining equations (1) and (2), we get the CINR PUSCH Max for each mobile Mi that ensures access to the highest
bearer:
M
TX ic
i
i
i
CINR PUSCH Max = T B + 1 FPC L Total
For each cell TXi(ic), the highest value is kept:

TX ic
i
i
CINR PUSCH Max = Max CINR PUSCH Max
All M
i
For each iteration k, the simulation process,

5. Determines the downlink and uplink C/(I+N) and bearers for each of these mobiles as explained in "C/(I+N) and Bearer
Calculation (DL)" on page 520 and "C/(I+N) and Bearer Calculation (UL)" on page 534 respectively.
6. Determines the channel throughputs at the mobile as explained in "Throughput Calculation" on page 541.
7. Performs radio resource management and scheduling to determine the amount of resources to allocate to each
mobile according to the service priorities and throughput demands of each mobile using the selected scheduler as
explained in "Scheduling and Radio Resource Management" on page 556.
8. Calculates the user throughputs after allocating resources to each mobile as explained in "User Throughput
Calculation" on page 565.
480

AT332_TRR_E0
Figure 6.1: LTE Simulation Algorithm

9. Updates the traffic loads, and noise rise values of all the cells according to the resources in use and the total resources
as follows:
Calculation of Traffic Loads:
Atoll calculates the traffic loads for all the cells TXi(ic).
TX ic
i
TL DL
TX ic
i
RDL and TLUL

i
Mi
RUL
i
Mi
TX i ic
For MU-MIMO, TL DL
MU MIMO DL
Mi
RC DL
TX i ic
and TL UL
MU MIMO DL
M
i
MU MIMO UL
Mi
RC UL
MU MIMO UL
M
i
Calculation of Uplink Noise Rise:

For each victim cell TXi(ic), the uplink noise rise is calculated and updated by considering each interfering mobile Mj
as explained in "Interference Calculation (UL)" on page 527.
Calculation of Downlink Cell-edge Traffic Ratio:
Atoll calculates the downlink cell-edge traffic ratio for all the cells as follows:
TX i ic
CE
Mi
R DL
CE
i
r DL CE = -------------------TX i ic
TL DL
481

Where
CE
i
R DL
is the sum of the percentages of the downlink cell resources allocated to mobiles in the cell-edge.
CE
i
Calculation of Downlink AAS Usage:

Atoll calculates the downlink AAS usages for all the cells as follows:
Mi
AAS
Mi
TX i ic
AAS
= ------------------------------TX i ic
TL DL
AU DL
Where
R DL
Mi
R DL
AAS
is the sum of the percentages of the downlink cell resources allocated to mobiles served by the
AAS
smart antennas.
10. Performs uplink noise rise control as follows:
For each cell TXi(ic), Atoll calculates the difference between the current and the maximum noise rise values (in terms
of IoT, i.e., the ratio of interference over thermal noise I/N which can be calculated from the noise rise:
IoT = I/N = (I+N)/N - 1):
TX ic
TX ic
NR i
NR i
UL
UL Max
-----------------------
------------------------------
10
10
1 10 Log 10
1
= 10 Log 10
TX i ic
NR UL
TX ic
i
Here NRUL
is the uplink noise rise of the cell TXi(ic) calculated in step 9.
The default method of uplink noise rise control is the best effort method. This means that uplink noise rise control is
not part of the simulation convergence criteria. In other words, a simulation will converge once the downlink and
uplink traffic loads and the uplink noise rise values are stable, irrespective of whether or not the noise rise control has
been successful. The resulting noise rise values may be higher than the maximum allowed values defined per cell. If
the resulting noise rise values are higher than the maximum allowed, this means that the noise rise control requires
more iterations for stabilising the overall networks noise rise than those needed by the simulation to converge. If you
wish to achieve optimum noise rise control, you should decrease the uplink noise rise convergence threshold defined
for the simulation so that the simulation takes more iterations to converge and allows noise rise control to reach its
goal. The best effort noise rise control works as follows:
TX i ic
If NR UL
0 , the cell TXi(ic) requests its neighbouring cells to decrease the uplink transmission powers of the
mobiles they serve (mobiles interfering TXi(ic)).

TX i ic
If 0 NR UL
If NR UL
TX i ic
M NRC , the cell TXi(ic) does not request any change.
M NRC , the cell TXi(ic) requests its neighbouring cells to increase the uplink transmission powers of
the mobiles they serve (mobiles interfering TXi(ic)).

Here M NRC is a noise rise control margin set to -1 dB by default. This value can be changed through Atoll.ini file by
adding the following lines and setting it to a value other than "1" (positive values are considered as negative margins):
[LTE]
NR_CONTROL_MARGIN_MIN = 1
The uplink transmission powers of the mobiles in neighbouring cells of the cell TXi(ic) are adjusted according to the
request in the next iteration by updating the maximum PUSCH C/(I+N) for the neighbouring cells TXj(jc):
TX j jc
CINR PUSCH Max
482
TX j jc
= Min CINR PUSCH Max
TX i ic
k1
NRUL
TX jc
j
CINR PUSCH Max CINR PUSCH Limit

AT332_TRR_E0
TX jc
j
Here CINR PUSCH Max

TX jc
j
CINR PUSCH Max
is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the current iteration k,
is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the previous iteration k-1,
k1
TX jc
j
CINR PUSCH Limit is an upper limit fixed at 50 dB, and CINR PUSCH Max is the maximum PUSCH C/(I+N) for the
neighbouring cell TXj(jc) as calculated in step 4.
If you wish to include the uplink noise rise control in the simulation convergence criteria, you can change the uplink
noise rise control method from best effort to strict by setting the following option in the Atoll.ini file:
[LTE]
ULNRControlMethod = 1
The strict uplink noise rise control method makes the uplink noise rise control a part of the simulation convergence
criteria. In other words, a simulation will converge once the downlink and uplink traffic loads and the uplink noise rise
values are stable, and the uplink noise rise values of all the cells are less than or equal to the defined maximum uplink
noise rise. The strict noise rise control works as follows:
TX i ic
If NR UL
0 , the cell TXi(ic) requests its neighbouring cells to decrease the uplink transmission powers of the
mobiles they serve (mobiles interfering TXi(ic)).
TX i ic
If NR UL
m NRC , the cell TXi(ic) requests its neighbouring cells to increase the uplink transmission powers of
the mobiles they serve (mobiles interfering TXi(ic)).

Here m NRC is a noise rise control precision level set to 0.5 dB by default. This value can be changed through Atoll.ini
file by adding the following lines:
[LTE]
ULNRControlPrecision = 5
Setting this option to X means that the precision will be taken as 0.X dB. The default value is 5 (= 0.5 dB).
The uplink transmission powers of the mobiles in neighbouring cells of the cell TXi(ic) are adjusted according to the
request in the next iteration by updating the maximum PUSCH C/(I+N) for the neighbouring cells TXj(jc):
TX j jc
CINR PUSCH Max
TX jc
j
= Min CINR PUSCH Max
TX jc
j
Here CINR PUSCH Max

TX j jc
CINR PUSCH Max
TX i ic
k1
NR UL
TX jc
j
CINR PUSCH Max CINR PUSCH Limit
is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the current iteration k,
is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the previous iteration k-1, and
k1
TX j jc
CINR PUSCH Max is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) as calculated in step 4.
At most six neighbouring cells are considered in uplink noise rise control. These six
neighbouring cells are those whose served mobiles generate the highest interference for
the studied cell.
11. Performs the convergence test to see whether the differences between the previous and current values are within the
TX i ic
TL DL
TX i ic
TL UL
TX ic
i
NR UL
TX ic
i
Max TL DL
All TX ic
i
TL UL
TX ic
i
Max TL UL
All TX ic
i
TX i ic
TL DL
TX i ic
TX ic
i
Max NR UL
All TX ic
i
k 1
TX ic
i
NR UL
k 1
k 1
483

TX ic
i
If TL DL
TX ic
i
Req
, TL UL
TX ic
i
Req
, and NR UL
Req
are the simulation convergence thresholds defined when creating
the simulation, Atoll stops the simulation in the following cases.

Convergence: Simulation has converged between iteration k - 1 and k, with the best effort uplink noise rise control, if:
TX ic
i
TL DL
TX ic
i
TL DL
TX ic
i
Req
AND TL UL
TX ic
i
TL UL
TX ic
i
Req
AND NR UL
TX ic
i
NR UL
Req
Simulation has converged between iteration k - 1 and k, with the strict uplink noise rise control, if:
TX i ic
TL DL
TX ic
i
NRUL
TX i ic
TL DL
Req
TX i ic
TL UL
AND
TX i ic
TL UL
TX i ic
NR UL
AND
Req
TX i ic
NR UL
Req
AND
TX ic
i
NR UL Max
No convergence: Simulation has not converged even after the defined maximum number of iterations, with the best
effort uplink noise rise control, if:
TX i ic
TL DL
TX i ic
TL DL
TX i ic
Req
OR TL UL
TX i ic
TL UL
TX i ic
Req
OR NR UL
TX i ic
NR UL
Req
Simulation has not converged even after the defined maximum number of iterations, with the strict uplink noise rise
control, if:
TX i ic
TL DL
TX i ic
NRUL
TX i ic
TL DL
Req
OR
TX i ic
TL UL
TX i ic
TL UL
Req
OR
TX i ic
NR UL
TX i ic
NR UL
Req
OR
TX i ic
NR UL Max
12. Repeats the above steps (from step 3.) for the iteration k+1 using the new calculated loads as the current loads.
Simulation Results
At the end of the simulation process, the main results obtained are:
Downlink traffic loads

Uplink traffic loads
Uplink noise rise
Downlink cell-edge traffic ratio
Uplink ICIC noise rise
Downlink AAS usage
Number of co-scheduled MU-MIMO users (DL)
Number of co-scheduled MU-MIMO users (UL)
Maximum PUSCH C/(I+N)
Number of connected users in downlink
Number of connected users in uplink
These results can be used as input for C/(I+N)-based coverage predictions.

In addition to the above parameters, the simulations also list the connection status of each mobile. Mobiles can be rejected
due to:
No Coverage: If an LTE mobile does not have any best serving cell with cell type "LTE" and if an LTE-A mobile does not
have any best serving primary cell with cell type "LTE-A PCell" (step 3.)
No Service: If the mobile is not able to access a bearer in the direction of its activity (step 5.), i.e., UL, DL, or DL+UL, or
if the mobiles minimum throughput demand is higher than the UE throughput capacity.
Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 7.). For LTE-A mobiles,
this applies to the mobiles selected for scheduling by their primary cells.
Resource Saturation: If all the cell resources are used up before allocation to the mobile or if, for a user active in
uplink, the minimum uplink throughput demand is higher than the uplink allocated bandwidth throughput (step 7.).
For LTE-A mobiles, this applies to their primary cells.
Backhaul Saturation: If allocating resources to a mobile makes the effective RLC aggregate site throughputs exceed
the maximum S1 interface throughputs defined for the site. This condition is only verified if the simulation was created
with the Backhaul capacity check box selected (step 7.)
Connected mobiles (step 7.) can be:
484
Connected UL: If a mobile active in UL is allocated resources in UL.

Connected DL: If a mobile active in DL is allocated resources in DL.
Connected DL+UL: If a mobile active in DL+UL is allocated resources in DL+UL.

AT332_TRR_E0
6.4 Calculation Details

The following sections describe all the calculation algorithms used in point analysis, calculation of coverage predictions,
calculations on subscriber lists, and Monte Carlo simulations.
6.4.1 Downlink Transmission Power Calculation

LTE eNode-Bs have a maximum transmission power which is shared by downlink channels. These channels include the
downlink reference signals, SSS, PSS, PBCH, PDCCH (which is considered to include the PHICH and PCFICH), and PDSCH. The
transmission powers of various channels are determined from the distribution of the total energy over a frame among the
resource elements corresponding to these channels. The energy per resource element (EPRE) of the downlink reference
signals is considered to be the reference with respect to which the EPRE of other channels is determined. You can either define
the reference signal EPRE for each cell, or let Atoll calculate it from the cells maximum power and the EPRE offsets of other
channels. The EPRE offsets of channels other than the downlink reference signals can be positive values meaning a relative
boost with respect to the downlink reference signals EPRE, or negative values meaning a reduction with respect to the
downlink reference signals EPRE.
Atoll first determines the EPRE for each channel in the downlink and then the transmission power corresponding to each
channel from the EPRE values.
Input
F : Subcarrier width (15 kHz).
W FB : Width of a frequency block (180 kHz).
N FB SS PBCH : Number of frequency blocks that carry the SS and the PBCH (6).
N Slot SF : Number of slots per subframe (2).
D CP
N SD Slot : Number of symbol durations per slot (7 is D CP
TX ic
i
N SD PDCCH
: Number of PDCCH symbol durations per subframe defined in the TXi(ic) frame configuration or,
otherwise, global network settings.
N FB
TX i ic
: Cyclic prefix duration defined in the TXi(ic) frame configuration or, otherwise, global network settings.
TX i ic
TX ic
i
TX jc
j
and N FB
TX i ic
is Normal, 6 if D CP
is Extended).
: Total number of frequency blocks defined in the frequency bands table for the channel
bandwidth used by the cell.
TX i ic
TX j jc
N FB CE0 and N FB CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the
channel bandwidth used by the cell and PSS ID 0.
TX i ic
TX j jc
TX i ic
TX j jc
TX i ic
N SF DL : Number of downlink subframes in the frame for the cell TXi(ic). It is equal to 10 for FDD frequency bands, and
is determined from the cells TDD frame configuration for TDD frequency bands.
TX i ic
N TDD SSF : Number of TDD special subframes (containing DwPTS, GP, and UpPTS) in the frame for the cell TXi(ic). It is
equal to 0 for FDD frequency bands, and is determined from the cells TDD frame configuration for TDD frequency
bands.
TX i ic
TX i ic
N SF DL and N TDD SSF are determined as follows:

TX i ic
TX i ic
Configuration
N SF DL
N TDD SSF
FDD
10
DSUUU-DSUUU
DSUUD-DSUUD
DSUDD-DSUDD
DSUUU-DSUUD
485


TX ic
i
TX ic
i
Configuration
N SF DL
N TDD SSF
DSUUU-DDDDD
DSUUD-DDDDD
DSUDD-DDDDD
TX i ic
N Ant TX : Number of transmission (downlink) antenna ports defined for the cell TXi(ic).
P Max : Maximum transmission power of the cell TXi(ic).
EPRE DLRS : Downlink reference signal EPRE of the cell TXi(ic).
TX i ic
TX ic
i
TX i ic
You can either set the P Max

TX ic
i
TX i ic
or EPRE DLRS for a cell.
EPRE SS
EPRE PBCH : Energy per resource element offset for the PBCH with respect to the downlink reference signals EPRE.
EPRE PDCCH : Energy per resource element offset for the PDCCH with respect to the downlink reference signals EPRE.
EPRE PDSCH : Energy per resource element offset for the PDSCH with respect to the downlink reference signals EPRE.
: Energy per resource element offset for the SS with respect to the downlink reference signals EPRE.
TX i ic
TX i ic
TX i ic
Calculations
If you have directly entered the downlink reference signal EPRE for the cell, you can skip the section "Calculation of Downlink
Reference Signal EPRE" on page 486 and go directly to the section "Calculation of Other EPREs and Per-channel Powers" on
page 491.
Calculation of Downlink Reference Signal EPRE
In LTE, a resource block (RB) is defined as 1 frequency block by 1 slot. However, schedulers are able to perform resource
allocation every subframe (2 slots). 1 frequency block by 1 subframe (2 slots) is called a scheduler resource block (SRB) in the
calculations below.
The number of modulation symbols (resource elements) per scheduler resource block is calculated as follows:
N Sym SRB = N SCa FB N SD Slot N Slot SF
Where N SCa FB is the number of subcarriers per frequency block calculated as follows:
W FB
N SCa FB = --------F
The number of modulation symbols (resource elements) corresponding to the DwPTS per scheduler resource block in the TDD
special subframes is calculated as follows:
DwPTS
DwPTS
N Sym SSF = N SCa FB N SD SSF

DwPTS
Where N SD SSF is the number of DwPTS symbol durations (OFDM symbols) per special subframe, determined from the TDD
special subframe configuration according to the 3GPP specifications as follows:
486

AT332_TRR_E0
Special
Subframe
Configuration
Cyclic Prefix = Normal

DwPTS
GP
N SD SSF
DwPTS
N SD SSF
Cyclic Prefix = Extended

UpPTS
DwPTS
GP
UpPTS
N SD SSF
DwPTS
N SD SSF
10
10
11
10
12
10
11
GP
N SD SSF
GP
UpPTS
UpPTS
N SD SSF
The total number of modulation symbols (resource elements) in downlink is calculated as follows:
TX i ic
TX i ic
N Sym DL = N FB
TX i ic
TX i ic
N Sym SRB N SF DL + N FB
TX i ic
DwPTS
N TDD SSF N Sym SSF
Out of the total number of modulation symbols, Atoll then determines the numbers of modulation symbols corresponding to
each control channel as follows:
The number of modulation symbols for the downlink reference signals
The number of modulation symbols reserved for downlink reference signal transmission in one scheduler resource
block depends on the number of transmission antenna ports:
TX i ic
For all subframes except the TDD special subframes: N Res SRB
= 16
24
TX ic
i
if N Ant TX = 1
TX ic
i
if N Ant TX = 2
TX ic
i
For TDD special subframes:

Special
Subframe
Configuration

DwPTS
N SD SSF
10
11
TX i ic

TX i ic
N Ant TX
N Res DwPTS
DwPTS
N SD SSF
TX i ic
TX i ic
N Ant TX
N Res DwPTS
12
12
20
20
20
20
12
12
20
20
20
20
12
16
20
24
20
24
10
487

Special
Subframe
Configuration

TX ic
i
DwPTS
N SD SSF
12
10
11

TX ic
i
N Ant TX
N Res DwPTS
16
24
24
12
20
20
12
20
20
12
20
20
12
12
TX ic
i
DwPTS
N SD SSF
TX ic
i
N Ant TX
N Res DwPTS
12
20
20
12
20
20
12
12
This gives a number of reserved modulation symbols per frame:

TX i ic
TX i ic
TX i ic
N Sym Res = N SF DL N FB
TX i ic
TX i ic
TX i ic
N Res SRB + N TDD SSF N FB
TX i ic
N Res DwPTS
The number of modulation symbols used for downlink reference signal transmission in one scheduler resource block
is:
TX i ic
For all subframes except the TDD special subframes: N DLRS SRB
= 8
TX ic
i
if N Ant TX = 1
TX ic
i
if N Ant TX = 2
TX ic
i
For TDD special subframes:

Special
Subframe
Configuration
488

DwPTS
N SD SSF
TX i ic
TX i ic
N Ant TX
N DLRS DwPTS
DwPTS
N SD SSF
TX i ic
TX i ic
N Ant TX
N DLRS DwPTS

AT332_TRR_E0
Special
Subframe
Configuration

DwPTS
N SD SSF
10
11
12
10
11
TX ic
i
TX ic
i
N Ant TX
N DLRS DwPTS
DwPTS
N SD SSF
10
TX ic
i
TX ic
i
N Ant TX
N DLRS DwPTS
This gives a number of downlink reference signal modulation symbols per frame:
TX i ic
TX i ic
TX i ic
N Sym DLRS = N SF DL N FB
TX i ic
TX i ic
TX i ic
N DLRS SRB + N TDD SSF N FB
TX i ic
N DLRS DwPTS
The number of modulation symbols for the SS

The primary and secondary synchonisation signals are transmitted on 1 symbol duration each in the 1st and the 6th
downlink subframes, over the center 6 frequency blocks. Therefore,
N Sym PSS = 2 N FB SS PBCH N SCa FB = 144
N Sym SSS = 2 N FB SS PBCH N SCa FB = 144
489

And, N Sym SS = N Sym PSS + N Sym SSS = 288

The number of modulation symbols for the PBCH
The physical broadcast channel is transmitted on four symbol durations in the 1st downlink subframe over the center
6 frequency blocks. The physical broadcast channel overlaps with the downlink reference signals, therefore, some
modulation symbols reserved for downlink reference signals are subtracted:
216 for extended cyclic prefix
240 for normal cyclic prefix
The number of modulation symbols for the PDCCH
The physical downlink control channel can be transmitted over up to 4 symbol durations in each subframe. The
number of symbol durations for the PDCCH is defined in the global network settings. The physical downlink control
channel overlaps with the downlink reference signals, therefore, some modulation symbols reserved for downlink
reference signals are subtracted:
TX ic
i
if N SD PDCCH = 0 :
TX i ic
N Sym PDCCH = 0
TX ic
TX ic
i
i
TX ic
TX i ic
N Sym PDCCH =
TX ic
i
N i
TX ic
TX i ic
N SF DL
TX ic
i
i
+ N SD PDCCH N SCa FB 4 N FB
TX i ic
N TDD SSF
Otherwise:
TX ic
TX i ic
N Sym PDCCH =
TX ic
TX ic
i
i
N i

TX ic
TX ic
TX i ic
N SF DL
TX ic
i
i
i
+ Min 2 N SD PDCCH N SCa FB 2 Min 4 N Ant TX N FB
TX i ic
N TDD SSF
The number of modulation symbols for the PDSCH

The total number of modulation symbols in the frame excluding all the control channel modulation symbols gives the
number of modulation symbols available for user data, i.e., for the PDSCH:
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
N Sym PDSCH = N Sym DL N Sym Res N Sym SS N Sym PBCH N Sym PDCCH
The energy per resource element for 1 modulation symbol (dBm/Sym) of the downlink reference signals is calculated as
follows:
If the reference signal EPRE calculation method is set to Calculated (equal distribution of unused EPRE):
TX ic
TX i ic
EPRE DLRS
P i
Max
-------------------
TX i ic
TX i ic
DwPTS
10
= 10 Log 10
TX ic
TX ic
i
i
EPRE SS
EPRE PBCH
TX ic
--------------------------------------------------------------------i
10
10
+ N Sym PBCH 10
10 L og N Sym DLRS + N Sym SS 10
+ N Sym PDCCH 10
490
TX i ic
EPRE PDCCH
-----------------------------------10
+ N Sym PDSCH 10
TX i ic
EPRE PDSCH
------------------------------------
10
If the reference signal EPRE calculation method is set to Calculated (with boost) or Calculated (without boost):

AT332_TRR_E0
TX ic
TX ic
i
EPRE DLRS
P i
Max
-------------------
TX ic
TX ic
i
i
DwPTS
10
= 10 Log 10
TX ic
TX ic
i
i
EPRE SS
EPRE PBCH
TX ic
--------------------------------------------------------------------i
10
10
10 L og N Sym Res + N Sym SS 10
+ N Sym PBCH 10
+ N Sym PDCCH 10
TX i ic
EPRE
PDCCH
-----------------------------------10
+ N Sym PDSCH 10
TX i ic
EPRE
PDSCH
------------------------------------
10
Calculation of Other EPREs and Per-channel Powers

The energy per resource element for 1 modulation symbol (dBm/Sym) of the SS is calculated as follows:
TX i ic
EPRE SS
TX i ic
TX i ic
= EPRE DLRS + EPRE SS
The energy per resource element for 1 modulation symbol (dBm/Sym) of the PBCH is calculated as follows:
TX i ic
TX i ic
TX i ic
EPRE PBCH = EPRE DLRS + EPRE PBCH
The energy per resource element for 1 modulation symbol (dBm/Sym) of the PDCCH is calculated as follows:
TX i ic
TX i ic
TX i ic
EPRE PDCCH = EPRE DLRS + EPRE PDCCH

The energy per resource element for 1 modulation symbol (dBm/Sym) of the PDSCH is calculated as follows:
TX i ic
TX i ic
TX i ic
EPRE PDSCH = EPRE DLRS + EPRE PDSCH

If the reference signal EPRE calculation method is set to Calculated (with boost), the "boosted" RS energy per resource
element is calculated as follows:
TX i ic
EPRE DLRS
TX i ic
= EPRE DLRS
N TXi ic
Sym Res
+ 10 Log ------------------------TX i ic
N Sym DLRS
The instantaneous downlink reference signal transmission power is calculated as follows:

TX i ic
P DLRS
TX i ic
TX i ic
= EPRE DLRS + 10 Log 2 N FB
TX ic
i
Where 2 NFB
implies that at the instant when downlink reference signals are transmitted, they are transmitted using 2
subcarriers in each frequency block.

The instantaneous SS transmission power is calculated as follows:
TX i ic
P SS
TX i ic
= EPRE SS
The instantaneous PBCH transmission power is calculated as follows:

TX i ic
TX i ic
P PBCH = EPRE PBCH + 10 Log N SCa FB N FB SS PBCH

Where N SCa FB N FB SS PBCH implies that at the instant when the SS and the PBCH are transmitted, they are transmitted
using all the subcarriers in the centre 6 consecutive frequency blocks.
The average PDCCH transmission power is calculated as follows:
TX i ic
P PDCCH
TX i ic
N SymPDCCH
-------------------------------------------------------------------------------------------------------------------------------------= EPRE PDCCH + 10 Log
TX
ic
TX
ic
TX
ic
TX
ic
i
i
i
i
N
+
Min
2
N
SD PDCCH
SF DL
SD PDCCH
TDD SSF
TX i ic
The average PDSCH transmission power is calculated as follows:
491

TX ic
i
P PDSCH
TX ic
i
TX ic
N SymPDSCH
i
= EPRE PDSCH + 10 Log --------------------------------------------------------------------------------------------------------------------
TX ic
TX ic
i
N i
N
N
Slot SF
SD PDCCH
SF DL
SD Slot
TX i ic
TX i ic
DwPTS
+ N SD SSF Min 2 N SD PDCCH N TDD SSF
As the number of subcarriers used for the PDCCH and PDSCH transmission varies over time, i.e., from one symbol duration to
the next, the instantaneous powers of the PDCCH and the PDSCH also vary over time. This is why average transmission powers
are calculated and used in Atoll.
EPRE and Transmission Power adjustment for ICIC
The following applies to RS, PDCCH, and PDSCH EPREs for cells using downlink static ICIC.
1. No ICIC, time-switched FFR, and hard FFR
Cell-edge and cell-centre frequency blocks are not transmitted at the same time. Therefore,
TX i ic
TX i ic
EPRE DLRS CC = EPRE DLRS
TX ic
i
EPRE PDCCH CC
TX i ic
EPRE PDSCH CC
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
N FB
N FB
and EPRE DLRS CE = EPRE DLRS ---------------- ----------------TX i ic
TX i ic
N FB CC
N FB CE
TX ic
i
EPRE PDCCH
TX i ic
TX i ic
TX i ic
TX i ic
TX ic
TX ic
N FB
N FB
i
i
----------------and EPRE PDCCH CE = EPRE PDCCH ----------------TX i ic
TX i ic
N FB CC
N FB CE
TX i ic
TX i ic
TX i ic
N FB
N FB
and EPRE PDSCH CE = EPRE PDSCH ----------------= EPRE PDSCH ----------------TX i ic
TX i ic
N FB CC
N FB CE
TX i ic
TX i ic
P DLRS CC = P DLRS CE = P DLRS

TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
P PDCCH CC = P PDCCH CE = P PDCCH

P PDSCH CC = P PDSCH CE = P PDSCH
2. Soft and partial soft FFR
Cell-edge and cell-centre frequency blocks are transmitted at the same time; therefore, power is divided among cellcentre and cell-edge frequency blocks. Therefore, we have,
TX i ic
TX i ic
EPRE DLRS CC = EPRE DLRS
TX i ic
EPRE PDCCH CC
TX i ic
EPRE PDSCH CC
TX i ic
TX i ic
TX i ic
TX i ic
N FB
= EPRE PDCCH ---------------------------------------------------------------------- and EPRE PDCCH CE = EPRE PDCCH CC CE
TX ic
TX ic
TX i ic
i N i
FB CE + N FB CC
CE
492
TX i ic
TX i ic
TX i ic
TX i ic
N FB
= EPRE PDSCH ---------------------------------------------------------------------- and EPRE PDSCH CE = EPRE PDSCH CC CE
TX i ic
TX i ic
TX i ic
N FB CE + N FB CC
CE
TX i ic
TX i ic
TX i ic
P DLRS CC = P DLRS
TX ic
i
P PDCCH CC
TX i ic
TX i ic
TX i ic
TX i ic
N FB
---------------------------------------------------------------------- and EPRE DLRS CE = EPRE DLRS CC CE
TX ic
TX ic
TX i ic
i N i
FB CE + N FB CC
CE
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic EPRE
EPRE DLRS CC
DLRS CE
- and P DLRS
---------------------------- CE = P DLRS ----------------------------TX i ic
TX i ic
EPRE DLRS
EPRE DLRS
TX ic
i
P PDCCH
TX i ic
TX i ic
TX ic
TX ic
EPRE PDCCH CC
EPRE PDCCH CE
i
i
- and P PDCCH
-------------------------------- CE = P PDCCH --------------------------------TX i ic
TX i ic
EPRE PDCCH
EPRE PDCCH

AT332_TRR_E0
TX ic
i
P PDSCH CC
TX ic
i
In the above, CE
If
TX ic
i
CE
TX ic
i
P PDSCH
TX ic
i
TX ic
i
TX ic
TX ic
EPRE PDSCH CC
EPRE PDSCH CE
i
i
- and P PDSCH
------------------------------- CE = P PDSCH -------------------------------TX ic
TX ic
i
i
EPRE PDSCH
EPRE PDSCH
EPRE CE
is the cell-edge power boost for cell TXi(ic)s frame configuration. By definition: CE = ----------------EPRE CC
is left empty, it is automatically calculated as follows:
TX i ic
TX ic
i
CE
TX i ic
N FB CC
= ----------------TX i ic
N FB CE
TX i ic
N FB CC and N FB CE are respectively the numbers of frequency blocks in cell centre and cell-edge of TXi(ic).
Number of frequency blocks in
ICIC mode
Cell centre
Cell edge
TX i ic
No FFR
N FB
Time-switched FFR
N FB
TX i ic
TX i ic
N FB CEx
TX i ic
Soft FFR
TX i ic
TX i ic
N FB CEx
TX i ic
Hard FFR
Partial soft FFR
TX i ic
N FB
N FB
TX i ic
N FB
TX i ic
TX ic
N FB CEx
TX i ic
TX i ic
N FB CEx
TX ic
N FB CEx
TX ic
i
i
i
N FB CE0 + N FB CE1 + N FB CE2
TX i ic
TX i ic
N FB CEx
TX i ic
Where N FB CEx can be N FB CE0 , N FB CE1 , or N FB CE2 depending on the PSS ID of TXi(ic).
Output
TX ic
i
EPRE DLRS : Energy per resource element of the downlink reference signals for cell TXi(ic).
EPRE SS
EPRE PBCH : Energy per resource element of the PBCH for cell TXi(ic).
EPRE PDCCH : Energy per resource element of the PDCCH for cell TXi(ic).
EPRE PDSCH : Energy per resource element of the PDSCH for cell TXi(ic).
P DLRS : Instantaneous transmission power of the downlink reference signals for cell TXi(ic).
P SS
TX i ic
: Energy per resource element of the SS for cell TXi(ic).
TX i ic
TX i ic
TX i ic
TX ic
i
TX i ic
: Instantaneous transmission power of the SS for cell TXi(ic).
TX i ic
P PBCH : Instantaneous transmission power of the PBCH for cell TXi(ic).
P PDCCH : Average transmission power of the PDCCH for cell TXi(ic).
P PDSCH : Average transmission power of the PDSCH for cell TXi(ic).
TX i ic
TX i ic

An LTE network can consist of cells that use different channel bandwidths. Therefore, the start and end frequencies of all the
channels may not exactly coincide. Channel bandwidths of cells can overlap each other with different ratios.
493

Figure 6.2: Co-Channel and Adjacent Channel Overlaps

The following sections describe how the co- and adjacent channel overlaps are calculated between the channels used by any
studied cell TXi(ic) and any other cell TXj(jc) of the network. In terms of interference calculation, the studied cell can be
considered a victim of interference received from the other cells that might be interfering the studied cell.
TX ic
i
If the studied cell is assigned a channel number N Channel , it receives co-channel interference on the channel bandwidth of
TX i ic
TX i ic
N Channel , and adjacent channel interference on the adjacent channel bandwidths, i.e., corresponding to N Channel 1 and
TX i ic
N Channel + 1 .
In order to calculate the co- and adjacent channel overlaps between two channels, it is necessary to calculate the start and
end frequencies of both channels (explained in "Conversion From Channel Numbers to Start and End Frequencies" on
page 494). Once the start and end frequencies are known for the studied and other cells, the co- and adjacent overlaps and
the total overlap ratio are calculated as respectively explained in:
"Co-Channel Overlap Calculation" on page 495.

"Adjacent Channel Overlap Calculation" on page 496.
"Total Overlap Ratio Calculation" on page 496.
6.4.2.1 Conversion From Channel Numbers to Start and End Frequencies

Input
TX i ic
TX j jc
F Start Band and F Start Band : Start frequencies of the frequency bands assigned to the cells TXi(ic) and TXj(jc).
F Start Band can be the start frequency of a TDD frequency band ( F Start TDD ), or the uplink or the downlink start
frequency of an FDD frequency band ( F Start FDD UL or F Start FDD DL ).
First TX i ic
First TX j jc
N Channel
N Channel and N Channel : Channel numbers assigned to cells TXi(ic) and TXj(jc).
TX i ic
and N Channel
: First channel numbers the frequency band assigned to the cells TXi(ic) and TXj(jc).
TX j jc
For FDD networks, Atoll considers that the same channel number is assigned to a cell in the downlink and uplink, i.e.,
the channel number you assign to a cell is considered for uplink and downlink both.
494
TX i ic
TX j jc
W Channel and W Channel : Bandwidths of the channels assigned to cells TXi(ic) and TXj(jc).
ICS Band and ICS Band : Inter-channel spacing of the frequency bands assigned to cells TXi(ic) and TXj(jc).
CN Band and CN Band : Channel number step of the frequency bands assigned to cells TXi(ic) and TXj(jc).
TX i ic
TX i ic
TX j jc
TX j jc

AT332_TRR_E0
Calculations
Channel numbers are converted into start and end frequencies as follows:
For cell TXi(ic):
N TXi ic N First TX i ic
TX ic
TX ic
TX ic
TX ic
i
i
i
i
Channel
Channel
-
F Start = F Start Band + W Channel + ICS Band ------------------------------------------------------TX i ic
CN Band
TX i ic
F End
TX i ic
TX i ic
= F Start + W Channel
For cell TXj(jc):

TX j jc
F Start
TX j jc
F End
N TXj jc N First TX j jc
TX j jc
TX j jc
Channel
Channel
-
= F Start Band + W Channel + ICS Band ------------------------------------------------------TX j jc
CN Band
TX j jc
TX j jc
TX j jc
Output
TX ic
i
TX jc
j
F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc).
F End
TX i ic
TX j jc
and F End
: End frequencies for the cells TXi(ic) and TXj(jc).
6.4.2.2 Co-Channel Overlap Calculation

Input
TX i ic
TX j jc
F Start
and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
Numbers to Start and End Frequencies" on page 494.
TX i ic
TX j jc
F End
and F End
: End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
TX ic
i
W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic).
Calculations
Atoll first verifies that co-channel overlap exists between the cells TXi(ic) and TXj(jc).
Co-channel overlap exists if:
TX i ic
TX j jc
F Start F End
TX i ic
AND F End
TX j jc
F Start
Otherwise there is no co-channel overlap.

Atoll calculates the bandwidth of the co-channel overlap as follows:
TX i ic TX j jc
W CCO
TX jc
TX ic
TX jc
TX ic
j
i
j
i
= Min FEnd F End Max F Start F Start
The co-channel overlap ratio is given by:

TX i ic TX j jc
r CCO
TX i ic TX j jc
W CCO
= ---------------------------------TX j jc
W Channel
Output
TX i ic TX j jc
r CCO
: Co-channel overlap ratio between the cells TXi(ic) and TXj(jc).
495

6.4.2.3 Adjacent Channel Overlap Calculation

Input
TX ic
i
TX jc
j
F Start
TX i ic
TX j jc
F End
and F End
TX i ic
Calculations
Atoll first verifies that adjacent channel overlaps exist between (the lower-frequency and the higher-frequency adjacent
channels of) the cells TXi(ic) and TXj(jc).
Adjacent channel overlap exists on the lower-frequency adjacent channel if:
TX i ic
TX i ic
TX j jc
F Start W Channel F End
TX i ic
TX j jc
AND F Start F Start
Adjacent channel overlap exists on the higher-frequency adjacent channel if:

TX i ic
F End
TX j jc
F End
TX i ic
AND F End
TX i ic
TX j jc
+ W Channel F Start
Otherwise there is no adjacent channel overlap.

Atoll determines the adjacent channel overlap ratio as follows:
Bandwidth of the lower-frequency adjacent channel overlap:
TX i ic TX j jc
W ACO
TX jc
TX ic
TX jc
TX ic
TX ic
j
i
j
i
i
= Min F End F Start Max F Start F Start W Channel
The lower-frequency adjacent channel overlap ratio is given by:

TX i ic TX j jc
W ACO
L
= ---------------------------------TX j jc
W Channel
TX ic TX jc
i
j
r ACO
L
Bandwidth of the higher-frequency adjacent channel overlap:

TX i ic TX j jc
W ACO
TX jc
TX ic
j
i
= Min F End F End
TX ic
TX jc
TX ic
i
j
i
The higher-frequency adjacent channel overlap ratio is given by:

TX i ic TX j jc
W ACO
H
= ---------------------------------TX j jc
W Channel
TX i ic TX j jc
r ACO
The adjacent channel overlap ratio is given by:

TX i ic TX j jc
r ACO
TX i ic TX j jc
= r ACO
TX i ic TX j jc
+ r ACO
Output
TX ic TX jc
i
j
r ACO
: Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
6.4.2.4 Total Overlap Ratio Calculation

Input
TX i ic TX j jc
r CCO
: Co-channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co-Channel Overlap
496

AT332_TRR_E0
TX ic TX jc
i
j
r ACO
: Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Adjacent Channel
Overlap Calculation" on page 496.

TX ic
i
f ACS
TX ic
i
W Channel
: Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic).
TX jc
j
and W Channel : Bandwidths of the channels assigned to the cells TXi(ic) and TXj(jc).
Calculations
The total overlap ratio is:
TX ic TX jc
i
j
rO
TX i ic TX j jc
r CCO
TX i ic TX j jc
+ r ACO
10
TX i ic
f
ACS FB--------------------------10
TX i ic
TX j jc
TX i ic
TX j jc
if W Channel W Channel
TX ic
f ACS FB
TX i ic
TX ic TX jc TX ic TX jc
----------------------------
W Channel
10
j
i
j
r i
--------------------+ r ACO
10
TX j jc
CCO
W Channel
TX i ic
W Channel
The multiplicative factor --------------------is used to normalise the transmission power of the interfering cell TXj(jc). This means that
TX j jc
W Channel
TX j jc
TX j jc
if the interfering cell transmits at X dBm over a bandwidth of W Channel , and it interferes over a bandwidth less than W Channel ,
TX ic
i
W Channel
the interference from this cell should not be considered at X dBm but less than that. The factor --------------------converts X dBm over
TX jc
j
W Channel
TX j jc
TX j jc
W Channel to Y dBm (which is less than X dBm) over less than W Channel .
Output
TX i ic TX j jc
rO
: Total co- and adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
6.4.3 Subframe Pattern Collision Calculation

Subframe transmission and reception patterns can be defined for each cell using the Almost Blank Subframe (ABS) Pattern
field. The ABS pattern is a bit map, i.e., a series of 0s and 1s where each bit corresponds to one subframe. In an ABS pattern,
each 0 signifies a normal subframe and 1 implies an almost blank subframe. Almost blank subframes do not carry any traffic.
Only reference signals are transmitted over an ABS. For example, the ABS pattern "0100001000" means that subframes 1 and
6 are almost blank subframes whereas all the other subframes are normal subframes carrying traffic.
ABS patterns are used in conjunction with cell range expansion for eICIC (enhanced inter-cell interference coordination, also
known as time-domain ICIC) in an effort to minimise cell-edge interference between macro and small cells in heterogeneous
LTE networks (HetNets).
In order to calculate the collision between normal and almost blank subframes, the cells ABS patterns must be normalised,
and the used downlink, uplink, and special subframe patterns determined from the ABS patterns. These calculations are
respectively explained in:
"Subframe Pattern Normalisation" on page 497.

"Determination of Effective Subframe Patterns" on page 498.
"Calculation of Subframe Collision Probabilities" on page 499.
6.4.3.1 Subframe Pattern Normalisation

Prior to the calculation of subframe collision probabilities, Atoll normalises the different ABS pattern lengths in order to
perform logical (bit by bit) AND and OR operations afterwards.
497

ABS patterns are normalised as follows:

1. The standard length of the ABS pattern of a cell is determined from its frequency bands duplexing method and, in
case the cells frequency band is a TDD band, from the cells TDD frame configuration. The standard lengths of the ABS
pattern bit maps as defined by the 3GPP are as follows:
FDD cells: 40 bits

TDD cells using the frame configuration 0: 70 bits
TDD cells using the frame configuration 1 through 5: 20 bits
TDD cells using the frame configuration 6: 60 bits
The normalised ABS pattern length used in Atoll is 80 bits, which covers all the standard lengths. ABS patterns of
different lengths are normalised to 80 bits by Atoll.
2. The ABS pattern is corrected to match the standard ABS pattern lengths determined in step 1.:
If the ABS pattern contains an asterisk, the pattern of 0s and 1s leading the asterisk is cyclically repeated until it
matches the standard ABS pattern length. Any 0s and 1s entered after an asterisk will be ignored.
FDD example: 0100010000* = 0100010000010001000001000100000100010000
If the ABS pattern is shorter than the standard ABS pattern length, it is filled with 0s to match the standard ABS
pattern length.
FDD example: 01000100000100010000 = 0100010000010001000000000000000000000000
If the ABS pattern is longer than the standard ABS pattern length, it is truncated to match the standard ABS pattern
length.
FDD example: 01000100000100010000010001000001000100000111110000 =
0100010000010001000001000100000100010000
If the ABS pattern is empty, it means that there are no almost blank subframes defined and all the subframes can
carry traffic.
FDD example: NULL = 0 = 0* = 0000000000000000000000000000000000000000
3. The ABS pattern determined in step 2. is resized to 80 bits. More precisely, the ABS pattern is concatenated with itself.
Examples:
FDD:
0100010000010001000001000100000100010000 =
01000100000100010000010001000001000100000100010000010001000001000100000100010000
TDD frame configuration 0:

0100010000010001000001000100000100010000010001000001000100000100010000 =
01000100000100010000010001000001000100000100010000010001000001000100000100010000
TDD frame configurations 1 through 5:

01000100000100010000 =
01000100000100010000010001000001000100000100010000010001000001000100000100010000
TDD frame configuration 6:

010001000001000100000100010000010001000001000100000100010000 =
01000100000100010000010001000001000100000100010000010001000001000100000100010000
Once the ABS pattern SFP ABS has been normalised, it is inverted to determine the used subframe pattern SFP Used that is
used in further calculations:
SFP Used = !SFP ABS
6.4.3.2 Determination of Effective Subframe Patterns

Effective downlink, uplink, and special subframe patterns ( SFP DL , SFP UL , and SFP SSF ) are determined as follows by applying
masks ( SFM DL , SFM UL , and SFM SSF ) to the normalised used subframe patterns SFP Used determined as explained in
"Subframe Pattern Normalisation" on page 497:
SFP DL = SFP Used AND SFM DL
SFP UL = SFP Used AND SFM UL
SFP SSF = SFP Used AND SFM SSF
498

AT332_TRR_E0
SFM DL , SFM UL , and SFM SSF are, respectively, the downlink, uplink and special subframe masks listed below:
FDD
SFM DL
SFM UL
11111111111111111111111111111111111111111111111111111111111111111111111111111111
SFM SSF
TDD frame configuration 0-DSUUU DSUUU
SFM DL
10000100001000010000100001000010000100001000010000100001000010000100001000010000
SFM UL
00111001110011100111001110011100111001110011100111001110011100111001110011100111
SFM SSF
01000010000100001000010000100001000010000100001000010000100001000010000100001000
TDD frame configuration 1-DSUUD DSUUD

SFM DL
10001100011000110001100011000110001100011000110001100011000110001100011000110001
SFM UL
00110001100011000110001100011000110001100011000110001100011000110001100011000110
SFM SSF
01000010000100001000010000100001000010000100001000010000100001000010000100001000
TDD frame configuration 2-DSUDD DSUDD

SFM DL
10011100111001110011100111001110011100111001110011100111001110011100111001110011
SFM UL
00100001000010000100001000010000100001000010000100001000010000100001000010000100
SFM SSF
01000010000100001000010000100001000010000100001000010000100001000010000100001000
TDD frame configuration 3-DSUUU DDDDD

SFM DL
10000111111000011111100001111110000111111000011111100001111110000111111000011111
SFM UL
00111000000011100000001110000000111000000011100000001110000000111000000011100000
SFM SSF
01000000000100000000010000000001000000000100000000010000000001000000000100000000
TDD frame configuration 4-DSUUD DDDDD

SFM DL
10001111111000111111100011111110001111111000111111100011111110001111111000111111
SFM UL
00110000000011000000001100000000110000000011000000001100000000110000000011000000
SFM SSF
01000000000100000000010000000001000000000100000000010000000001000000000100000000
TDD frame configuration 5-DSUDD DDDDD

SFM DL
10011111111001111111100111111110011111111001111111100111111110011111111001111111
SFM UL
00100000000010000000001000000000100000000010000000001000000000100000000010000000
SFM SSF
01000000000100000000010000000001000000000100000000010000000001000000000100000000
TDD frame configuration 6-DSUUU DSUUD

SFM DL
10000100011000010001100001000110000100011000010001100001000110000100011000010001
SFM UL
00111001100011100110001110011000111001100011100110001110011000111001100011100110
SFM SSF
01000010000100001000010000100001000010000100001000010000100001000010000100001000
6.4.3.3 Calculation of Subframe Collision Probabilities

The probabilities of collision of subframes between a studied cell TXi(ic) and any interfering cell TXj(jc) are calculated as
follows.
499

In the following equations, the operator
implies the sum of 1s in a given
series of bits, X.
In the following equations, AND and OR are logical bit-by-bit operators.
Method 1: ABS Patterns Used Only at Cell Edges

By default, ABS patterns are considered only to be used for serving users at cell edges. This means that all subframes are
considered non-ABS subframes in the cell centre. This is equivalent to setting the following Atoll.ini option:
[LTE]
UseABSonCellEdgeOnly = 1
Different collision probabilities are calculated depending on the location of the served pixel, subscriber, or mobile in cell
TXi(ic):
Subframe collision between cell centre of TXi(ic) and cell centre of TXj(jc):
TX i ic TX j jc
p ABS DL CC
j
AND SFM j
OR SFMSSF
DL
1
= -----------------------------------------------------------------------------------------------------------------------------------------------------------------------TX ic
TX ic
i
SFM i
OR SFM SSF
DL
TX i ic
SFMDL
TX i ic
OR SFM SSF
TX jc
TX jc
TX i ic TX j jc
p ABS UL CC
TX jc
TX jc
j
j
OR SFMSSF
AND SFM UL
1
= -----------------------------------------------------------------------------------------------------------------------------------------------------------------------TX i ic
TX i ic
SFM
OR
SFM
UL
SSF
TX ic
i
SFMUL
TX ic
i
OR SFM SSF
Subframe collision between cell edge of TXi(ic) and cell centre of TXj(jc):
TX i ic TX j jc
p ABS DL CC
j
AND SFM j
OR SFM SSF
DL
1
= -------------------------------------------------------------------------------------------------------------------------------------------------------------------TX i ic
TX i ic
SFP
OR SFPSSF
DL
TX i ic
SFPDL
TX i ic
OR SFPSSF
TX jc
TX jc
TX i ic TX j jc
p ABS UL CC
j
AND SFM j
OR SFM SSF
UL
1
= -------------------------------------------------------------------------------------------------------------------------------------------------------------------TX ic
TX i ic
SFP i
OR SFPSSF
UL
TX i ic
SFPUL
TX i ic
OR SFPSSF
TX jc
TX jc
Subframe collision between cell centre of TXi(ic) and cell edge of TXj(jc):
TX i ic TX j jc
p ABS DL CE
j
AND SFP j
OR SFP SSF
DL
1
= -------------------------------------------------------------------------------------------------------------------------------------------------------------------TX i ic
TX i ic
SFM
OR SFMSSF
DL
TX i ic
SFMDL
TX i ic
OR SFM SSF
TX jc
TX jc
TX i ic TX j jc
p ABS UL CE
j
AND SFP j
OR SFP SSF
UL
1
= -------------------------------------------------------------------------------------------------------------------------------------------------------------------TX ic
TX i ic
SFM i
OR SFMSSF
UL
TX i ic
SFMUL
TX i ic
OR SFM SSF
TX jc
TX jc
Subframe collision between cell edge of TXi(ic) and cell edge of TXj(jc):
TX i ic TX j jc
p ABS DL CE
j
AND SFP j
OR SFP SSF
DL
1
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------TX i ic
TX i ic
SFP
OR
SFP
DL
SSF
TX i ic
SFPDL
TX i ic
OR SFPSSF
500
TX jc
TX jc

AT332_TRR_E0
TX ic TX jc
i
j
p ABS UL CE
TX jc
TX jc
j
j
OR SFP SSF
AND SFP UL
1
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------TX ic
TX ic
i
SFP i
OR SFP SSF
UL
TX ic
i
TX ic
i
SFPUL
OR SFPSSF
This method enables you to include the cell-edge traffic ratio in the calculation of interference. The downlink interference
reduction factor due to subframe collisions for any served pixel, subscriber, or mobile in cell TXi(ic) is calculated as follows:
TX i ic TX j jc
f ABS DL
TX jc
TX ic TX jc
TX jc
TX ic TX jc
j
i
j
j
i
j
= 10 Log r DL CE p ABS DL CE + 1 r DL CE p ABS DL CC
The uplink interference reduction factor due to subframe collisions for any served pixel, subscriber, or mobile in cell TXi(ic) is
TX i ic TX j jc
f ABS UL
TX ic TX jc
TX ic TX j jc
i
j
i
= 10 Log p ABS UL CE or f ABS UL
TX ic TX jc
i
j
= 10 Log p ABS UL CC
Method 2: ABS Patterns Used Throughout the Cell

If you wish to apply the ABS patterns throughout the cell, irrespective of the cell-edge area and the cell-edge traffic ratio, you
can do so by adding the following lines in the Atoll.ini file:
[LTE]
UseABSonCellEdgeOnly = 0
The following collision probabilities are calculated between cells TXi(ic) and TXj(jc):
TX i ic TX j jc
p ABS DL
j
AND SFP j
OR SFPSSF
DL
1
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------TX ic
TX i ic
SFP i
OR SFP SSF
DL
TX i ic
SFPDL
TX i ic
OR SFP SSF
TX jc
TX jc
TX i ic TX j jc
p ABS UL
j
AND SFP j
OR SFPSSF
UL
1
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------TX ic
TX i ic
SFP i
OR SFP SSF
UL
TX i ic
SFPUL
TX i ic
OR SFP SSF
TX jc
TX jc
The downlink interference reduction factor due to subframe collisions for any pixel, subscriber, or mobile is calculated as
follows:
TX i ic TX j jc
f ABS DL
TX i ic TX j jc
= 10 Log p ABS DL
The uplink interference reduction factor due to subframe collisions for any pixel, subscriber, or mobile is calculated as follows:
TX i ic TX j jc
f ABS UL
TX ic TX j jc
i
= 10 Log p ABS UL
6.4.4 Signal Level and Signal Quality Calculations

These calculations include the calculation of the received signal levels, and noise and interference. The following sections
describe how the received signal levels, the noise and interference, C/N, and C/(I+N) ratios are calculated in Atoll:

"Noise Calculation (DL)" on page 507.
"Interference Calculation (DL)" on page 507.
"Noise Calculation (UL)" on page 527.
"Interference Calculation (UL)" on page 527.
"Noise Rise Calculation (UL)" on page 530.
501

6.4.4.1 Signal Level Calculation (DL)

Input
TX ic
i
P Max : Max power of the cell TXi(ic).
P DLRS : Transmission power of the downlink reference signals for cell TXi(ic) as calculated in "Downlink Transmission
TX i ic
Power Calculation" on page 485.
TX i ic
P SS
: Transmission power of the SS for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX i ic
P PBCH : Transmission power of the PBCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation"
on page 485.
TX i ic
P PDCCH : Transmission power of the PDCCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation"
on page 485.
TX i ic
P PDSCH : Transmission power of the PDSCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation"
on page 485.
TX i ic
EPRE DLRS : Energy per resource element of the downlink reference signals for cell TXi(ic) as calculated in "Downlink
Transmission Power Calculation" on page 485.
TX i ic
EPRE SS
: Energy per resource element of the SS for cell TXi(ic) as calculated in "Downlink Transmission Power
TX i ic
EPRE PBCH : Energy per resource element of the PBCH for cell TXi(ic) as calculated in "Downlink Transmission Power
TX ic
i
EPRE PDCCH : Energy per resource element of the PDCCH for cell TXi(ic) as calculated in "Downlink Transmission Power
TX ic
i
EPRE PDSCH : Energy per resource element of the PDSCH for cell TXi(ic) as calculated in "Downlink Transmission Power
TX
E SA : Number of antenna elements defined for the smart antenna equipment used by the transmitter TXi.
G SA
Combining
G SA
G SA : Smart antenna diversity gain (for cross-polarised smart antennas) defined per clutter class.
G Ant : Transmitter antenna gain for the antenna used by the transmitter TXi.
G SA : Smart antenna gain in the direction of the served pixel, subscriber, or mobile Mi. For more information on
Array
: Smart antenna array gain offset defined per clutter class.

: Smart power combining gain offset defined per clutter class.
Div
TX i
TX i
TX i
the calculation of G SA , see "Beamforming Smart Antenna Models" on page 43.

TX i
Array
The smart antenna gain ( G SA ) and the smart antenna array gain offset ( G SA
) are
applied only if the AAS criterion (RS C/N, RS C/(I+N), or PDSCH C/(I+N)) is less than the
DL
AAS threshold ( T AAS ) defined in the properties of the reception equipment used by the
pixel, subscriber, or mobile Mi.
502
TX
: Total transmitter losses for the transmitter TXi ( L
TX
= L Total DL ).
L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model.
L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi.
M Shadowing Model : Shadowing margin based on the model standard deviation.
TX i

AT332_TRR_E0
In coverage predictions, shadowing margins are taken into account when the option "Shadowing taken into account"
is selected. If the terminal used for the coverage prediction supports CoMP and the serving cell being studied at any
pixel belongs to a CoMP set than uses downlink dynamic point selection or coherent joint transmission, the shadowing
MD
margin is adjusted by the macro-diversity gain ( G CoMP ). The adjusted shadowing margin becomes:
MD
M Shadowing Model = M Shadowing Model + G CoMP

For more information, see "Shadow Fading Model" on page 90.
L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected.
Mi
Mi
: Receiver terminal losses for the pixel, subscriber, or mobile Mi.

: Receiver terminals antenna gain for the pixel, subscriber, or mobile Mi.
Mi
L Ant : Receiver terminals antenna attenuation calculated for the pixel, subscriber, or mobile Mi.
M
For calculating the useful signal level from the best serving cell, L Ant is determined in the direction (H,V) = (0,0) from
Mi
the antenna patterns of the antenna used by Mi. For calculating the interfering signal level from any interferer, L Ant
is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi, while the
antenna is pointed towards Mis best serving cell.
L Body : Body loss defined for the service used by the pixel, subscriber, or mobile Mi.
Mi
, G
Mi
Mi
Mi
, L Ant , and L Body are not used in the calculations performed for the point
analysis tools profile tab, delta path loss calculation, and the downlink reference signal
level based coverage predictions.
TX i ic
D CP
: Cyclic prefix duration defined in the TXi(ic) frame configuration or, otherwise, in the global network settings.
Calculations
TX ic
i
C Max
TX ic
i
= EIRP Max L Path M Shadowing Model L Indoor + G

TX ic
i
Without smart antennas: EIRP Max

TX i ic
With smart antennas: EIRP Max

TX i ic
TX ic
i
TX
TX i ic
TX i
TX i
TX i ic
C DLRS = EIRP1 DLRS L Path M Shadowing Model L Indoor + G

TX i ic
Without smart antennas: EIRP1 DLRS

TX ic
i
TX i ic
C SS
TX i ic
= EIRP1 SS
TX ic
i
TX i ic
With smart antennas: EIRP1 SS

TX i ic
TX
TX i ic
= P SS
TX i ic
= P SS
Mi
TX
Mi
TX i
TX i
TX i
C PBCH = EIRP1 PBCH L Path M Shadowing Model L Indoor + G
L Ant L Body + f CP
TX i
Mi
Mi
Mi
Combining
Div
+ G SA
Mi
L Ant L Body + f CP
TX i
TX
i
Combining
+ G Ant L
+ G Ant L
TX i
= P DLRS + G Ant L
TX i ic
TX i ic
TX i ic
TX
= P DLRS + G Ant L
Without smart antennas: EIRP1 SS
= P Max + G Ant L
= P Max + G Ant L
With smart antennas: EIRP1 DLRS
Mi
Mi
Mi
L Ant L Body + f CP
TX i
TX i
Mi
Mi
Combining
Div
+ G SA
Mi
L Ant L Body + f CP
503


TX ic
i
TX ic
i
TX
Without smart antennas: EIRP1 PBCH = P PBCH + G Ant L

TX ic
i
TX ic
i
TX
With smart antennas: EIRP1 PBCH = P PBCH + G Ant L

TX i ic
TX
TX i ic
C PDCCH = EIRP1 PDCCH L Path M Shadowing Model L Indoor + G

TX i ic
TX i ic
TX
TX
i
Combining
Div
+ G SA
Mi
TX i
Without smart antennas: EIRP1 PDCCH = P PDCCH + G Ant L

TX i ic
TX i ic
TX i
With smart antennas: EIRP1 PDCCH = P PDCCH + G Ant L

TX i ic
TX i
TX i ic
C PDSCH = EIRP1 PDSCH L Path M Shadowing Model L Indoor + G

TX i ic
TX i ic
Mi
TX i ic
Mi
Mi
Mi
L Ant L Body + f CP
TX i
TX
i
Combining
Div
+ G SA
TX i
Without smart antennas: EIRP1 PDSCH = P PDSCH + G Ant L

TX i ic
Mi
Mi
Mi
L Ant L Body + f CP
TX i
TX i
Array
With smart antennas: EIRP1 PDSCH = P PDSCH + G SA + G SA
Combining
+ G SA
Div
+ G SA L
TX i
The energy per resource element (dBm/Sym) received from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi
as follows:
TX i ic
RSRP: E DLRS
TX i ic
= EIRP2 DLRS L Path M Shadowing Model L Indoor + G

TX i ic
Without smart antennas: EIRP2 DLRS

TX i ic
With smart antennas: EIRP2 DLRS

TX i ic
E SS
TX i ic
= EIRP2 SS
TX i ic
TX i
= EPRE DLRS + G Ant L

TX i ic
TX i
= EPRE DLRS + G Ant L

TX i ic
Without smart antennas: EIRP2 SS
Mi
TX i ic
= EPRE SS
Mi
TX i
Mi
Mi
Mi
L Ant L Body + f CP
TX i
TX
i
Combining
Mi
TX i
+ G Ant L
Mi
Mi
L Ant L Body + f CP
TX i

TX i ic
Server: EIRP2 SS
Interferer: EIRP2 SS
TX i ic
TX i ic
= EPRE SS
TX i ic
TX i
+ G Ant L
TX i ic
= EPRE SS
TX i
TX
i
Combining
Div
+ G SA
TX i
+ G Ant L
TX i ic
E PBCH = EIRP2 PBCH L Path M Shadowing Model L Indoor + G

TX i ic
TX
TX i
Mi
TX i ic
i
Combining
Mi
TX i
Without smart antennas: EIRP2 PBCH = EPRE PBCH + G Ant L
Mi
Mi
L Ant L Body + f CP
TX i

TX i ic
TX i ic
TX i
TX i
Server: EIRP2 PBCH = EPRE PBCH + G Ant L
Interferer: EIRP2 PBCH = EPRE PBCH + G Ant L
TX i ic
TX i ic
TX i ic
TX i
TX i ic
TX i
TX i
E PDCCH = EIRP2 PDCCH L Path M Shadowing Model L Indoor + G

TX i ic
TX i ic
Mi
Combining
TX i
Div
+ G SA
TX i
Mi
Without smart antennas: EIRP2 PDCCH = EPRE PDCCH + G Ant L
Mi
Combining
Mi
L Ant L Body + f CP
TX i
504
TX i ic
TX i ic
TX i
Server: EIRP2 PDCCH = EPRE PDCCH + G Ant L
TX i
TX
i
Combining
Div
+ G SA

AT332_TRR_E0
TX ic
i
TX ic
i
TX ic
i
TX
Interferer: EIRP2 PDCCH = EPRE PDCCH + G Ant L

TX ic
i
E PDSCH = EIRP2 PDSCH L Path M Shadowing Model L Indoor + G

TX ic
i
TX ic
i
TX
TX
TX
Without smart antennas: EIRP2 PDSCH = EPRE PDSCH + G Ant L
Combining
L Ant L Body + f CP
TX

TX i ic
TX i ic
TX i
Array
Server: EIRP2 PDSCH = EPRE PDSCH + G SA + G SA
Interferer: EIRP2 PDSCH = EPRE PDSCH + G SA + G SA
TX i ic
TX i ic
TX i
Combining
+ G SA
Array
Div
+ G SA L
Combining
+ G SA
TX i
TX i
In the above, L Path is the path loss (dB) calculated as follows:

TX i
L Path = L Model + L Ant

Furthermore, the total losses between the cell and the pixel, subscriber, or mobile Mi can be calculated as follows:
L Total = L Path + L
TX i
TX i
+L
Mi
Mi
Mi
Mi
+ L Ant + L Body
f CP is the cyclic prefix factor, i.e., the ratio of the useful symbol energy to the total symbol energy.
The total symbol duration of a modulation symbol comprises the useful symbol duration, carrying the actual data bits, and a
cyclic prefix, added to the useful data bits as padding against multi-path to avoid inter-symbol interference. Hence, the total
energy within a modulation symbol belongs in part to the useful data bits and in part to the cyclic prefix. Once a modulation
symbol is received, only the energy of the useful data bits can be used for extracting the data. The energy belonging to the
cyclic prefix is lost once it has served its purpose of combatting inter-symbol interference. Therefore, f CP implies that the
energy belonging to the cyclic prefix is excluded from the useful signal level.
f CP
TX ic
10 Log 7 7,5 If D CPi = Normal
TX i ic
=
= Extended
0
If TX i ic is an interferer
The cyclic prefix energy and the useful data bits energy are both taken into account when calculating interfering signal levels.
For downlink coherent joint transmission CoMP, the resulting signal level at any pixel, subscriber, or mobile Mi is the sum of
signal levels from all the coordinated servers:
TX ic
CoMP JTC
C Max
C i
Max -
-------------------10
10
= 10 Log
TXi ic
TX ic
CoMP JTC
C DLRS
TX ic
C i
E i
DLRS
DLRS
-------------------
-------------------CoMP
JTC
10
10
10
10
and E DLRS
= 10 Log
= 10 Log
TXi ic
TXi ic
TX ic
CoMP JTC
C SS
TX ic
C i
E i
SS
SS
-------------------
-------------------
-
CoMP
JTC
10
10
and E SS
10
10
= 10 Log
= 10 Log
TXi ic
TXi ic
TX ic
CoMP JTC
C PBCH
TX ic
C i
E i
PBCH -
PBCH
-------------------
-------------------CoMP
JTC
10
10
and E PBCH
= 10 Log
= 10 Log
10
10
TXi ic
TXi ic
505


TX ic
CoMP JTC
C PDCCH
TX ic
C i
E i
PDCCH
PDCCH
--------------------
--------------------CoMP JTC
10
10
10
and E PDCCH
10
= 10 Log
= 10 Log
TX i ic
TXi ic
TX ic
CoMP JTC
C PDSCH
TX ic
C i
E i
PDSCH
PDSCH
-------------------
-------------------CoMP JTC
10
10
10
10
and E PDSCH
= 10 Log
= 10 Log
TX i ic
TXi ic
Output
TX ic
i
C Max : Received max signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
C DLRS : Received downlink reference signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
C SS
TX i ic
TX i ic
: Received SS signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i ic
C PBCH : Received PBCH signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
C PDCCH : Received PDCCH signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
C PDSCH : Received PDSCH signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
E DLRS : Received downlink reference signal energy per resource element from the cell TXi(ic) at the pixel, subscriber,
TX i ic
TX ic
i
TX i ic
or mobile Mi.
TX ic
i
E SS
: Received SS energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i ic
E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
E PDSCH : Received PDSCH eneregy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
L Path : Path loss between the cell TXi(ic) and the pixel, subscriber, or mobile Mi.
L Total : Total losses between the cell TXi(ic) and the pixel, subscriber, or mobile Mi.
C Max
TX i ic
TX i ic
CoMP JTC
: Received max signal level from all the downlink coherent joint transmission servers TXi(ic) at the pixel,
subscriber, or mobile Mi.
CoMP JTC
C DLRS
: Received downlink reference signal level from all the downlink coherent joint transmission servers TXi(ic)
at the pixel, subscriber, or mobile Mi.
CoMP JTC
C SS
: Received SS signal level from all the downlink coherent joint transmission servers TXi(ic) at the pixel,
CoMP JTC
C PBCH
: Received PBCH signal level from all the downlink coherent joint transmission servers TXi(ic) at the pixel,
CoMP JTC
C PDCCH
: Received PDCCH signal level from all the downlink coherent joint transmission servers TXi(ic) at the pixel,
CoMP JTC
C PDSCH
: Received PDSCH signal level from all the downlink coherent joint transmission servers TXi(ic) at the pixel,
CoMP JTC
E DLRS
: Received downlink reference signal energy per resource element from all the downlink coherent joint
transmission servers TXi(ic) at the pixel, subscriber, or mobile Mi.
CoMP JTC
E SS
: Received SS energy per resource element from all the downlink coherent joint transmission servers TXi(ic)
at the pixel, subscriber, or mobile Mi.
CoMP JTC
E PBCH
: Received PBCH energy per resource element from all the downlink coherent joint transmission servers
TXi(ic) at the pixel, subscriber, or mobile Mi.
506

AT332_TRR_E0
CoMP JTC
E PDCCH
: Received PDCCH energy per resource element from all the downlink coherent joint transmission servers
CoMP JTC
E PDSCH
: Received PDSCH energy per resource element from all the downlink coherent joint transmission servers
6.4.4.2 Noise Calculation (DL)

For determining the C/N and C/(I+N), Atoll calculates the downlink noise which comprises thermal noise and the noise figure
of the equipment. The thermal noise density depends on the temperature, i.e., it remains constant for a given temperature.
However, the value of the thermal noise varies with the used bandwidth.
Input
K: Boltzmanns constant.
T: Temperature in Kelvin.
nf
Mi
: Noise figure of the terminal used for calculations by the pixel, subscriber, or mobile Mi.
Calculations
The power spectral density of thermal noise is calculated as follows:
n 0 = 10 Log K T 1000 = 174 dBm/Hz
The thermal noise for one resource element, i.e., over one subcarrier, is calculated as follows:
TX i ic
n 0 Sym = n 0 + 10 Log F
The downlink noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel,
subscriber, or mobile Mi. The downlink noise for one resource element, i.e., over one subcarrier, is calculated as follows:
TX i ic
n Sym
TX i ic
= n 0 Sym + nf
Mi
Output
TX i ic
n Sym
: Downlink noise for one subcarrier.
6.4.4.3 Interference Calculation (DL)

The interference received by any pixel, subscriber, or mobile, served by a cell TXi(ic) from other cells TXj(jc) can be defined as
the signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the channels used by the
cells TXi(ic) and TXj(jc), on the traffic loads of the interfering cells TXj(jc), and whether the cells support ICIC or not.
Input
TX j jc
E DLRS : Received downlink reference energy per resource element received from any interfering cell TXj(jc) as
calculated in "Signal Level Calculation (DL)" on page 502 at the pixel, subscriber, or mobile Mi covered by the cell
TXi(ic).
E SS
TX jc
j
: Received SS energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal
Level Calculation (DL)" on page 502 at the pixel, subscriber, or mobile Mi covered by the cell TXi(ic).
TX j jc
E PBCH : Received PBCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal
TX jc
j
E PDCCH : Received PDCCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal
TX j jc
E PDSCH : Received PDSCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal
507

Interfering cells TXj(jc) do not include any downlink coherent joint transmission CoMP
servers.
TX
G SA : Smart antenna gain in the direction . For more information, see "Beamforming Smart Antenna Models" on
page 43.
TX j
G SA : Smart antenna gain in the direction calculated from the average array correlation matrix:
H
G SA = g n S R Avg S . For more information, see "Beamforming Smart Antenna Models" on page 43.
M Shadowing C I : Shadowing margin based on the C/I standard deviation.

In Monte Carlo simulations, the received energies per resource element from interferers already include
M Shadowing Model , as explained in "Signal Level Calculation (DL)" on page 502.
In coverage predictions, the ratio M Shadowing Model M Shadowing C I is applied to the interfering signals (for more
information, see "Shadow Fading Model" on page 90). As the received energies per resource element from interferers
MD
MD
already include M Shadowing Model + G CoMP , M Shadowing C I + G CoMP is added to the received energies per resource
element from interferers in order to achieve the ratio M Shadowing Model M Shadowing C I :
E
TX j jc
= E
TX j jc
DPS
+ M Shadowing C I + G CoMP
is selected.
TX j jc
N Sym DLRS : Number of downlink reference signal resource elements as calculated in "Downlink Transmission Power

N Sym SS : Number of SS resource elements as calculated in "Downlink Transmission Power Calculation" on page 485.
N Sym PBCH : Number of PBCH resource elements as calculated in "Downlink Transmission Power Calculation" on
TX j jc
page 485.
TX j jc
N Sym PDCCH : Number of PDCCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX j jc
N Sym PDSCH : Number of PDSCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX j jc
N Sym DL : Total number of downlink resource elements as calculated in "Downlink Transmission Power Calculation"
on page 485.
TX ic TX jc
i
j
rO
: Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent
Channel Overlaps Calculation" on page 493.
TL DL
TX jc
j
: Downlink traffic load of the interfering cell TXj(jc).
Traffic loads can either be calculated using Monte Carlo simulations, or entered manually for each cell. Calculation of
traffic loads is explained in "Simulation Process" on page 479.
TX jc
j
AU DL
W FB : Width of a frequency block in the frequency domain (180 kHz).
N FB
TX i ic
: Downlink AAS usage of the interfering cell TXj(jc).
TX j jc
and N FB
TX i ic
TX j jc
508

AT332_TRR_E0
TX ic
i
TX jc
j
TX ic
i
TX jc
j
TX i ic
TX j jc
F Start and F Start : Start frequencies of the channels assigned to the cells TXi(ic) and TXj(jc) calculated as explained
in "Conversion From Channel Numbers to Start and End Frequencies" on page 494.
TX i ic
TX j jc
ID
TX ic
i
r DL CE
N Ant TX : Number of transmission (downlink) antenna ports defined for the cell TXj(jc).
TX i ic
TX j jc
and ID
and
TX jc
j
r DL CE :
: Physical cell IDs of the cells TXi(ic) and TXj(jc).
Downlink cell-edge traffic ratios of the cells TXi(ic) and TXj(jc).
TX i ic
TX j jc
Calculations
Two interference calculation methods exist in Atoll.
TX j jc
TX j jc
Calculations of f MIMO , f TL
TX i ic TX j jc
, f ICIC DL
TX i ic TX j jc
, f PDCCH
TX i ic TX j jc
, and f PDSCH
are
explained at the end of this section.
TX j jc
In the calculations below, E DLRS

probability
TX ic TX jc
i
j
f ABS DL
is weighted by the downlink subframe collision
when the relevant option is set in the Atoll.ini file:
[LTE]
eICIConRS = 1
Method 1: Synchronised Transmission and Reception
Atoll calculates the interference between two cells using this method when:
The frequency channels assigned to the interfered and interfering cells have the same centre frequency, and
The interfered and interfering cells both have an even number of frequency blocks or both have an odd number of
frequency blocks, and
The Atoll.ini file does not contain the following option:
[LTE]
SameItf_PDSCH_RS_PDCCH = 1
Synchronised transmission and reception means that the OFDM symbols of the interfered and interfering frames overlap and
match each other in time.
The interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell TXj(jc) at a
pixel, subscriber, or mobile Mi is calculated as follows:
RS of the interfered cell TXi(ic) collide only with RS of the interfering cell TXj(jc)
TX i ic
This occurs when ID PSS
TX j jc
= ID PSS
TX j jc
TX i ic
and N Ant TX N Ant TX
For the calculation of the probability of collision, here N Ant TX = Min 4 NAnt TX .
TX jc
TX j jc
DLRS
j
TX ic
E DLRS
N i
---------------------
TX i ic TX j jc
10
Ant TX
= 10 Log ------------------ 10
+ fO
TX jc
j
N Ant TX
RS of the interfered cell TXi(ic) collide with RS, PDCCH, and PDSCH of the interfering cell TXj(jc)
509

TX ic
i
TX jc
j
= ID PSS
TX jc
j
TX ic
i
For the calculation of the probability of collision, here N Ant TX = Min 4 N Ant TX .
With 1 or 2 antenna ports:

TX jc
TX j jc
DLRS
j
TX jc
E DLRS
N j
---------------------
TX ic TX j jc
10
Ant
TX
+f i
= 10 Log ------------------ 10
O
TX i ic
N Ant TX
TX jc
TX ic TX jc
TX jc
TX ic TX jc
TX jc
TX ic TX jc
TX jc
TX ic TX jc
j
i
j
j
i
j
E PDCCH + f PDCCH
E PDSCH + f PDSCH
TX j jc
---------------------------------------------------------------------------------------------------------------------------------------------
TXi ic
10
10
N Ant TX N Ant TX 10
+ 3 10
------------------------------------------------------------------------------------------------------------------------------------
+ 10 L og --------------------------------------------TX i ic
4
N Ant TX
TX i ic
With 4 or 8 antenna ports and N SD PDCCH = 1 :

TX jc
TX j jc
DLRS
j
TX jc
E
DLRS
N j
---------------------
TX ic TX j jc
10
Ant
TX
+f i
= 10 Log ------------------ 10
O
TX
ic
i
N Ant TX
TX jc
TX ic TX jc
j
i
j
j
i
j
E PDCCH + f PDCCH
E PDSCH + f PDSCH
TX j jc
---------------------------------------------------------------------------------------------------------------------------------------------
TXi ic
10
10
+ 5 10
------------------------------------------------------------------------------------------------------------------------------------
+ 10 L og --------------------------------------------TX i ic
6
N Ant TX
TX i ic
With 4 or 8 antenna ports and N SD PDCCH 1 :

TX jc
TX j jc
DLRS
j
TX jc
E DLRS
N j
---------------------
TX i ic TX j jc
10
Ant TX
= 10 Log ------------------ 10
TX i ic
+ fO
N Ant TX
TX jc
TX ic TX jc
j
i
j
j
i
j
E PDCCH + f PDCCH
E PDSCH + f PDSCH
TX j jc
---------------------------------------------------------------------------------------------------------------------------------------------
TXi ic
10
10
+ 2 10
------------------------------------------------------------------------------------------------------------------------------------
+ 10 L og --------------------------------------------TX i ic
3
N Ant TX
RS of the interfered cell TXi(ic) collide only with PDCCH and PDSCH of the interfering cell TXj(jc)
This
occurs
TX i ic
ID PSS
when
TX i ic
( ID PSS
TX j jc
= ID PSS
and
TX i ic
Shift
TX j jc
= Shift 3
and
TX i ic
TX j jc
ID PSS
With 1 or 2 antenna ports:

TX jc
TX j jc
DLRS
TX ic TX jc
TX ic TX jc
j
i
j
j
i
j
EPDCCH
+ f PDCCH
E PDSCH + f PDSCH
-----------------------------------------------------------------------
----------------------------------------------------------------------10
10
10
TX i ic TX j jc
+ 3 10
= 10 Log ------------------------------------------------------------------------------------------------------------------------------------ + f O
4
TX i ic
With 4 or 8 antenna ports and N SD PDCCH = 1 :
510
TX jc
TX j jc
N Ant TX = N Ant TX = 1 )
OR

AT332_TRR_E0
TX jc
TX jc
j
DLRS
TX ic TX jc
TX jc
TX ic TX jc
TX jc
TX ic TX jc
j
j
i
j
i
j
EPDCCH
E
+f
+f
PDSCH PDSCH
PDCCH
-----------------------------------------------------------------------
----------------------------------------------------------------------10
10
TX ic TX jc
10
+ 5 10
i
j
= 10 Log ------------------------------------------------------------------------------------------------------------------------------------ + f O
6
TX i ic
With 4 or 8 antenna ports and N SD PDCCH 1 :

TX jc
TX j jc
DLRS
TX ic TX jc
j
i
j
j
i
j
EPDCCH
+ f PDCCH
E PDSCH + f PDSCH
-----------------------------------------------------------------------
----------------------------------------------------------------------10
10
TX i ic TX j jc
10
+ 2 10
= 10 Log ------------------------------------------------------------------------------------------------------------------------------------ + f O
3
The interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi is calculated as follows:
TX j jc
TX j jc
SS PBCH
TX j jc
E PBCH
ESS
--------------------TX j jc
-------------------
10
10
TX i ic TX j jc
TX j jc
10
N Sym SS + 10
N Sym PBCH
------------------------------------------------------------------------------------------------------------= 10 Log
+ f MIMO
+ fO
TX j jc
The interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel, subscriber,
or mobile Mi is calculated as follows:
PDCCH of the interfered cell TXi(ic) collides with PDCCH and all the RS of the interfering cell TXj(jc)
This
occurs
TX ic
i
ID PSS
when
TX i ic
( ID PSS
TX j jc
= ID PSS
and
TX i ic
Shift
TX j jc
= Shift 3
and
TX i ic
TX j jc
OR
TX jc
j
ID PSS
TX jc
TX jc
j
PDCCH
E DLRS
TX j jc
---------------------
TX ic TX jc
N
1
j
Sym DLRS in PDCCH
- 10 10 + f O i
----------------------------------------= 10 Log ------------------TX j jc
TX i ic
N Ant TX
N Sym PDCCH
TX jc
TX ic TX jc
j
i
j
TX ic
E PDCCH + f PDCCH
TX j jc
N i
-----------------------------------------------------------------------
N
10
Sym PDCCH
Sym DLRS in PDCCH
- 10
+ 10 L og ----------------------------------------------------------------------------TX i ic
N Sym PDCCH
Here, N Sym DLRS in PDCCH is the number of downlink reference signal resource elements that fall within the PDCCH, and
N Sym PDCCH is the number of PDCCH resource elements per frame.
PDCCH of the interfered cell TXi(ic) collides with PDCCH and some RS of the interfering cell TXj(jc)
TX ic
i
TX jc
j
= ID PSS
TX jc
j
TX ic
i
511


TX jc
TX jc
j
PDCCH
j
TX ic
E
TX jc
TX ic
DLRS
j
i
N i
-------------------TX ic TX jc
N
N
j
10
Ant
TX
Sym
DLRS
in
PDCCH
Sym
DLRS
in
PDCCH
+f i
------------------------------------------------------------------------------------------- 10
= 10 Log ------------------O
TX jc
TX ic
j
i
N Ant TX
N Sym PDCCH
TX jc
TX ic TX jc
j
i
j
TX jc
E PDCCH + f
PDCCH
N j
-----------------------------------------------------------------------
10
Sym PDCCH
10
+ 10 L og ---------------------------- TXi ic
N Sym PDCCH
Here, N Sym DLRS in PDCCH is the number of downlink reference signal resource elements that fall within the PDCCH, and
N Sym PDCCH is the number of PDCCH resource elements per frame.
PDCCH of the interfered cell TXi(ic) collides only with PDCCH of the interfering cell TXj(jc)
TX i ic

TX j jc
TX j jc
TX j jc
= ID PSS
TX i ic TX j jc
PDCCH = E PDCCH + f PDCCH
TX j jc
TX i ic
TX i ic TX j jc
+ fO
The interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel, subscriber,
PDSCH of the interfered cell TXi(ic) collides with PDSCH and all the RS of the interfering cell TXj(jc)
This
occurs
TX i ic
ID PSS
when
TX i ic
( ID PSS
TX j jc
= ID PSS
and
TX i ic
Shift
TX j jc
= Shift 3
and
TX i ic
TX j jc
OR
TX j jc
ID PSS
TX jc
TX j jc
PDSCH
E DLRS
TX j jc
--------------------
TX ic TX j jc
N Sym DLRS in PDSCH
1
- 10 10 + f O i
----------------------------------------= 10 Log ------------------TX j jc
TX i ic
N Ant TX
N Sym PDSCH
TX jc
TX ic TX jc
j
i
j
TX ic
E PDSCH + f PDSCH
TX j jc
N i
-----------------------------------------------------------------------
N
10
Sym PDSCH
Sym DLRS in PDSCH
10
+ 10 L og ----------------------------------------------------------------------------TX i ic
N Sym PDSCH
Here, N Sym DLRS in PDSCH is the number of downlink reference signal resource elements that fall within the PDSCH, and
N Sym PDSCH is the number of PDSCH resource elements per frame.
PDSCH of the interfered cell TXi(ic) collides with PDSCH and some RS of the interfering cell TXj(jc)
TX i ic
TX j jc
= ID PSS
TX j jc
TX i ic
512

AT332_TRR_E0
TX jc
TX jc
j
PDSCH
j
TX ic
E
TX jc
TX ic
DLRS
j
i
N i
-------------------TX ic TX jc
N
N
j
10
Ant
TX
Sym
DLRS
in
PDSCH
Sym
DLRS
in
PDSCH
+f i
------------------------------------------------------------------------------------------ 10
= 10 Log ------------------O
TX jc
TX ic
j
i
N Ant TX
N Sym PDSCH
TX jc
TX ic TX jc
j
i
j
TX jc
E PDSCH + f
PDSCH
N j
-----------------------------------------------------------------------
10
Sym PDSCH
- 10
+ 10 L og --------------------------- TXi ic
N Sym PDSCH
Here, N Sym DLRS in PDSCH is the number of downlink reference signal resource elements that fall within the PDSCH, and
N Sym PDSCH is the number of PDSCH resource elements per frame.
PDSCH of the interfered cell TXi(ic) collides only with PDSCH of the interfering cell TXj(jc)
TX i ic
TX j jc
= ID PSS

TX j jc
TX j jc
TX i ic TX j jc
PDSCH = E PDSCH + f PDSCH
TX j jc
TX i ic
TX i ic TX j jc
+ fO
Method 2: Non-synchronised Transmission and Reception

Atoll calculates the interference between two cells using this method when:
The frequency channels assigned to the interfered and interfering cells do not have the same centre frequency, or
The interfered and interfering cells do not both have an even number of frequency blocks or do not both have an odd
number of frequency blocks, or
The Atoll.ini file contains the following option:
[LTE]
SameItf_PDSCH_RS_PDCCH = 1
This method is also used for calculating the interference received from LTE cells of an external network in co-planning mode,
i.e., inter-technology interference received from LTE cells calculated using the inter-technology IRFs.
The interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell TXj(jc) at a
TX jc
TX j jc
DLRS
TX jc
TX ic TX jc
j
i
j
E j
E PDCCH + f PDCCH
TX j jc
TX jc
DLRS ------------------------------------------------------------------------------------------ N j
N
10
10
Sym DLRS
Sym PDCCH
- + 10
------------------------ ----------------------------= 10 Log 10
TX j jc
TX j jc
N Sym DL
N Sym DL
+ 10
E PDSCH + f PDSCH
----------------------------------------------------------------------10
TX j jc
TX ic TX j jc
N Sym PDSCH
- + fO i
---------------------------TX j jc
N Sym DL
The interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc) at a pixel,
TX j jc
TX jc
j
SS PBCH
TX j jc
TX j jc
TX j jc
+ f MIMO
E PBCH + f MIMO
ESS
----------------------------------------------TX j jc
---------------------------------------------
10
10
TX ic TX jc
10
N Sym SS + 10
N Sym PBCH
j
- + f O i
= 10 Log --------------------------------------------------------------------------------------------------------------------------------------------------TX j jc
The interfering energy per resource element (dBm/Sym) received over the PDSCH and the PDCCH from any cell TXj(jc) at a
513


TX jc
TX jc
+ 10
TX jc TX ic TX jc
j
i
j
E PDSCH + f
PDSCH
----------------------------------------------------------------------10
TX jc
j
TX ic TX j jc
N Sym PDSCH
- + fO i
---------------------------TX jc
j
N Sym DL
TX jc
TX j jc
TX ic TX jc
j
i
j
E j
E
TX jc
TX jc
DLRS
PDCCH + f PDCCH
j
j
-----------------------------------------------------------------------------------------N Sym DLRS
N Sym PDCCH
10
10
----------------------------------------------------= 10 Log 10
TX jc + 10
TX jc
j
j
N Sym DL
N Sym DL
TX jc
j
PDSCH
PDCCH
TX jc
TX ic TX jc
j
i
j
E j
E PDCCH + f PDCCH
TX j jc
TX j jc
DLRS -----------------------------------------------------------------------------------------N Sym DLRS
N Sym PDCCH
10
10
- + 10
---------------------------- ------------------------
= 10 Log 10
TX j jc
TX j jc
N Sym DL
N Sym DL
+ 10
E PDSCH + f PDSCH
----------------------------------------------------------------------10
TX j jc
TX ic TX jc
N Sym PDSCH
j
- + fO i
---------------------------TX j jc
N Sym DL
E-UTRA carrier RSSI is measured on the OFDM symbols that contain reference signals. Therefore, the interfering energy per
frequency block (dBm/RB) received from any cell TXj(jc) at a pixel, subscriber, or mobile Mi over 1 frequency block during an
OFDM symbol carrying reference signals, is given as follows:
For cells using more than 1 antenna port, the encircled 10 in the formulas below is
replaced with 8.
TX j jc
TX j jc
RSSI
TX i ic TX j jc
EPDSCH + fPDSCH
TX j jc
---------------------------------------------------------------------10
N Sym PDSCH
10
10
= 10 Log --------------------------------------------------------------------------------------------TX j jc
TX j jc
N
+
N
Sym PDSCH
Sym PDCCH
TX j jc
TX i ic TX j jc
E PDCCH + f PDCCH
----------------------------------------------------------------------10
TX j jc
N Sym PDCCH
10
- 10 + 10
+ --------------------------------------------------------------------------------------------TX jc
TX jc
j
j
N Sym PDSCH + N Sym PDCCH
TX j jc
E DLRS
--------------------10
TX j jc
TX i ic TX j jc
2 Min 2 N Ant TX + f O
Calculation of PDCCH and PDSCH Interference Weighting Factors

TX i ic TX j jc
The PDCCH and PDSCH interference weighting factors ( f PDCCH
TX i ic TX j jc
f PDCCH
TX i ic TX j jc
TX jc
TX ic TX jc
TX ic TX jc
j
j
i
j
i
j
DL
f
+f
+f
+f
+f
MIMO
TL
ICIC DL
ABS DL
CoMP
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
TX j jc
10
1 AU DL 10
= 10 Log
TX
TX
TX
ic
TX
jc
TX
ic
TX
jc
j
j
i
j
i
j
DL
G SA G SA + f ICIC DL
+ f ABS DL
+ f CoMP
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
TX jc
10
j
+ AU DL
10
Calculation of MIMO/Antenna Diversity Interference Factors
514
) are calculated as follows:
TX j jc TX j jc TX i ic TX j jc TX i ic TX j jc DL
f MIMO + f TL
+ f ICIC DL
+ f ABS DL
+ f CoMP
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------
TX j jc
10
1 AU
10
DL
= 10 Log
TX j jc TX ic TX jc TX ic TX jc
i
j
i
j
DL
f
+f
+f
+f
TL
ICIC DL
ABS DL
CoMP
----------------------------------------------------------------------------------------------------------------------------------------------TX j jc
10
10
+ AU DL
TX jc
f PDSCH
TX i ic TX j jc
and f PDSCH

AT332_TRR_E0
TX jc
TX jc
TX jc
j
j
j
f MIMO is the interference increment due to more than one transmission antenna port: f MIMO = 10 Log N Ant TX
TX jc
j
If you do not wish to apply f MIMO , add the following lines in the Atoll.ini file:
[LTE]
MultiAntennaInterference = 0
MultiAntennaInterference is set to 1 by default.
Calculation of Interference Reduction Factors
Calculations for the interference reduction factors are explained below:
Interference reduction due to the traffic loads of the interfering cells:
Interference reduction due to the traffic loads of the interfering cells TXj(jc) is calculated as follows:
TX j jc
f TL
TX j jc
= 10 Log TL DL
Interference reduction due to the co- and adjacent channel overlap between the studied and the interfering cells:
Interference reduction due to the co- and adjacent channel overlap between the cells TXi(ic) and TXj(jc) is calculated
as follows:
TX i ic TX j jc
fO
TX i ic TX j jc
= 10 Log r O
Interference reduction due to static downlink ICIC using fractional frequency reuse:
The interference reduction factor due to static downlink ICIC using fractional frequency reuse for any pixel, subscriber,
TX i ic TX j jc
f ICIC DL
TX i ic TX j jc
= 10 Log p Collision
Whether a pixel, subscriber, or mobile Mi is in cell centre or cell-edge is determined as explained in "Best Server
TX i ic TX j jc
Depending on the ICIC mode defined for the frame configuration of the cells TXi(ic) and TXj(jc), f ICIC DL
is
If both TXi(ic) and TXj(jc) use time-switched FFR

The cell-edge and cell-centre resources are time-divided. Therefore, an interfered user may receive interference
from the cell-edge and cell-centre parts of the frame depending on time-domain switching points between the
cell-edge and cell-centre parts of the frames.
Atoll determines the switching point between the ICIC and the non-ICIC parts of the frame using the ICIC ratios.
The switching points between the ICIC and non-ICIC parts of the frame of the victim and interfering cells, TXi(ic)
and TXj(jc) respectively, are calculated as follows:
SP
TX i ic
TX i ic
TX j jc
TX j jc
r DL CE
r DL CE
= ---------------------------------------------------------------------------- and SP
= ---------------------------------------------------------------------------TX ic
TX jc
i
j
TX i ic
TX i ic
TX j jc
TX j jc
N
N
FB CE
FB CE
r DL CE + 1 r DL CE ----------------r DL CE + 1 r DL CE ----------------TX ic
TX jc
i
j
N FB
N FB
Where, SP is the switching point between the ICIC and the non-ICIC parts of the frame, and r DL CE is the downlink
cell-edge traffic ratios of the cells.
The ICIC ratio is used to partition the total downlink traffic load into ICIC and non-ICIC
parts of the frame. Therefore, the switching point formula is derived from the equation:
r DL CE TL DL
1 r DL CE TL DL
--------------------------------------------------------- = ---------------------------------------------- 1 SP W Channel
N FB CE
SP WChannel ----------------N FB
515

With cells using static downlink ICIC, there can be four different interference scenarios.
i.
Between the ICIC part of the victim and the ICIC part of the interferer.
ii. Between the ICIC part of the victim and the non-ICIC part of the interferer.
iii. Between the non-ICIC part of the victim and the ICIC part of the interferer.
iv. Between the non-ICIC part of the victim and the non-ICIC part of the interferer.
Therefore, Atoll calculates the probabilities of collision for each scenario and weights the total interference
according to the total collision probability. The probability of collision p Coll for each scenario is:
Case
Interfered cell
TX i ic
Interfering cell
TX j jc
p Coll
ICIC
ICIC
N FB CE
--------------------TX i ic
N FB CE
ii
ICIC
Non ICIC
Common
Common
iii
Non ICIC
ICIC
N FB CE
--------------------TX i ic
N FB
iv
Non ICIC
Non ICIC
Common
Where, N FB CE
TX ic
i
is the number of cell-edge frequency blocks common in TXi(ic) and TXj(jc), and N FB CE is the
number of cell-edge frequency blocks in the cell TXi(ic).

For a pixel, subscriber, or mobile Mi in the cell-edge of TXi(ic), the total collision probability for the pixel,
TX i ic TX j jc
p Collision
TX j jc
TX i ic
i
p Coll
If SP
SP
TX j jc
TX i ic
TX j jc
= i
ii
+ p Coll SP
SP
TX j jc
TX i ic
p Coll SP
------------------------------------------------------------------------------------------------------------ SP
If SP
TX i ic
SP
For a pixel, subscriber, or mobile Mi in the cell centre of TXi(ic), the total collision probability for the pixel,
TX i ic TX j jc
p Collision
TX j jc
TX i ic
iv
p Coll
If SP
SP
TX j jc
TX jc
TX ic
+ p iii SP j SP i
= p iv
TX jc
TX ic
Coll 1 SP
Coll
--------------------------------------------------------------------------------------------------------------------------- If SP j SP i
TX ic
1 SP i
Other combinations of ICIC modes

TX i ic TX j jc
Separate probabilities of collisions, p Collision

TX ic TX jc
i
j
Cell centre: p Collision
TX i ic TX j jc
Cell-edge: p Collision
Common
Where, N FB CC
, are calculated for cell-centre and cell-edge cases as follows:
Common
N FB CC
= -------------------TX i ic
N FB CC
Common
N FB CE
= -------------------TX ic
i
N FB CE
Common
is the number of common frequency blocks in TXi(ic) and TXj(jc) in cell centre, N FB CE
TX i ic
TX i ic
is the
number of common frequency blocks in TXi(ic) and TXj(jc) on cell-edge, N FB CC and N FB CE are respectively the
numbers of frequency blocks in cell centre and cell-edge of TXi(ic).
516

AT332_TRR_E0
ICIC mode
Cell centre
Cell edge
TX ic
i
No FFR
N FB
Time-switched FFR
N FB
TX ic
i
N FB
TX ic
i
TX ic
i
N FB CEx
TX i ic
Hard FFR
TX i ic
N FB CEx
TX ic
i
Soft FFR
N FB
TX i ic
Partial soft FFR
N FB
N FB CEx
TX ic
i
TX ic
i
N FB CEx
TX ic
TX ic
N FB CEx
TX ic
i
i
i
TX ic
i
TX ic
i
TX ic
i
TX i ic
N FB CEx
TX ic
i
Interference reduction due to subframe collision probabilities:
TX ic TX jc
i
j
The interference reduction factor due to downlink subframe collision probabilities f ABS DL
is calculated as
explained in "Subframe Pattern Collision Calculation" on page 497.

Interference reduction due to downlink coordinated scheduling CoMP:
Interference reduction due to the downlink coordinated scheduling CoMP is calculated as follows:
DL
NCoMP
TX k kc
TL DL
10 Log P DL = f ----------------------------------k
for k downlink CoMP servers
=
DL
CoMP
N CoMP
0
DL
f CoMP
DL
P CoMP is the collision probability read from the CoMP collision probability (DL) graph defined for the CoMP set to
N CoMP
TX k kc
TL DL
DL
k
which the servers belong. ----------------------------------is the average downlink traffic load of the N CoMP coordinated scheduling
DL
N CoMP
CoMP servers.
Calculation of the Downlink Inter-technology Interference
The downlink inter-technology interference is calculated as follows:
Inter Tech
I DL
TX k
P DL Rec
--------------------------------------
F
TX
ic
TX
i
k
TX k ICP DL
TX k
Here P DL Rec is the received downlink power from an interfering cell TXk belonging to another technology, and
F TX i ic TX k
ICPDL
is the inter-technology downlink channel protection ratio for a frequency offset F between the interfered
and interfering frequency channels of TXi(ic) and TXk.

TX k
P DL Rec is calculated based on the EIRP from GSM cells, total power from UMTS, CDMA2000, and TD-SCDMA cells, maximum
power from LTE cells, preamble power from WiMAX cells, and downlink cell power from Wi-Fi cells.
Output
TX j jc
DLRS : Interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell
TXj(jc) at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic).
517

TX jc
j
SS PBCH : Interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc)
at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic).
TX jc
j
PDSCH : Interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi covered by a cell TXi(ic).
TX j jc
PDCCH : Interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi covered by a cell TXi(ic).
TX j jc
RSSI : Interfering energy per frequency block (dBm/RB) received from any cell TXj(jc) at a pixel, subscriber, or mobile
Mi over 1 frequency block during an OFDM symbol carrying reference signals.
Inter Tech
I DL
: Downlink inter-technology interference.
6.4.4.4 C/N Calculation (DL)

Input
TX i ic
or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 502.
TX i ic
E SS
: Received SS energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
calculated in "Signal Level Calculation (DL)" on page 502.
TX i ic
E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i ic
E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i ic
E PDSCH : Received PDSCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i ic
n Sym
: Downlink noise for one subcarrier for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 507.
CINR DLRS : Downlink reference signal C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi as calculated in "C/
TX ic
i
(I+N) and Bearer Calculation (DL)" on page 520.

Mi
T SU MIMO DL : SU-MIMO threshold defined in the reception equipment used by Mis terminal.
T B : Bearer selection thresholds of the bearers defined in the reception equipment used by Mis terminal.
B DL Highest Service : Highest downlink bearer defined in the properties of the service used by the pixel, subscriber,
Mi
Mi
or mobile Mi.
Mi
B DL Lowest Service : Lowest downlink bearer defined in the properties of the service used by the pixel, subscriber,
or mobile Mi.
TX ic
i
N Ant RX : Number of reception (downlink) antenna ports defined for the terminal used by the pixel, subscriber, or
Mi
mobile Mi.
Mobility M i : Mobility used for the calculations.

M
i
BLER BDL : Downlink block error rate read from the graphs available in the reception equipment assigned to the

terminal used by the pixel, subscriber, or mobile Mi.
Mi
G Div PBCH : PBCH diversity gain defined in the reception equipment of the terminal used by the pixel, subscriber, or
mobile Mi.
G Div PDCCH : PDCCH diversity gain defined in the reception equipment of the terminal used by the pixel, subscriber,
or mobile Mi.
518

AT332_TRR_E0
DL
G Div : Additional downlink diversity gain defined for the clutter class where the pixel, subscriber, or mobile Mi is
located.
Calculations
The C/N for cell TXi(ic) are calculated as follows for any pixel, subscriber, or mobile Mi:
TX i ic
CNR DLRS
TX i ic
CNR SS
TX i ic
TX i ic
= E DLRS n Sym
TX i ic
= E SS
TX i ic
n Sym
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
CNR PBCH = E PBCH n Sym
CNR PDCCH = E PDCCH n Sym

CNR PDSCH = E PDSCH n Sym
Bearer Determination:
The bearers available for selection in the pixel, subscriber, or mobile Mis reception equipment are the ones:
Which are common between Mis and TXi(ic)s equipment (bearer indexes for which selection thresholds are
defined in both equipment), if the corresponding option has been set in the Atoll.ini file. For more information,
Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service
being accessed by Mi.
Whose selection thresholds are less than the PDSCH C/N at Mi: T B CNR PDSCH
Mi
TX i ic
If the cell supports MIMO, and according to the Mi diversity mode depending on the SU-MIMO and MU-MIMO
Mi
thresholds and criteria, transmit diversity, SU-MIMO diversity, or MU-MIMO diversity gain, G Div DL ,
corresponding to the bearer is applied to its selection threshold. The gain is read from the properties of the
TX i ic
Mi
reception equipment assigned to the pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX , Mobility M i ,
M
i
BLER B DL .

DL
The additional downlink diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is
also applied. Therefore, the bearers available for selection are all the bearers defined in the reception equipment
for which the following is true:
M
TX ic
i
DL
T B G Div DL G Div CNR PDSCH

The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic).
Bearer Index
From among the bearers available for selection, the selected bearer is the one with the highest index.
Peak RLC Throughput

From among the bearers available for selection, the selected bearer is the one with the highest downlink peak RLC
channel throughput as calculated in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and
Per-user Throughput Calculation" on page 551.

From among the bearers available for selection, the selected bearer is the one with the highest downlink effective
RLC channel throughput as calculated in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput,
and Per-user Throughput Calculation" on page 551.
MIMO Diversity Gain:

With MIMO, the PBCH and PDCCH C/N become:
TX i ic
TX i ic
Mi
DL
CNR PBCH = CNR PBCH + G Div PBCH + G Div
519

TX ic
i
TX ic
i

M
DL
CNR PDCCH = CNR PDCCH + G Div PDCCH + G Div

The PBCH and PDCCH diversity gains are applied to the C/N when the cell and the terminal both support any form of
MIMO in downlink. The additional downlink diversity gain defined per clutter is also applied.
Once the bearer is known, the PDSCH C/N calculated above becomes:
TX i ic
TX i ic
Mi
DL
CNR PDSCH = CNR PDSCH + G Div DL + G Div

Mi
Where G Div DL is the transmit diversity, SU-MIMO diversity, or MU-MIMO diversity gain applied if the cell supports
MIMO and depending on the Mi diversity mode, the SU-MIMO and MU-MIMO thresholds and criteria.
Output
TX i ic
CNR DLRS : Downlink reference signal C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
CNR SS
CNR PBCH : PBCH C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
CNR PDCCH : PDCCH C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
CNR PDSCH : PDSCH C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i ic
: SS C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i ic
TX i ic
TX ic
i
6.4.4.5 C/(I+N) and Bearer Calculation (DL)

The carrier signal to interference and noise ratio is calculated in three steps. First Atoll calculates the received signal level from
the studied cell (as explained in "Signal Level Calculation (DL)" on page 502) at the pixel, subscriber, or mobile under study.
Next, Atoll calculates the interference received at the same studied pixel, subscriber, or mobile from all the interfering cells
(as explained in "Interference Calculation (DL)" on page 507). Interference from each cell is weighted according to the co- and
adjacent channel overlap between the studied and the interfering cells, the traffic loads of the interfering cells, and the
probability of collision in case ICIC is used by the cells. Finally, Atoll takes the ratio of the signal level and the sum of the total
interference from other cells and the downlink noise (as calculated in "Noise Calculation (DL)" on page 507).
The receiver terminal is always considered to be oriented towards its best server, except when the "Lock Status" is set to
"Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited. In the case of NLOS between
the receiver and the best server, Atoll does not try to find the direction of the strongest signal, the receiver is oriented towards
the best server just as in the case of LOS.
Input
D CP
TX i ic
: Cyclic prefix duration defined in TXi(ic) frame configuration or, otherwise, in the global network settings.
TX i ic
TX i ic
N FB
TX i ic
is Extended).
: Number of frequency blocks, defined in the frequency bands table, for the channel bandwidth used by the
cell TXi(ic).
TX i ic
TX i ic
bands.
TX i ic
TX i ic
520

AT332_TRR_E0
TX ic
i
TX ic
i
Configuration
N SF DL
N TDD SSF
FDD
10
DSUUU-DSUUU
DSUUD-DSUUD
DSUDD-DSUDD
DSUUU-DSUUD
DSUUU-DDDDD
DSUUD-DDDDD
DSUDD-DDDDD
TX i ic
or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 502.
TX ic
i
E SS
: Received SS energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX ic
i
TX i ic
E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i ic
E PDSCH : Received PDSCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i ic
N Sym PDCCH : Number of PDCCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX i ic
N Sym PDSCH : Number of PDSCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX i ic
n Sym
: Downlink noise for one subcarrier for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 507.
DLRS : Interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell
TX j jc
TXj(jc) at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on
page 507.
TX jc
j
SS PBCH : Interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc)
at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 507.
TX j jc
PDSCH : Interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 507.
TX jc
j
PDCCH : Interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 507.
TX j jc
RSSI : Interfering energy per frequency block (dBm/RB) received over 1 frequency block during an OFDM symbol
carrying reference signals from any cell TXj(jc) at a pixel, subscriber, or mobile Mi as calculated in "Interference
Calculation (DL)" on page 507.
Inter Tech
NR DL
: Inter-technology downlink noise rise.
TX i ic
CNR DLRS : Downlink reference signal C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi as calculated in "C/N
Mi
T SU MIMO DL : SU-MIMO threshold defined in the reception equipment used by Mis terminal.
T B : Bearer selection thresholds of the bearers defined in the reception equipment used by Mis terminal.
Mi
521

M
or mobile Mi.
M
or mobile Mi.
TX i ic
N Ant RX : Number of reception (downlink) antenna ports defined for the terminal used by the pixel, subscriber, or
Mi
mobile Mi.
i
BLER BDL : Downlink block error rate read from the graphs available in the reception equipment assigned to the

Mi
G Div PBCH : PBCH diversity gain defined in the reception equipment of the terminal used by the pixel, subscriber, or
mobile Mi.
Mi
G Div PDCCH : PDCCH diversity gain defined in the reception equipment of the terminal used by the pixel, subscriber,
or mobile Mi.
DL
located.
Inter Tech
I DL
: Downlink inter-technology interference as calculated in "Interference Calculation (DL)" on page 507.
Calculations
The downlink reference signal C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc
TX i ic
CINR DLRS
TX ic
i
j
n Sym
DLRS
------------------
---------------------
TX i ic
= E DLRS 10 Log
DL
DL
All TXj jc
The SS C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc
TX i ic
CINR SS
TX i ic
= E SS
TX ic
i
j
n Sym
SS PBCH
------------------------
---------------------
Inter
Tech
Inter
Tech
10
10
10
+I
+ NR
+ 10
10 Log
DL
DL
All TXj jc
The PBCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc
TX i ic
CINR PBCH
TX ic
n Sym
SS PBCH
---------------------------------------------
TX i ic
= E PBCH 10 Log
DL
DL
All TXj jc
The PDCCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc
TX i ic
CINR PDCCH
TX ic
i
j
n Sym
PDCCH-
-------------------
---------------------
TX i ic
Inter Tech
Inter Tech
10
10
+10
= E PDCCH 10 Log
10
+ I DL
+ NR DL
All TXj jc
The PDSCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc
TX i ic
CINR PDSCH
522
TX ic
j
n Sym
PDSCH-
----------------------------------------
TX i ic
10
10
Inter
Tech
Inter
Tech
+ NR
10
+I
+10
= E PDSCH 10 Log
DL
DL
All TXj jc

AT332_TRR_E0
The downlink reference signal received quality (RSRQ) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile
Mi:
RSRQ
TX ic
i
TX ic
i
= 10 Log N FB
TX i ic
Where E DLRS
TX ic
i
+ E DLRS RSSI
is the cells RSRP and RSSI
TX i ic
TX ic
i
is the received signal strength indicator, i.e., the received signals from the
server (TXi(ic)), and all the interfering cells (TXj(jc)), calculated as follows:
For cells using more than 1 antenna port, the encircled 10 in the formulas below is
replaced with 8.
TX jc
RSSI
TX i ic
TX ic
j
n Sym
RSSI
TX ic
------------------
--------------------TX ic
i
10 10 + I Inter Tech + 10 10 12 + NR Inter Tech + 10 Log N i
= 10 Log RSSI +
DL
FB
DL
All TX j jc
The downlink reference signal total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc
TX i ic
I + N DLRS
TX ic
j
n Sym
DLRS -
------------------
---------------------
TX ic
10 10 + I Inter Tech + 10 10 + NR Inter Tech + 10 Log 2 N i
= 10 Log
DL
DL
FB
All TXj jc
The SS and PBCH total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc
TX i ic
I + N SS PBCH
TX ic
i
j
n Sym
SS PBCH-
------------------------
---------------------
10 10 + I Inter Tech + 10 10 + NRInter Tech + 10 Log N
= 10 Log
DL
SCa FB N FB SS PBCH
DL
All TX j jc
The PDSCH and PDCCH total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
Method 1: Synchronised Transmission and Reception
For details, see "Interference Calculation (DL)" on page 455.
TX i ic
I + N PDCCH
TX i ic
I + N PDSCH
TX jc
TX ic
TX jc
TX ic
j
n Sym
TX ic
PDCCH-
i
----------------------------------------
N Sym PDCCH
10
10
Inter Tech
Tech
- + NR Inter
= 10 Log
10
+ I DL
+ 10
+ 10 Log ---------------------------------------------DL
TX ic
TX i ic
i
N SF DL + N TDD SSF
All TXj jc
j
n Sym
TX i ic
PDSCH-
-------------------
---------------------
N Sym PDSCH
10
10
Inter Tech
Tech
----------------------------------------------- + NR Inter
= 10 Log
+ 10
DL
TX i ic
10
+ I DL
+ 10 Log TXi ic
All TXj jc
N SF DL + N TDD SSF
Method 2: Non-synchronised Transmission and Reception

For details, see "Interference Calculation (DL)" on page 455.
TX jc
TX i ic
I + N PDCCH
TX jc
TX i ic
I + N PDSCH
TX ic
j
n Sym
TX i ic
PDCCH-
------------------- N TXi ic
---------------------
Inter
Inter Tech
10
10
------------------------------------------------------------------= 10 Log
10
+
+
10
+
+ NR DL
I
10
Log
TX i ic
DL
All TXj jc
TX ic
j
n Sym
TX i ic
PDSCH
------------------- N TXi ic
---------------------
Inter Tech
10
10
- + NR Inter
= 10 Log
10
+ I DL
+ 10
+ 10 Log ------------------------------------------------------------------DL
TX i ic
All TXj jc
523

With N SCa FB calculated as follows:

W FB
N SCa FB = --------F
The bearers available for selection in the pixel, subscriber, or mobile Mis reception equipment are the ones:
Whose selection thresholds are less than the PDSCH C/(I+N) at Mi: T B CINR PDSCH
TX ic
i
Mi
thresholds and criteria, transmit diversity, SU-MIMO diversity, or MU-MIMO diversity gain, G Div DL ,
TX ic
i
reception equipment assigned to the pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX , Mobility M i ,
M
i
BLER B DL .

DL
The additional downlink diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is
also applied. Therefore, the bearers available for selection are all the bearers defined in the reception equipment
for which the following is true:
Mi
Mi
TX i ic
DL
T B G Div DL G Div CINR PDSCH

Bearer Index
Peak RLC Throughput

From among the bearers available for selection, the selected bearer is the one with the highest downlink peak RLC


With MIMO, the PBCH and PDCCH C/(I+N) become:
TX i ic
TX i ic
Mi
TX ic
i
TX ic
i
DL
CINR PBCH = CINR PBCH + G Div PBCH + G Div

i
DL
CINR PDCCH = CINR PDCCH + G Div PDCCH + G Div

The PBCH and PDCCH diversity gains are applied to the C/(I+N) when the cell and the terminal both support any form
of MIMO in downlink. The additional downlink diversity gain defined per clutter is also applied.
Once the bearer is known, the PDSCH C/(I+N) calculated above becomes:
TX i ic
TX i ic
Mi
DL
CINR PDSCH = CINR PDSCH + G Div DL + G Div

Mi
Where G Div DL is the transmit diversity, SU-MIMO diversity, or MU-MIMO diversity gain applied if the cell supports
MIMO and depending on the Mi diversity mode, the SU-MIMO and MU-MIMO thresholds and criteria.
524

AT332_TRR_E0
Output
TX ic
i
CINR DLRS : Downlink reference signal C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
CINR SS
TX ic
i
: SS C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i ic
CINR PBCH : PBCH C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
CINR PDCCH : PDCCH C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
CINR PDSCH : PDSCH C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
RSRQ
TX ic
i
TX i ic
TX i ic
: Downlink reference signal received quality from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i ic
RSSI
: Received signal strength indicator, i.e., the received signals from the server (TXi(ic)), and all the interfering
cells (TXj(jc)), at pixel, subscriber, or mobile Mi.
I + N DLRS : Downlink reference signals total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile
TX i ic
Mi covered by a cell TXi(ic).
TX i ic
I + N SS PBCH : SS and PBCH total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered
by a cell TXi(ic).
TX ic
i
I + N PDCCH : PDCCH total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by a
cell TXi(ic).
TX i ic
I + N PDSCH : PDSCH total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by a
cell TXi(ic).
B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink.
6.4.4.6 Signal Level Calculation (UL)

Input
TX i ic
CINR PUSCH Max : Maximum PUSCH C/(I+N) defined for the cell TXi(ic).
NR UL
TX i ic
: Uplink noise rise of the cell TXi(ic). This value can be user-defined or calculated as explained in "Interference
Calculation (UL)" on page 527.

TX i ic
NR UL ICIC : ICIC uplink noise rise of the cell TXi(ic). This value can be user-defined or calculated as explained in
n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic).
N FB
TX ic
i
TX i ic
cell TXi(ic).
TX i ic
FPC
: Fractional uplink power control factor defined for the cell TXi(ic).
Mi
P Max : Maximum transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi after power control
Mi
adjustment as calculated in "C/(I+N) and Bearer Calculation (UL)" on page 534.

TX i
G Ant : Transmitter antenna gain for the antenna used by the transmitter TXi.
TX i
TX i
= L Total UL ).
TX i
L Path : Path loss ( L Path = L Model + L Ant ).
L Total : Total loss calculated as explained in "Signal Level Calculation (DL)" on page 502.
525

TX

is selected.

i
Mi
M
Mi
D CP
TX i ic
Calculations
Atoll first calculates the allowed maximum transmission power for the terminal used by the pixel, subscriber, or mobile Mi.
This power is calculated by performing fractional power control.
Fractional Power Control:
Fractional power control imposes a limitation on the maximum transmission power of the terminal. A nominal PUSCH
power is indicated by the cell to all the pixels, subscribers, or mobiles. This nominal PUSCH power is calculated as
follows:
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
P O_PUSCH = CINR PUSCH Max + NRUL
TX ic
TX ic
i
i
+ n PUSCH PUCCH 10 Log N FB
TX i ic
for cell centre.
TX i ic
P O_PUSCH = CINR PUSCH Max + NRUL ICIC + n PUSCH PUCCH 10 Log NFB
TX ic
for cell-edge.
TX ic
i
i
Where n PUSCH PUCCH 10 Log N FB
corresponds to the uplink noise over 1 frequency block.
Next, the maximum allowed transmission power for the terminal used by the pixel, subscriber, or mobile Mi is
Mi
TX i ic
TX i ic
TX i ic
Mi
P Allowed = Min P Max 10 Log NFB + P O_PUSCH + FPC L Total
Once the maximum allowed power has been calculated, it is used as an upper limit for transmission power in all the remaining
calculations.
The received PUSCH and PUCCH signal level (dBm) from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated
as follows:
Mi
Mi
TX i
C PUSCH PUCCH = EIRP PUSCH PUCCH L Path M Shadowing Model L Indoor + G Ant L
TX i
Mi
Mi
L Ant L Body + f CP
Where EIRP is the effective isotropic radiated power of the terminal calculated as follows:
M
EIRP PUSCH PUCCH = P

With P
Mi
+G
= P Allowed without power control adjustment at the start of the calculations, and is P
adjustment.
526
Mi
Mi
Mi
= P Eff after power control

AT332_TRR_E0
f CP is the cyclic prefix factor, i.e., the ratio of the useful symbol energy to the total symbol energy.
The total symbol duration of a modulation symbol comprises the useful symbol duration, carrying the actual data bits, and a
cyclic prefix, added to the useful data bits as padding against multi-path to avoid inter-symbol interference. Hence, the total
energy within a modulation symbol belongs in part to the useful data bits and in part to the cyclic prefix. Once a modulation
symbol is received, only the energy of the useful data bits can be used for extracting the data. The energy belonging to the
cyclic prefix is lost once it has served its purpose of combatting inter-symbol interference. Therefore, f CP implies that the
energy belonging to the cyclic prefix is excluded from the useful signal level.
f CP
TX ic
10 Log 7 7,5 If D CPi = Normal
TX i ic
=
= Extended
0
If M i is an interferer
The cyclic prefix energy and the useful data bits energy are both taken into account when calculating interfering signal levels.
Output
Mi
C PUSCH PUCCH : Received PUSCH and PUCCH signal level from the pixel, subscriber, or mobile Mi at a cell TXi(ic).
P Allowed : Maximum allowed transmission power for the terminal used by the pixel, subscriber, or mobile Mi.
Mi
6.4.4.7 Noise Calculation (UL)

For determining the C/N and C/(I+N), Atoll calculates the uplink noise over the channel bandwidth used by the cell. The used
bandwidth depends on the number of used subcarriers.
The uplink noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on the
temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the used
bandwidth.
Input
W FB : Width of a frequency block in the frequency domain (180 kHz).
N FB
TX i ic
cell TXi(ic).
nf
TX i ic
: Noise figure of the cell TXi(ic).
Calculations
n 0 = 10 Log K T 1000 = 174 dBm/Hz
The thermal noise for the PUSCH and the PUCCH is calculated as:
TX ic
TX ic
i
i
n 0 PUSCH PUCCH = n 0 + 10 Log N FB
W FB 1000
The uplink noise is the sum of the thermal noise and the noise figure of the cell TXi(ic).
TX i ic
TX i ic
n PUSCH PUCCH = n 0 PUSCH PUCCH + nf
TX i ic
Output
TX i ic
n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic).
6.4.4.8 Interference Calculation (UL)

The PUSCH and PUCCH interference is only calculated during Monte Carlo simulations. In coverage predictions, the uplink
noise rise values already available in simulation results or in the Cells table are used.
527

The interference received by a cell TXi(ic) from an interfering mobile covered by a cell TXj(jc) can be defined as the PUSCH and
PUCCH signal level received from the interfering mobile Mj depending on the overlap that exists between the channels used
by the cells TXi(ic) and TXj(jc) and on the traffic load of the interfering mobile Mj.
The calculation of uplink interference can be divided into two parts:
6.4.4.8.1
Calculation of the uplink interference from each individual interfering mobile as explained in "Interfering Signal Level
Calculation of the uplink noise rise which represents the total uplink interference from all interfering mobiles as
explained in "Noise Rise Calculation (UL)" on page 530.
Interfering Signal Level Calculation (UL)

Input
TX i ic
N FB
TX j jc
and N FB
TX i ic
TX j jc
TX i ic
TX j jc
TX i ic
TX j jc
Mj
C PUSCH PUCCH : PUSCH and PUCCH signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell
TXj(jc) as calculated in "Signal Level Calculation (UL)" on page 525.

In Monte Carlo simulations, interfering signal levels already include M Shadowing Model , as explained in "Signal Level
information, see "Shadow Fading Model" on page 90). As the interfering signal levels already include
M Shadowing Model , M Shadowing C I is added to the received interfering signal levels in order to achieve the ratio
M Shadowing Model M Shadowing C I :
Mj
Mj
C PUSCH PUCCH = C PUSCH PUCCH + M Shadowing C I

is selected.
TX i ic TX j jc
TL UL
rO
TX j jc
: Uplink traffic load of the interfering cell TXj(jc).
Cell traffic loads can either be calculated using Monte Carlo simulations, or entered manually for each cell. Calculation
of traffic loads is explained in "Simulation Process" on page 479.
Mj
TL UL : Uplink traffic load of the interfering mobile Mj.

Mobile traffic loads are calculated during Monte Carlo simulations as explained in "Scheduling and Radio Resource
Allocation" on page 556.
Calculations
The uplink interference received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) is calculated as follows:
Mj
Mj
TX i ic TX j jc
I PUSCH PUCCH = C PUSCH PUCCH + f O
528
Mj
TX i ic TX j jc
+ f TL UL + f ICIC UL
TX i ic TX j jc
+ f ABS UL
UL
+ f CoMP

AT332_TRR_E0
M
Where f TL UL is an interference reduction factor due to the uplink traffic load of the interfering mobile Mj, calculated as
follows:
M
j
j
f TL UL = 10 Log TL UL
Calculations for other interference reduction factors are explained below:

as follows:
TX i ic TX j jc
fO
TX i ic TX j jc
= 10 Log r O
Interference reduction due to static uplink ICIC using fractional frequency reuse:
If the cell supports static ICIC in uplink, it means that a part of the LTE frame may use a fraction of the channel
bandwidth. The interference reduction factor due to static uplink ICIC using fractional frequency reuse is calculated as
follows:
TX ic TX jc
i
j
f ICIC UL
TX ic TX jc
j
i
= 10 Log p Collision
TX i ic TX j jc
Where p Collision
is the collision probability between the subcarriers used by the interfered and interfering cells.
It is determined during Monte Carlo simulations as follows:

TX i ic TX j jc
Cell centre: p Collision
TX i ic TX j jc
Cell-edge: p Collision
Common
Where, N FB CC
Common
N FB CC
= -------------------TX i ic
N FB CC
Common
N FB CE
= -------------------TX i ic
N FB CE
Common
is the number of common frequency blocks in TXi(ic) and TXj(jc) in cell centre, N FB CE
TX i ic
is the
TX i ic
number of common frequency blocks in TXi(ic) and TXj(jc) on cell-edge, N FB CC and N FB CE are respectively the
numbers of frequency blocks in cell centre and cell-edge of TXi(ic).
ICIC mode
Cell centre
Cell edge
TX ic
i
No FFR
N FB
Time-switched FFR
N FB
TX ic
i
N FB
TX i ic
TX i ic
N FB CEx
TX i ic
Hard FFR
TX i ic
N FB CEx
TX i ic
Soft FFR
N FB
Partial soft FFR
TX i ic
TX i ic
N FB
N FB CEx
TX i ic
TX i ic
N FB CEx
TX ic
TX ic
N FB CEx
TX ic
i
i
i
TX i ic
TX i ic
TX i ic
N FB CEx
TX i ic
Whether a pixel, subscriber, or mobile is located in the cell-edge is determined as explained in "Best Server
Interference reduction due to subframe collision probabilities:
TX i ic TX j jc
The interference reduction factor due to uplink subframe collision probabilities f ABS UL
is calculated as
explained in "Subframe Pattern Collision Calculation" on page 497.
529

Interference reduction due to uplink coordinated scheduling CoMP:

Interference reduction due to the uplink coordinated scheduling CoMP is calculated as follows:
UL
NCoMP
TX kc
TL UL
UL
k
for k uplink CoMP servers
= 10 Log P CoMP = f ----------------------------------UL
N
CoMP
0
UL
f CoMP
UL
P CoMP is the collision probability read from the CoMP collision probability (UL) graph defined for the CoMP set to
N CoMP
TX kc
k
TL UL
UL
k
which the servers belong. ----------------------------------is the average uplink traffic load of the N CoMP coordinated scheduling
UL
N CoMP
CoMP servers.
In Monte Carlo simulations, Atoll calculates two separate noise rise values; for the mobiles located in the cell-edge of the
interfered cell Atoll calculates the ICIC UL Noise Rise, and for the mobiles located in the cell centre of the interfered cell Atoll
calculates the UL Noise Rise.
In coverage predictions, point analysis, and calculations on subscriber lists, according to the zone, cell centre or cell-edge,
where the pixel, receiver, or subscriber is located, Atoll uses either the ICIC UL Noise Rise or the UL Noise Rise to calculate the
PUSCH and PUCCH C/(I+N). For more information on the calculation of the uplink noise rise, see "Noise Rise Calculation (UL)"
on page 530.
Output
6.4.4.8.2
Mj
I PUSCH PUCCH : PUSCH and PUCCH interference signal level received at a cell TXi(ic) from an interfering mobile Mj
covered by a cell TXj(jc).
Noise Rise Calculation (UL)

The uplink noise rise is defined as the ratio of the total uplink interference received by any cell TXi(ic) from all interfering
mobiles Mj present in the coverage areas of all other cells TXj(jc) to the uplink noise of the cell TXi(ic). In other words, it is the
ratio (I+N)/N.
Input
Mj
I PUSCH PUCCH : PUSCH and PUCCH interference signal levels received at a cell TXi(ic) from interfering mobiles Mj
covered by other cells TXj(jc) as calculated in "Interfering Signal Level Calculation (UL)" on page 528.
TX i ic
n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic) as calculated in "Noise Calculation (UL)"
on page 527.
Inter Tech
NRUL
: Inter-technology uplink noise rise.
Calculations
For any mobile Mi in the cell centre of the interfered cell TXi(ic), Atoll calculates the UL Noise Rise as follows:
M
TX i ic
NR UL
TX i ic
I PUSCH PUCCH
n
non-ICIC M
PUSCH PUCCH
------------------------------------------------------------------------------i
-------------------------------------------
TX ic
10
10
+ NR Inter Tech n i
= 10 Log
10
+ 10
UL
PUSCH PUCCH
All M j
All TXj jc
For any pixel, subscriber, or mobile Mi in the cell centre of the interfered cell TXi(ic), Atoll calculates the PUSCH and PUCCH
total noise (I+N) as follows:
TX i ic
TX i ic
I + N PUSCH PUCCH = NR UL
530
TX i ic
+ n PUSCH PUCCH

AT332_TRR_E0
For any mobile Mi in the cell-edge of the interfered cell TXi(ic), Atoll calculates the ICIC UL Noise Rise as follows:
M
TX ic
IPUSCH PUCCH
n
ICIC M
PUSCH PUCCH
i
-------------------------------------------
-------------------------------------------------------------------
TX ic
10
10
+ NR Inter Tech n i
= 10 Log
10
+ 10
UL
PUSCH PUCCH
All M j
All TX j jc
TX ic
i
NRUL ICIC
For any pixel, subscriber, or mobile Mi in cell-edge of the interfered cell TXi(ic), Atoll calculates the PUSCH and PUCCH total
noise (I+N) as follows:
TX i ic
TX i ic
TX i ic
I + N PUSCH PUCCH = NR UL ICIC + n PUSCH PUCCH

Output
TX ic
i
NR UL
: Uplink noise rise for the cell TXi(ic).
TX i ic
NR UL ICIC : ICIC uplink noise rise for the cell TXi(ic).
I + N PUSCH PUCCH : PUSCH and PUCCH total noise for a cell TXi(ic) calculated for any pixel, subscriber, or mobile Mi.
TX i ic
6.4.4.9 C/N Calculation (UL)

Input
Mi
C PUSCH PUCCH : Received PUSCH and PUCCH signal level from the pixel, subscriber, or mobile Mi at its serving cell TXi(ic)
as calculated in "Signal Level Calculation (UL)" on page 525.
TX i ic
n PUSCH PUCCH : PUSCH and PUCCH noise for the cell TXi(ic) as calculated in "Noise Calculation (UL)" on page 527.
T SU MIMO UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(ic).
N FB
TX i ic
TX ic
i
: Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by
the cell TXi(ic).
TX i ic
N FB CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth
used by the cell TXi(ic) and PSS ID 0.
TX ic
i
TX i ic
TX i ic
T B Lowest : Bearer selection threshold of the lowest bearer in the reception equipment assigned to the cell TXi(ic).
P Allowed : Maximum allowed transmission power of the terminal used by the pixel, subscriber, or mobile Mi as
Mi
calculated in "Signal Level Calculation (UL)" on page 525.

Mi
P Min : Minimum transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
M PC : Power control adjustment margin defined in the global network settings.
T B : Bearer selection thresholds of the bearers defined in the reception equipment used bythe cell TXi(ic).
B UL Highest Service : Highest downlink bearer defined in the properties of the service used by the pixel, subscriber,
Mi
Mi
or mobile Mi.
Mi
B UL Lowest Service : Lowest downlink bearer defined in the properties of the service used by the pixel, subscriber,
or mobile Mi.
531

N Ant TX : Number of transmission (uplink) antenna ports defined for the terminal used by the pixel, subscriber, or
mobile Mi.
TX ic
i
N Ant RX : Number of reception (uplink) antenna ports defined for the cell TXi(ic).
i
BLER BUL : Uplink block error rate read from the graphs available in the reception equipment assigned to the cell

TXi(ic).
Calculations
The PUSCH and PUCCH C/N from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated as follows:
Mi
TX i ic
Mi
CNR PUSCH PUCCH = C PUSCH PUCCH n PUSCH PUCCH

The bearers available for selection in the cell TXi(ic)s reception equipment are the ones:
Whose selection thresholds are less than the PUSCH and PUCCH C/N at Mi: T B CNR PUSCH PUCCH
Mi
Mi
TX i ic
thresholds and criteria, receive diversity, SU-MIMO diversity, or MU-MIMO diversity gain, G Div UL ,
M
TX ic
i
i
i
reception equipment assigned to the cell TXi(ic) for N Ant TX , N Ant RX , Mobility M i , BLER B UL .
UL
The additional uplink diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is also
applied. Therefore, the bearers available for selection are all the bearers defined in the reception equipment for
which the following is true:
Mi
TX i ic
Mi
UL
T B G Div UL G Div CNR PUSCH PUCCH

Bearer Index
Peak RLC Throughput

From among the bearers available for selection, the selected bearer is the one with the highest uplink peak RLC

From among the bearers available for selection, the selected bearer is the one with the highest uplink effective

Once the bearer is known, the PUSCH and PUCCH C/N calculated above become:
M
TX ic
i
UL
CNR PUSCH PUCCH = CNR PUSCH PUCCH + G Div UL + G Div

TX i ic
Where G Div UL is the receive diversity, SU-MIMO diversity, or MU-MIMO diversity gain applied if the cell supports
MIMO and depending on the Mi diversity mode, the SU-MIMO and MU-MIMO threshold and criteria.
532

AT332_TRR_E0
Uplink Bandwidth Allocation (No. of Used Frequency Blocks):

The uplink bandwidth allocation depends on the target defined for the scheduler used by the cell TXi(ic). The PUSCH
and PUCCH C/N calculated above is given for the total number of frequency blocks associated with the channel
TX ic
i
bandwidth of the cell, i.e., N FB
. Bandwidth allocation is performed for all the pixels, subscribers, or mobiles in the
uplink, and may reduce the number of used frequency blocks in order to satisfy the selected target.
Full Bandwidth
Full channel width is used by each mobile in the uplink. As there is no reduction in the bandwidth used for
transmission, there is no gain in the PUSCH and PUCCH C/N.
Maintain Connection
The bandwidth used for transmission by a mobile is reduced only if the PUSCH and PUCCH C/N is not enough to
even access the lowest bearer. For example, as a mobile moves from good to bad radio conditions, the number of
frequency blocks used by it for transmission in uplink are reduced one by one in order to improve the PUSCH and
PUCCH C/N. The calculation of the gain introduced by the bandwidth reduction is explained below.
Best Bearer
The bandwidth used for transmission by a mobile is reduced in order to improve the PUSCH and PUCCH C/N
enough to access the best bearer. For example, if using 5 frequency blocks, a mobile is able to access the best
bearer, and using 6 it would only get access to the second best, it will be assigned 5 frequency blocks as the used
uplink bandwidth. Although using 4 frequency blocks, its PUSCH and PUCCH C/N will be better than when using 5,
the uplink bandwidth is not reduced to 4 because it does not provide any gain in terms of the bearer, i.e., the
mobile already has the best bearer using 5 frequency blocks. The calculation of the gain introduced by the
bandwidth reduction is explained below.
The definition of the best bearer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic),
i.e., bearer with the highest index, with the highest peak RLC throughput, or with the highest effective RLC
throughput.
The uplink bandwidth allocation may result in the use of a number of frequency blocks which is less than the number
of frequency blocks associated with the channel bandwidth of the cell. The gain related to this bandwidth reduction
is applied to the PUSCH and PUCCH C/N:
N TX i ic
Mi
Mi
FB
-
CNR PUSCH PUCCH = CNR PUSCH PUCCH + 10 Log ----------------Mi
All FB
Final
N FB UL
TX i ic
Mi
Min
Where N FB UL Service N FB UL N FB CC for any pixel, subscriber, or mobile Mi in the cell centre of the interfered
TX i ic
Mi
Min
cell TXi(ic), and N FB UL Service N FB UL N FB CE for any pixel, subscriber, or mobile Mi in the cell-edge of the
interfered cell TXi(ic).
ICIC mode
Cell centre
Cell edge
TX ic
i
No FFR
N FB
Time-switched FFR
N FB
TX ic
i
N FB
TX i ic
TX i ic
N FB CEx
TX i ic
Hard FFR
TX i ic
N FB CEx
TX i ic
Soft FFR
N FB
Partial soft FFR
TX i ic
TX i ic
N FB
N FB CEx
TX i ic
TX i ic
N FB CEx
TX ic
TX ic
N FB CEx
TX ic
i
i
i
TX i ic
TX i ic
TX i ic
N FB CEx
TX i ic
Uplink Power Control Adjustment:
533

Once the bandwidth allocation is performed, Atoll continues to work with the C/N given by the bandwidth allocation,
M
i.e., CNR PUSCH PUCCH = CNR PUSCH PUCCH .

Final
The pixel, subscriber, or mobile Mi reduces its transmission power so that the PUSCH and PUCCH C/N from it at its cell
is just enough to get the selected bearer.
If with P
Mi
Mi
Mi
= P Allowed AND CNR PUSCH PUCCH T
TX i ic
Mi
+ M PC , where T
B UL
TX i ic
Mi
B UL
is the bearer selection threshold, from
the reception equipment assigned to the cell TXi(ic), for the bearer selected for the pixel, subscriber, or mobile Mi.
The transmission power of Mi is reduced to determine the effective transmission power from the pixel, subscriber, or
mobile Mi as follows:
Mi
Mi
Mi
TXi ic
Mi
P Eff = Max PAllowed CNR PUSCH PUCCH T M + M PC P Min
i
UL
Mi
Mi
CNR PUSCH PUCCH is calculated again using P Eff .

Output
Mi
CNR PUSCH PUCCH : PUSCH and PUCCH C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
6.4.4.10 C/(I+N) and Bearer Calculation (UL)

The carrier signal to interference and noise ratio is calculated in three steps. First, Atoll calculates the received signal level
from each pixel, subscriber, or mobile at its serving cell using the effective power of the terminal used by the pixel, subscriber,
or mobile as explained in "Signal Level Calculation (UL)" on page 525. Next, Atoll calculates the uplink carrier to noise ratio as
explained in "C/N Calculation (UL)" on page 531. Finally, determines the uplink C/(I+N) by dividing the previously calculated
uplink C/N by the uplink noise rise value of the cell as calculated in "Noise Rise Calculation (UL)" on page 530.
The uplink noise rise can be set by the user manually for each cell or calculated using Monte Carlo simulations.
Input
Mi
CNR PUSCH PUCCH : PUSCH and PUCCH C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic) as calculated
in "C/N Calculation (UL)" on page 531.
TX i ic
NRUL
: Uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 530.
TX i ic
NRUL ICIC : ICIC uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 530.
T SU MIMO UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(c).
N FB
TX i ic
TX ic
i
the cell TXi(ic).
TX i ic
TX i ic
TX i ic
534
TX i ic
T B Lowest : Bearer selection threshold of the lowest bearer in the reception equipment assigned to the cell TXi(ic).

AT332_TRR_E0
M
P Allowed : Maximum allowed transmission power of the terminal used by the pixel, subscriber, or mobile Mi as
calculated in "Signal Level Calculation (UL)" on page 525.
M
M PC : Power control adjustment margin defined in the global network settings.
T B : Bearer selection thresholds of the bearers defined in the reception equipment used bythe cell TXi(ic).
Mi
or mobile Mi.
M
or mobile Mi.
Mi
N Ant TX : Number of transmission (uplink) antenna ports defined for the terminal used by the pixel, subscriber, or
mobile Mi.
TX ic
i
N Ant RX : Number of reception (uplink) antenna ports defined for the cell TXi(ic).
i
BLER B UL : Uplink block error rate read from the graphs available in the reception equipment assigned to the cell
TXi(ic).
Calculations
For any pixel, subscriber, or mobile Mi in the cell centre of the interfered cell TXi(ic), Atoll calculates the PUSCH and PUCCH C/
(I+N) as follows:
Mi
TX i ic
Mi
CINR PUSCH PUCCH = CNR PUSCH PUCCH NR UL
For any pixel, subscriber, or mobile Mi in the cell-edge of the interfered cell TXi(ic), Atoll calculates the PUSCH and PUCCH C/
(I+N) as follows:
M
TX ic
i
CINR PUSCH PUCCH = CNR PUSCH PUCCH NR UL ICIC

The bearers available for selection in the cell TXi(ic)s reception equipment are the ones:
Whose selection thresholds are less than the PUSCH and PUCCH C/(I+N) at Mi: T B CINR PUSCH PUCCH
Mi
Mi
TX i ic
thresholds and criteria, receive diversity, SU-MIMO diversity, or MU-MIMO diversity gain, G Div UL ,
M
TX ic
i
i
i
reception equipment assigned to the cell TXi(ic) for N Ant TX , N Ant RX , Mobility M i , BLER BUL .

UL
The additional uplink diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is also
applied. Therefore, the bearers available for selection are all the bearers defined in the reception equipment for
Mi
TX i ic
UL
Mi
T B G Div UL G Div CINR PUSCH PUCCH

Bearer Index
535

Peak RLC Throughput

From among the bearers available for selection, the selected bearer is the one with the highest uplink peak RLC


Once the bearer is known, the PUSCH and PUCCH C/(I+N) calculated above become:
M
TX ic
i
UL
CINR PUSCH PUCCH = CINR PUSCH PUCCH + G Div UL + G Div

TX i ic
Where G Div UL is the receive diversity, SU-MIMO diversity, or MU-MIMO diversity gain applied if the cell supports
MIMO and depending on the Mi diversity mode, the SU-MIMO and MU-MIMO threshold and criteria.
Uplink Bandwidth Allocation (No. of Used Frequency Blocks):
The uplink bandwidth allocation depends on the target defined for the scheduler used by the cell TXi(ic). The PUSCH
and PUCCH C/(I+N) calculated above is given for the total number of frequency blocks associated with the channel
TX i ic
bandwidth of the cell, i.e., N FB
. Bandwidth allocation is performed for all the pixels, subscribers, or mobiles in the
uplink, and may reduce the number of used frequency blocks in order to satisfy the selected target.
Full Bandwidth
transmission, there is no gain in the PUSCH and PUCCH C/(I+N).
Maintain Connection
The bandwidth used for transmission by a mobile is reduced only if the PUSCH and PUCCH C/(I+N) is not enough
to even access the lowest bearer. For example, as a mobile moves from good to bad radio conditions, the number
of frequency blocks used by it for transmission in uplink are reduced one by one in order to improve the PUSCH
and PUCCH C/(I+N). The calculation of the gain introduced by the bandwidth reduction is explained below.
Best Bearer
The bandwidth used for transmission by a mobile is reduced in order to improve the PUSCH and PUCCH C/(I+N)
enough to access the best bearer. For example, if using 5 frequency blocks, a mobile is able to access the best
bearer, and using 6 it would only get access to the second best, it will be assigned 5 frequency blocks as the used
uplink bandwidth. Although using 4 frequency blocks, its PUSCH and PUCCH C/(I+N) will be better than when using
5, the uplink bandwidth is not reduced to 4 because it does not provide any gain in terms of the bearer, i.e., the
mobile already has the best bearer using 5 frequency blocks. The calculation of the gain introduced by the
bandwidth reduction is explained below.
i.e., bearer with the highest index, with the highest peak RLC throughput, or with the highest effective RLC
throughput.
The uplink bandwidth allocation may result in the use of a number of frequency blocks which is less than the number
of frequency blocks associated with the channel bandwidth of the cell. The gain related to this bandwidth reduction
is applied to the PUSCH and PUCCH C/(I+N):
N TX i ic
Mi
Mi
FB
-
CINR PUSCH PUCCH = CINR PUSCH PUCCH + 10 Log ----------------Mi
All FB
Final
N FB UL
Mi
Min
TX i ic
Where N FB UL Service N FB UL N FB CC for any pixel, subscriber, or mobile Mi in the cell centre of the interfered
Min
Mi
TX i ic
cell TXi(ic), and N FB UL Service N FB UL N FB CE for any pixel, subscriber, or mobile Mi in the cell-edge of the
interfered cell TXi(ic).
536

AT332_TRR_E0
ICIC mode
Cell centre
Cell edge
TX ic
i
No FFR
N FB
Time-switched FFR
N FB
TX ic
i
N FB
TX ic
i
TX ic
i
N FB CEx
TX i ic
Hard FFR
TX i ic
N FB CEx
TX ic
i
Soft FFR
N FB
TX i ic
Partial soft FFR
N FB
TX ic
i
N FB CEx
TX ic
i
TX ic
i
N FB CEx
TX ic
N FB CEx
TX ic
TX ic
TX i ic
i
i
i
TX ic
i
TX ic
i
N FB CEx
TX ic
i
Uplink Power Control Adjustment:
Once the bandwidth allocation is performed, Atoll continues to work with the C/(I+N) given by the bandwidth
Mi
Mi
allocation, i.e., CINR PUSCH PUCCH = CINR PUSCH PUCCH .

Final
The pixel, subscriber, or mobile Mi reduces its transmission power so that the PUSCH and PUCCH C/(I+N) from it at its
cell is just enough to get the selected bearer.
If with P
Mi
Mi
Mi
= P Allowed AND CINR PUSCH PUCCH T
TX i ic
Mi
+ M PC , where T
B UL
TX i ic
Mi
B UL
is the bearer selection threshold, from
the reception equipment assigned to the cell TXi(ic), for the bearer selected for the pixel, subscriber, or mobile Mi.
Mi
Mi
Mi
TX i ic
Mi
P Eff = Max P Allowed CINR PUSCH PUCCH T M + M PC P Min

i
UL
Mi
Mi
CINR PUSCH PUCCH is calculated again using P Eff .

Output
Mi
CINR PUSCH PUCCH : PUSCH and PUCCH C/(I+N) from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
N FB UL : Number of frequency blocks used by the pixel, subscriber, or mobile Mi after uplink bandwidth allocation.
P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink.
Mi
Mi
Mi
6.4.5 Best Server Determination

In LTE, best server refers to a cell ("serving transmitter"-"reference cell" pair) that best covers a pixel, subscriber, or mobile
Mi and provides the best service.
537

Input
TX ic
i
C DLRS : Downlink reference signal level received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated
in "Signal Level Calculation (DL)" on page 502 using the terminal and service parameters ( L
,G
, L Ant , and L Body )
of Mi.
TX ic
i
E DLRS : Received downlink reference signal energy per resource element (RSRP) from any cell TXi(ic) at a pixel,
subscriber, or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 502.
T RSRP : Minimum RSRP defined for the cell TXi(ic).
T Selection : Cell selection threshold defined for the cell TXi(ic).
O Individual : Cell individual offset defined for the cell TXi(ic).
M HO
TX i ic
TX ic
i
TX i ic
TX i ic
: Handover margin defined for the cell TXi(ic).
TX i ic
p Layer : Priority defined for the layer assigned to for any cell TXi(ic).
N SCell
Max DL
: Maximum number of downlink secondary cells defined for the terminal used by the pixel, subscriber, or
mobile Mi.
Max UL
N SCell
: Maximum number of uplink secondary cells defined for the terminal used by the pixel, subscriber, or
mobile Mi.
Calculations
The serving cell selected for coverage predictions is based on the Standard serving cell selection method. The serving cell
selected for Monte Carlo simulations can also be based on the Random method instead of the Standard method. If no serving
cell is found for a mobile Mi, it is rejected for No Coverage.
The best server selection for any pixel, subscriber, or mobile Mi BSM is performed as follows:
i
Standard cell selection based on 3GPP specifications for connected mode mobility:
a. Qualification:
To qualify as potential servers, cells must fulfill the following requirements:
PRACH
preamble
format
a.
The cells layers must be supported by the pixel, subscriber, or mobile Mis service and terminal.
The speed defined in the pixel, subscriber, or mobile Mis mobility type must be less than or equal to the
maximum speed supported by the cells layers.
The pixel, subscriber, or mobile Mi must be located within the round-trip time distances corresponding to the
cells PRACH preamble format.
Cyclic prefix
Preamble sequence
Window
size
Guard
period
RTT distance
Tsa
Sec.
Ts
Sec.
Sec.
Sec.
Sec.
Metres
3168
0.00010
24576
0.00080
0.00090
0.00100
0.00010
14521
21024
0.00068
24576
0.00080
0.00148
0.00200
0.00052
77290
6240
0.00020
49152
0.00160
0.00180
0.00200
0.00020
29511
21024
0.00068
49152
0.00160
0.00228
0.00300
0.00072
107269
448
0.00001
4096
0.00013
0.00015
0.00017
0.00002
2811
The basic unit of time in LTE: Ts = 1/(15000 x 2048) seconds.
For potential serving cells that belong to layers of higher priorities, the RSRP received at the pixel, subscriber,
or mobile Mi must be higher than or equal to the cells Min RSRP plus the cell selection threshold:
TX ic
TX ic
TX ic
i
i
i
E DLRS T RSRP + Max 0 T Selection .
538
Cyclic
prefix +
preamble
sequence

AT332_TRR_E0
For the potential serving cells that belong to the layer of the lowest priority, the RSRP received at the pixel,
TX ic
i
TX ic
i
subscriber, or mobile Mi must be higher than or equal to the cells Min RSRP: E DLRS T RSRP .
b. Preselection:
From the list of cells that qualify as potential servers in step a., the cell that fulfills the following conditions is
preselected as the serving cell ( S 0 ):
The cell belonging to the highest priority layer, and
From which the pixel, subscriber, or mobile Mi receives the highest reference signal level or RSRP ( C DLRS or
TX i ic
TX i ic
E DLRS ) according to the defined best server selection criterion.

c. Final selection:
Among the cells that qualify as potential servers, other than the preselected server ( S 0 ), and fulfill the following
condition:
TX i ic
TX i ic
S0
S0
S0
E DLRS + O Individual E DLRS + O Individual + M HO

Atoll selects as the best server the cell from which the pixel, subscriber, or mobile Mi receives the highest
TX i ic
S0
TX i ic
S0
reference signal level or RSRP plus the cell individual offset ( C DLRS + O Individual or E DLRS + O Individual ),
according to the defined best server selection criterion.
If no cell fulfils the above condition, then the preselected server ( S 0 ) is selected as the best server.
Random cell selection:

a. Qualification:
To qualify as potential servers, cells must fulfill the following requirements:
The cells layers must be supported by the pixel, subscriber, or mobile Mis service and terminal.
The cells frequency band must be supported by the pixel, subscriber, or mobile Mis terminal.
The speed defined in the pixel, subscriber, or mobile Mis mobility type must be less than or equal to the
maximum speed supported by the cells layers.
The pixel, subscriber, or mobile Mi must be located within the round-trip time distances corresponding to the
cells PRACH preamble format (see table above).
The RSRP received at the pixel, subscriber, or mobile Mi must be higher than or equal to the cells Min RSRP:
TX i ic
TX i ic
E DLRS T RSRP .
b. Final selection:
From the list of cells that qualify as potential servers in step a., Atoll keeps only one potential server per layer, i.e.,
per layer the cell from which the pixel, subscriber, or mobile Mi receives the highest reference signal level or RSRP,
and then selects among these cells one cell as the best server at random.
For carrier aggregation, Atoll selects multiple servers by processing lists of potential servers according to the Standard or
Random cell selection method:
LTE users:
a. A list of potential serving cells whose cell type includes LTE
LTE-A users:
b. A list of potential primary serving cells whose cell type includes LTE and LTE-A PCell
c. A list of potential secondary serving cells whose may include LTE-A SCell DL and LTE-A SCell UL
Atoll selects the serving cell for LTE users from the list a. and a primary serving cell for LTE-A users from the remaining
list b.
Once a primary serving cell has been selected, Atoll eliminates the selected cell as well as any other co-channel cell
from list c. Here, co-channel cells are cells whose channels overlap the channel being used the primary serving cell.
539

In intra-eNode-B carrier aggregation, at this stage Atoll also eliminates cells belonging to
other eNode-Bs than that of the selected primary cell.
In group-based carrier aggregation, at this stage Atoll also eliminates cells not belonging
to the carrier aggregation groups to which the selected primary cell belongs.
If the primary serving cell belongs to more than one carrier aggregation groups, Atoll
searches for secondary serving cells in the in the first carrier aggregation group among
the largest carrier aggregation groups (most member cells) sorted alphabetically.
You can switch between carrier aggregation modes, using the Atoll.ini file. For more
information, see the Administrator Manual.
For LTE-A users with a primary serving cell of type LTE-A PCell selected from list b., Atoll selects secondary serving
cells from list c.
This step is carried out until either list c. is empty, or the numbers of downlink or uplink secondary serving cells
Max DL
assigned to the user become equal to the maximum numbers defined in the terminal properties ( N SCell
Max UL
N SCell
and
). Secondary cells are selected based on the reference signal level or RSRP, according to the defined best
server selection criterion. Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell
DL
activation threshold defined in the terminal reception equipment properties ( T SCell ) are activated for aggregation in
downlink. Similarly, only secondary cells whose PDSCH C/(I+N) and PUSCH C/(I+N) are both higher than or equal to the
DL
secondary cell activation threshold defined in the terminal and cell reception equipment properties ( T SCell and
UL
T SCell ), respectively, are activated for aggregation in uplink.

The primary and secondary serving cells once assigned to a mobile do not change during a Monte Carlo simulation.
For downlink or uplink CoMP, Atoll also determines additional CoMP servers of type LTE-A PCell:
From the CoMP set to which the primary serving cell belongs,
Within the best servers cell-edge region, and
Among cells using frequency channels with the same centre frequency.
The number of CoMP servers, including the primary serving cell, is limited to the CoMP transmission and reception set sizes
Max DL
( N CoMP
Max UL
and N CoMP
) defined for the CoMP set of the best server.
If the best server belongs to more than one CoMP set, Atoll searches for additional CoMP servers in the CoMP set that uses
the CoMP scheme providing the highest gains: non-coherent joint transmission then coherent joint transmission then
coordinated scheduling then dynamic point selection12. If the best server belongs to more than one CoMP set using the same
CoMP scheme, Atoll searches for additional CoMP servers in the first CoMP set among the largest CoMP sets (most member
cells) sorted alphabetically.
Atoll determines whether the pixel, subscriber, or mobile Mi is in the cell-edge or cell centre of TXi(ic) by:
Either calculating the difference between the total loss from the second best server and the best server, and
comparing it with the delta path loss threshold defined for the best server of the pixel, subscriber, or mobile Mi.
Therefore,
2ndBSM
L Total
pixel,
subscriber,
BS M 2ndBS M
i
i
+ 10 Log r O
or
BS M
mobile
BS M
Mi
is
considered
to
be
cell
edge
if
BS M
L Total L Path , and it is considered to be in cell centre otherwise. Here, L Total

2ndBS
is the total loss from Mis best server and L Total
Mi
is the total loss from Mis second best server calculated as

BS M
explained in "Signal Level Calculation (DL)" on page 452. L Path is the delta path loss threshold defined for the best
server of the pixel, subscriber, or mobile Mi.
12.
540
Or calculating the difference between the RSRP from the second best server and the best server, and comparing it
with the delta RSRP threshold defined for the best server of the pixel, subscriber, or mobile Mi.
Based on CoMP gains published in the R&D test results on CoMP used by the 3GPP (R1-111944, R1-111922, R1-112811)
and a research paper published in IEEE ComSoc Magazine February 2012.

AT332_TRR_E0
Therefore,
2ndBS
E DLRS
a
2ndBS
pixel,
M
subscriber,
BS
+ O Individual + 10 Log r O
BS
or
2ndBS
i
mobile
M
BS
Mi
is
BS
considered
BS
to
be
cell
edge
if
i
i
i
E
+ O Individual E DLRS , and it is considered to be in cell
DLRS
2ndBS
centre otherwise. Here, E DLRS is the RSRP from Mis best server and E DLRS
BS M
is the RSRP from Mis second best server

i
calculated as explained in "Signal Level Calculation (DL)" on page 452. O Individual is the cell individual offset defined
2ndBS M
BS M
for Mis best server and O Individual is the cell individual offset defined for Mis second best server. E DLRS is the delta
RSRP threshold defined for the best server of the pixel, subscriber, or mobile Mi.
If you set the cell edge calculation method to use RSRP rather than path losses, Atoll calculates the cell-edge regions
for CoMP by comparing the cell-edge margin with the difference between the best server RSRP and the second best
server RSRP belonging to the same CoMP set.
BSM 2ndBS M
i
i
rO
is the total channel overlap ratio between the best server and the second best server as calculated in "Co- and
Adjacent Channel Overlaps Calculation" on page 493.

Output
BS M : Best serving cell of the pixel, subscriber, or mobile Mi.

i
6.4.6 Throughput Calculation

Throughputs are calculated in two steps.
Calculation of uplink and downlink total resources in a cell as explained in "Calculation of Total Cell Resources" on
page 541.
Calculation of uplink and downlink UE capacities as explained in "Calculation UE Capacities" on page 549.
Calculation of throughputs as explained in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and
6.4.6.1 Calculation of Total Cell Resources

The total amount of resources in a cell is the number of modulation symbols that can be used for data transfer in each frame.
The total cell resources can be calculated separately for the downlink and uplink as described in:
6.4.6.1.1
"Calculation of Downlink Cell Resources" on page 541.

"Calculation of Uplink Cell Resources" on page 546.
Calculation of Downlink Cell Resources

Input
D CP
TX ic
i
TX i ic
TX i ic
is Extended).
TX i ic
N SD PDCCH : Number of PDCCH symbol durations per subframe defined in TXi(ic) frame configuration or, otherwise,
in the global network settings.
TX i ic
N FB
the cell TXi(ic).
TX i ic
541

TX ic
i
TX ic
i
TX i ic
TX i ic
bands.
TX i ic
TX i ic
TX i ic
TX i ic
Configuration
N SF DL
N TDD SSF
FDD
10
DSUUU-DSUUU
DSUUD-DSUUD
DSUDD-DSUDD
DSUUU-DSUUD
DSUUU-DDDDD
DSUUD-DDDDD
DSUDD-DDDDD
TX ic
i
Calculations
calculations below.
The number of modulation symbols (resource elements) per scheduler resource block is calculated as follows:
W FB
N SCa FB = --------F
The number of modulation symbols (resource elements) corresponding to the DwPTS per scheduler resource block in the TDD
special subframes is calculated as follows:
DwPTS
DwPTS
N Sym SSF = N SCa FB N SD SSF

DwPTS
Where N SD SSF is the number of DwPTS symbol durations (OFDM symbols) per special subframe, determined from the TDD
special subframe configuration according to the 3GPP specifications as follows:
542

AT332_TRR_E0
Special
Subframe
Configuration
DwPTS
GP
N SD SSF
DwPTS
N SD SSF

UpPTS
DwPTS
GP
UpPTS
N SD SSF
DwPTS
N SD SSF
10
10
11
10
12
10
11
GP
N SD SSF
GP
UpPTS
UpPTS
N SD SSF
The total number of modulation symbols (resource elements) in downlink is calculated as follows:
TX i ic
TX i ic
N Sym DL = N FB
TX i ic
TX i ic
N Sym SRB N SF DL + N Sym DwPTS
TX i ic
TX i ic
Where N Sym DwPTS = N FB
TX i ic
DwPTS
N TDD SSF N Sym SSF

TX i ic
The total downlink cell resources, i.e., R DL

TX i ic
R DL
TX i ic
TX i ic
, are calculated as follows:

TX i ic
TX i ic
TX i ic
= N Sym DL O DLRS O PSS O SSS O PBCH O PDCCH O DMRS

TX i ic
The downlink DwPTS resources, i.e., R DwPTS , are calculated as follows:

TX i ic
TX i ic
TX i ic
TX i ic
R DwPTS = N Sym DwPTS O DLRS DwPTS O PDCCH DwPTS

TX i ic
Where O DLRS is the overhead corresponding to the downlink reference signals, O PSS is the overhead corresponding to the
TX ic
i
primary synchronisation signals, O SSS is the overhead corresponding to the secondary synchronisation signals, O PBCH is the
TX i ic
overhead corresponding to the physical broadcast channel, and O PDCCH is the overhead corresponding to the physical
TX i ic
downlink control channel. O DMRS is the overhead corresponding to the UE-specific reference signals transmitted on the
logical antenna port 5 or the demodulation reference signals transmitted using antenna ports 7 and 8 or 7 through 14.
These overheads are calculated as follows:
Downlink reference signal overhead
The downlink reference signal overhead depends on the number of transmission antenna ports:
TX i ic
O DLRS
TX i ic
= N FB
TX i ic
TX i ic
TX i ic
Where O DLRS DwPTS = N FB
TX i ic
N DLRS SRB
TX i ic
TX i ic
N DLRS SRB N SF DL + O DLRS DwPTS
= 16
24
TX i ic
TX i ic
N DLRS DwPTS N TDD SSF ,

TX ic
i
if N Ant TX = 1
TX ic
i
if N Ant TX = 2
TX ic
i
TX i ic
And N DLRS DwPTS is determined from the table below:
543

Special
Subframe
Configuration

DwPTS
N SD SSF
10
11
12
10
11
TX ic
i
N Ant TX
TX ic
i
N DLRS DwPTS
DwPTS
N SD SSF
TX ic
i
TX ic
i
N Ant TX
N DLRS DwPTS
12
12
20
20
20
20
12
12
20
20
20
20
12
16
20
24
20
24
16
24
24
12
20
20
12
12
20
20
20
20
12
20
20
12
20
20
12
12
10
12
12
PSS and SSS overhead

The primary and secondary synchonisation signals are transmitted on 1 symbol duration each in the 1st and the 6th
downlink subframes, over the centre 6 frequency blocks. Therefore,
O PSS = 2 N FB SS PBCH N SCa FB = 144 symbols
O SSS = 2 N FB SS PBCH N SCa FB = 144 symbols
544

AT332_TRR_E0
PBCH overhead
The physical broadcast channel is transmitted on four symbol durations in the 1st downlink subframe over the center
6 frequency blocks. The physical broadcast channel overlaps with the downlink reference signals, therefore, some
downlink reference signal modulation symbols are subtracted:
216 for extended cyclic prefix
240 for normal cyclic prefix
PDCCH overhead
The physical downlink control channel can be transmitted over up to 4 symbol durations in each subframe. The
number of symbol durations for the PDCCH is defined in the global network settings. The PDCCH overlaps some
downlink reference signal symbols. These downlink reference signal symbols are subtracted from the PDCCH
overhead:
TX ic
i
if N SD PDCCH = 0 :
TX i ic
O PDCCH = 0
TX ic
TX ic
i
i
TX ic
TX ic
TX ic
i
i
i
O PDCCH = N SD PDCCH N SCa FB 4 N FB
TX i ic
Where O PDCCH DwPTS =
TX i ic
TX i ic
TX ic
TX ic
i
N i
TX i ic
N TDD SSF
Otherwise:
TX ic
TX ic
TX ic
TX ic
i
i
i
i
O PDCCH = N SD PDCCH N SCa FB 2 Min 4 N Ant TX N FB
TX ic
TX i ic
TX i ic
TX ic
TX ic
TX ic
i
i
i
i
Where O PDCCH DwPTS = Min 2 N SD PDCCH N SCa FB 2 Min 4 NAnt TX N FB
TX i ic
N TDD SSF
UE-specific and demodulation reference signal overhead

UE-specific reference signals are transmitted on the logical antenna port 5, DMRS are transmitted on antenna ports 7
and 8, or on 7 through 14. The DMRS overhead is:
TX ic
i
Without smart antennas and MIMO: O DMRS = 0
With smart antennas and without MIMO: O DMRS = 12 N FB
TX i ic
TX i ic
TX i ic
N SF DL
When DMRS transmitted on antenna ports 7 and 8 or on 7 through 14 is considered, i.e., the following Atoll.ini option
is set:
[LTE]
ApplyDMRSOverhead = 1
TX ic
i
TX ic
i
TX ic
i
With smart antennas and with MIMO: O DMRS = 24 N FB
N SF DL
Without smart antennas and with SU-MIMO or MU-MIMO and N Ant TX 4 : O DMRS = 24 N FB
TX i ic
TX i ic
TX i ic
TX i ic
N SF DL
TX i ic
Once R DL
is known, it is scaled down according to the ICIC mode used by the cell TXi(ic) depending on whether the
downlink cell resources are being calculated for a cell-centre or cell-edge pixel, subscriber, or mobile.
TX i ic
R DL
TX i ic
= R DL
ICIC
ABS
f Scaling f Scaling
ICIC
f Scaling is calculated as follows for the different ICIC modes:
545

ICIC
ICIC cell resource scaling factor f Scaling for
ICIC mode
Cell centre
No FFR
Cell edge
Time-switched FFR
1
TX ic
i
N FB CEx
-------------------TX i ic
N FB
TX i ic
TX i ic
Hard FFR
N FB CEx
-------------------TX i ic
N FB
Soft FFR
N FB
N FB CEx
-----------------------------------------TX i ic
N FB
Partial soft FFR
N FB
-----------------------------------------------------------------------------------------------------TX i ic
N FB
TX i ic
TX i ic
TX i ic
TX i ic
N FB CEx
-------------------TX i ic
N FB
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
N FB CEx
-------------------TX i ic
N FB
TX i ic
TX i ic
N FB CEx
-------------------TX i ic
N FB
TX i ic
ABS
f Scaling is calculated as follows:
ABS
f Scaling
1
Cell centre
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
SFP DL
R DL
R DwPTS +
SFP SSF R DwPTS
=
------------------------------------------------------------------------------------------------------------------------------------------------1
1
- Cell edge
TX ic
i
80 R DL

TX i ic
ABS
SFPDL
TX i ic
R DL
TX ic
i
R DwPTS +
TX i ic
SFPSSF
TX ic
i
R DwPTS
1
1
f Scaling = ------------------------------------------------------------------------------------------------------------------------------------------------TX i ic
80 R DL
TX i ic
For more information on SFP DL
TX i ic
and SFP SSF
, see "Subframe Pattern Collision Calculation" on page 497.
Output
6.4.6.1.2
TX i ic
R DL
: Amount of downlink resources in the cell TXi(ic).
Calculation of Uplink Cell Resources

Input
D CP
TX ic
i
N FB PUCCH
TX i ic
TX i ic
is Extended).
: Average number of PUCCH frequency blocks per frame defined in TXi(ic) frame configuration or,
otherwise, in the global network settings.
546
TX i ic

AT332_TRR_E0
TX ic
i
N FB
the cell TXi(ic).
TX ic
i
TX i ic
TX i ic
TX i ic
N SF UL : Number of uplink subframes in the frame for the cell TXi(ic). It is equal to 10 for FDD frequency bands, and
TX i ic
bands.
TX i ic
TX i ic
N SF UL and N TDD SSF are determined as follows:

TX i ic
TX i ic
Configuration
N SF UL
N TDD SSF
FDD
10
DSUUU-DSUUU
DSUUD-DSUUD
DSUDD-DSUDD
DSUUU-DSUUD
DSUUU-DDDDD
DSUUD-DDDDD
DSUDD-DDDDD
UpPTS is used for SRS (sounding reference signals) if the UpPTS duration is 1 OFDM
symbol, and for SRS and PRACH if the UpPTS duration is 2 OFDM symbols. Therefore, the
uplink cell capacity can be determined without considering the UpPTS symbols.
Calculations
calculations below.
The number of modulation symbols (resource elements) per resource block is calculated as follows:
W FB
N SCa FB = --------F
The total number of modulation symbols (resource elements) in uplink is calculated as follows:
TX i ic
TX i ic
N Sym UL = N FB
TX i ic
N Sym SRB N SF UL
TX i ic
The total uplink cell resources, i.e., R UL

TX i ic
R UL
TX i ic
TX i ic
TX i ic
, are calculated as follows:

TX i ic
= N Sym UL O PUCCH O ULSRS O ULDRS
547

TX ic
i
TX ic
i
Where O PUCCH is the PUCCH overhead, O ULSRS is the overhead corresponding to the uplink sounding reference signals, and
TX ic
i
O ULDRS is the overhead corresponding to the uplink demodulation reference signals. These control channel overheads are
Calculations of uplink control channel overheads
The PUCCH overhead is calculated as follows:
TX i ic
TX i ic
TX i ic
O PUCCH = N FB PUCCH N Sym SRB N SF UL

The uplink sounding reference signals are transmitted on 1 symbol duration in each uplink subframe. Therefore,
TX ic
TX ic
N SCa FB
i
i
- N Sym
O ULSRS = -------------------- UL
N Sym SRB
The uplink demodulation reference signals are transmitted on two symbol durations in each uplink subframe.
Therefore,
TX i ic
TX i ic
N SCa FB
- N Sym
O ULDRS = 2 -------------------- UL
N Sym SRB
TX i ic
Once R UL
is known, it is scaled down according to the ICIC mode used by the cell TXi(ic) depending on whether the uplink
cell resources are being calculated for a cell-centre or cell-edge pixel, subscriber, or mobile.
TX i ic
R UL
TX i ic
= R UL
ICIC
ABS
f Scaling f Scaling
ICIC
f Scaling is calculated as follows for the different ICIC modes:

ICIC
ICIC cell resource scaling factor f Scaling for
ICIC mode
Cell centre
Cell edge
Time-switched FFR
N FB CEx
-------------------TX i ic
N FB
Hard FFR
N FB CEx
-------------------TX i ic
N FB
Soft FFR
N FB
N FB CEx
-----------------------------------------TX ic
i
N FB
Partial soft FFR
N FB
-----------------------------------------------------------------------------------------------------TX i ic
N FB
No FFR
TX i ic
TX i ic
TX ic
i
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
N FB CEx
-------------------TX i ic
N FB
TX ic
i
TX i ic
TX i ic
TX ic
i
N FB CEx
-------------------TX ic
i
N FB
TX i ic
ABS
f Scaling is calculated as follows:
ABS
f Scaling
548
1
Cell centre
TX i ic
=
SFP UL
1
------------------------------ Cell edge
80
TX i ic
N FB CEx
-------------------TX i ic
N FB

AT332_TRR_E0
TX ic
i
SFPUL
ABS
1
f Scaling = ----------------------------80
TX ic
i
For more information on SFP UL
, see "Subframe Pattern Collision Calculation" on page 497.
Output
TX i ic
R UL
: Amount of uplink resources in the cell TXi(ic).
6.4.6.2 Calculation UE Capacities

The UE category parameters define the maximum throughput that can be supported by a UE in downlink and uplink. The UE
capacities are calculated for the downlink and uplink as described in:
6.4.6.2.1
"Calculation of Downlink UE Capacity" on page 549.

"Calculation of Uplink UE Capacity" on page 550.
Calculation of Downlink UE Capacity

Input
D Frame : Frame duration.
N TBB TTI : Maximum number of transport block bits per TTI (subframe) in downlink defined for a UE category.
Max DL
TX i ic
TX i ic
bands.
TX ic
i
TX ic
i

TX ic
i
TX ic
i
Configuration
N SF DL
N TDD SSF
FDD
10
DSUUU-DSUUU
DSUUD-DSUUD
DSUDD-DSUDD
DSUUU-DSUUD
DSUUU-DDDDD
DSUUD-DDDDD
DSUDD-DDDDD
Calculations
In LTE, the maximum throughput that can be supported by a user equipment is defined through its UE category parameter
Transport Block Size. This is the maximum number of transport block bits that the UE can carry per subframe.
The downlink UE capacity in terms of the maximum throughput supported by a UE in downlink is calculated as follows:
TX ic
Max
TP UE DL
Max DL
N TBB TTI
TX ic
i
N i
SF DL + N TDD SSF
---------------------------------------------------D Frame
The maximum transport block sizes defined by the 3GPP for different UE categories correspond to the following maximum
throughput capacities in FDD:
UE Category
Max DL
N TBB TTI (bits/TTI)
10296
51024
102048
150752
299552
301504
301504
2998560
549

UE Category
Max
TP UE DL (Mbps)
10.296
51.024
102.048
150.752
299.552
301.504
301.504
2998.560
Output
Max
TP UE DL : Maximum downlink throughput capacity of a UE category.
6.4.6.2.2
Calculation of Uplink UE Capacity

Input
N TBB TTI : Maximum number of transport block bits per TTI (subframe) in uplink defined for a UE category.
N SF UL : Number of uplink subframes in the frame for the cell TXi(ic). It is equal to 10 for FDD frequency bands, and
Max UL
TX i ic
TX i ic
bands.
TX i ic
TX i ic
N SF UL and N TDD SSF are determined as follows:

TX i ic
TX i ic
Configuration
N SF UL
N TDD SSF
FDD
10
DSUUU-DSUUU
DSUUD-DSUUD
DSUDD-DSUDD
DSUUU-DSUUD
DSUUU-DDDDD
DSUUD-DDDDD
DSUDD-DDDDD
Calculations
In LTE, the maximum throughput that can be supported by a user equipment is defined through its UE category parameter
Transport Block Size. This is the maximum number of transport block bits that the UE can carry per subframe.
The uplink UE capacity in terms of the maximum throughput supported by a UE in uplink is calculated as follows:
Max
TP UE UL
Max UL
N TBB TTI
TX i ic
N SF UL
----------------D Frame
The maximum transport block sizes defined by the 3GPP for different UE categories correspond to the following maximum
throughput capacities in FDD:
UE Category
Max UL
5160
25456
51024
51024
75376
51024
102048
1497760
Max
5.16
25.456
51.024
51.024
75.376
51.024
102.048
1497.760
N TBB TTI (bits/TTI)

TP UE UL (Mbps)
Output
550
Max
TP UE UL : Maximum uplink throughput capacity of a UE category.

AT332_TRR_E0
6.4.6.3 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput,

and Per-user Throughput Calculation
Channel throughputs are calculated for the entire channel resources allocated to the pixel, subscriber, or mobile Mi. Cell
capacities are similar to channel throughputs but upper-bound by the maximum downlink and uplink traffic loads. Allocated
bandwidth throughputs are calculated for the number of used frequency blocks in uplink allocated to the pixel, subscriber, or
mobile Mi. Per-user throughputs are calculated by dividing the downlink cell capacities or uplink allocated bandwidth
throughputs by the average number of downlink or uplink users defined for the cell, respectively.
Input
TX ic
i
TL DL Max : Maximum downlink traffic load for the cell TXi(ic).
TL UL Max : Maximum uplink traffic load for the cell TXi(ic).
R DL
TX i ic
TX i ic
: Amount of downlink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on
page 541.
TX i ic
R UL
page 541.
M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the downlink in

M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the uplink in "C/
(I+N) and Bearer Calculation (UL)" on page 534.

T SU MIMO UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(ic).
T SU MIMO DL : SU-MIMO threshold defined in the reception equipment of the pixel, subscriber, or mobile Mi.
T MU MIMO UL : MU-MIMO threshold defined in the reception equipment of the cell TXi(ic).
T MU MIMO DL : MU-MIMO threshold defined in the reception equipment of the pixel, subscriber, or mobile Mi.
G MU MIMO UL : Average number of co-scheduled MU-MIMO users in uplink for the cell TXi(ic).
G MU MIMO DL : Average number of co-scheduled MU-MIMO users in downlink for the cell TXi(ic).
: Amount of uplink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on
i
B DL
i
B UL
TX i ic
Mi
TX i ic
TX ic
i
TX i ic
TX i ic
TX ic
i
i
BLER B DL : Downlink block error rate read from the BLER vs. CINR PDSCH graph available in the reception equipment

assigned to the terminal used by the pixel, subscriber, or mobile Mi.
i
i
BLER B UL : Uplink block error rate read from the BLER vs. CINR PUSCH PUCCH graph available in the reception

equipment assigned to the cell TXi(ic).
DL
T SCell : Secondary cell activation threshold of the reception equipment assigned to the pixel, subscriber, or mobile Mi.
T SCell : Secondary cell activation threshold of the reception equipment assigned to the cell TXi(ic).
f TP Scaling : Throughput scaling factor defined in the properties of the service used by the pixel, subscriber, or mobile
UL
Mi
Mi.
M
TP Offset : Throughput offset defined in the properties of the service used by the pixel, subscriber, or mobile Mi.
N FB
TX i ic
cell TXi(ic).
N FB UL : Number of frequency blocks used by the pixel, subscriber, or mobile Mi after uplink bandwidth allocation as
calculated in "C/(I+N) and Bearer Calculation (UL)" on page 534.
TX ic
i
N Users DL : Number of users connected to the cell TXi(ic) in downlink.
551

TX ic
i
N Users UL : Number of users connected to the cell TXi(ic) in uplink.
Calculations
Downlink:
TX i ic
R DL
Mi
Mi
B DL
Peak RLC Channel Throughput: CTP P DL = --------------------------------D Frame

In the above formula, the actual value of D Frame is used to calculate the channel throughput for coverage predictions,
while D Frame = 1 sec for Monte Carlo simulations.
TX ic
i
For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG DL read
Mi
from the scheduler properties for the bearer B DL , Mobility M i , and the number of users connected to the cell in
downlink.
TX i ic
R DL
Mi
Mi
B DL
TX ic
i
CTP P DL = --------------------------------- G MUG DL
D Frame
TX i ic
Mi
Max
G MUG DL = 1 if CINR PDSCH CINR MUG

If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph, it is
interpolated from the gain values available for the numbers of users just less than and just greater than the actual
number of users.
MIMO SU-MIMO Gain:
If the cell supports MIMO, and according to the Mi diversity mode depending on the SU-MIMO threshold and criterion,
Max M i
the SU-MIMO gain, G SU MIMO DL , corresponding to the bearer is applied to its efficiency. The gain is read from the
TX i ic
Mi
properties of the reception equipment assigned to the pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX ,
M
i
Mobility M i , BLER B DL .
Atoll also takes into account the SU-MIMO Gain Factor f SU MIMO defined for the clutter class where the pixel,
subscriber, or mobile Mi is located.
Mi
Max M
DL
Mi
i
1 + f SU MIMO G SU MIMO DL 1
DL
If the Max SU-MIMO Gain for the exact value of the C/(I+N) is not available in the table, it is interpolated from the gain
values available for the C/(I+N) just less than and just greater than the actual C/(I+N).
MIMO MU-MIMO Gain (for throughput coverage predictions):
If the cell supports MU-MIMO, and according to the Mi diversity mode depending on the MU-MIMO threshold and
TX i ic
criterion, the MU-MIMO gain, G MU MIMO DL , which is the average number of co-scheduled users, is applied to the
channel throughput.
Mi
Mi
TX i ic
CTP P DL = CTP P DL G MU MIMO DL
552
i
i
i
Effective RLC Channel Throughput: CTP E DL = CTP P DL 1 BLER B DL
M
i
CTP A DL
i
CTP E DL
Application Channel Throughput:
Peak RLC Cell Capacity: Cap P DL = CTP P DL TL DL Max
Mi
Mi
TX i ic
Mi
M
f TP Scaling
i
- TP Offset
-----------------------100

AT332_TRR_E0
i
i
i
Effective RLC Cell Capacity: Cap E DL = Cap P DL 1 BLER B DL
Application Cell Capacity:
i
Cap A DL
i
Cap E DL
M
f TP Scaling
i
------------------------- TP Offset
100
i
Mi
Cap P DL
Peak RLC Throughput per User: PUTP P DL = ----------------------TX i ic
N Users DL
Mi
Cap E DL
= ----------------------TX i ic
N Users DL
Mi
Effective RLC Throughput per User: PUTP E DL
Application Throughput per User:
i
PUTP A DL
i
PUTP E DL
Mi
M
f TP Scaling
i
- TP Offset
-----------------------100
Carrier Aggregation and Downlink Non-coherent Joint Transmission CoMP:

Aggregated throughputs are calculated by summing the throughputs from each serving
cell taking part in carrier aggregation or downlink non-coherent joint transmission for
any LTE-A pixel, subscriber, or mobile. If the sum of the throughputs exceeds the
maximum throughput supported by the UE category, the aggregated throughput is
scaled down by the following ratio:
Mi
Max
CTP P DL
Min TP UE DL
TX i ic
r = --------------------------------------------------------------------------Mi
CTP P DL
TX i ic
Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell
DL
activation threshold ( T SCell ) defined in the terminal reception equipment properties are
activated for aggregation.
Uplink:
TX ic
i
Mi
R UL
M
i
B UL
Peak RLC Channel Throughput: CTP P UL = --------------------------------D Frame

TX i ic
For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG UL read
Mi
from the scheduler properties for the bearer B UL , Mobility M i , and the number of users connected to the cell in
uplink.
TX i ic
R UL
Mi
Mi
B UL
TX ic
i
CTP P UL = --------------------------------- G MUG UL
D Frame
TX i ic
Mi
Max
G MUG UL = 1 if CINR PUSCH PUCCH CINR MUG

number of users.
553

MIMO SU-MIMO Gain:

If the cell supports MIMO, and according to the Mi diversity mode depending on the SU-MIMO threshold and criterion,
Max TX ic
i
the SU-MIMO gain, G SU MIMO UL , corresponding to the bearer is applied to its efficiency. The gain is read from the
TX ic
i
i
i
properties of the reception equipment assigned to the TXi(ic) for N Ant RX , N Ant TX , Mobility M i , BLER B UL .

i
B UL
Max TX ic
i
B UL
i
1 + f SU MIMO G SU MIMO UL 1
MIMO MU-MIMO Gain (for throughput coverage predictions):
If the cell supports MU-MIMO, and according to the Mi diversity mode depending on the MU-MIMO threshold and
TX ic
i
criterion, the MU-MIMO gain, G MU MIMO UL , which is the average number of co-scheduled users, is applied to the
channel throughput.
Mi
Mi
TX i ic
CTP P UL = CTP P UL G MU MIMO UL
i
i
i
Effective RLC Channel Throughput: CTP E UL = CTP P UL 1 BLER B UL

Mi
Mi
f TP Scaling
- TP Offset
= CTP E UL -----------------------100
Mi
Mi
Application Channel Throughput: CTP A UL
Peak RLC Cell Capacity: Cap P UL = CTP P UL TL UL Max
i
i
i
Effective RLC Cell Capacity: Cap E UL = Cap P UL 1 BLER B UL
Mi
TX i ic
Mi
Mi
Application Cell Capacity: Cap A UL
Mi
Mi
f TP Scaling
- TPOffset
= Cap E UL -----------------------100
Mi
Mi
Peak RLC Allocated Bandwidth Throughput: ABTP P UL
i
i
i
Effective RLC Allocated Bandwidth Throughput: ABTP E UL = ABTP P UL 1 BLER B UL
Mi
Application Allocated Bandwidth Throughput: ABTP A UL

Mi
Peak RLC Throughput per User: PUTP P UL
Mi
Mi
f TP Scaling
- TP Offset
= ABTP E UL -----------------------100
Mi
Cap M i
M
P UL
- ABTP P i UL
= Min ---------------------- TXi ic
N Users UL
i
PUTP E UL
Cap M i
M
E UL
- ABTP E i UL
= Min ----------------------TX i ic
N Users UL
Effective RLC Throughput per User:
Mi
Mi
Mi
f TP Scaling
Application Throughput per User: PUTP A UL = PUTP E UL ------------------------- TPOffset
100
Mi
554
Mi
N FB UL
= CTP P UL ----------------TX i ic
N FB
Mi

AT332_TRR_E0

any LTE-A pixel, subscriber, or mobile. If the sum of the throughputs exceeds the
maximum throughput supported by the UE category, the aggregated throughput is
scaled down by the following ratio:
M
Max
i
CTP P UL
Min TP UE UL
TX i ic
r = --------------------------------------------------------------------------M
i
CTP P UL
TX i ic
Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell
DL
activation threshold ( T SCell ) defined in the terminal reception equipment properties and
UL
PUSCH C/(I+N) is higher than or equal to the secondary cell activation threshold ( T SCell )
defined in the cell reception equipment properties are activated for aggregation.
Output
Mi
CTP P DL : Downlink peak RLC channel throughput at the pixel, subscriber, or mobile Mi.
CTP E DL : Downlink effective RLC channel throughput at the pixel, subscriber, or mobile Mi.
CTP A DL : Downlink application channel throughput at the pixel, subscriber, or mobile Mi.
Cap P DL : Downlink peak RLC cell capacity at the pixel, subscriber, or mobile Mi.
Cap E DL : Downlink effective RLC cell capacity at the pixel, subscriber, or mobile Mi.
Cap A DL : Downlink application cell capacity at the pixel, subscriber, or mobile Mi.
PUTP P DL : Downlink peak RLC throughput per user at the pixel, subscriber, or mobile Mi.
PUTP E DL : Downlink effective RLC throughput per user at the pixel, subscriber, or mobile Mi.
PUTP A DL : Downlink application throughput per user at the pixel, subscriber, or mobile Mi.
CTP P UL : Uplink peak RLC channel throughput at the pixel, subscriber, or mobile Mi.
CTP E UL : Uplink effective RLC channel throughput at the pixel, subscriber, or mobile Mi.
CTP A UL : Uplink application channel throughput at the pixel, subscriber, or mobile Mi.
Cap P UL : Uplink peak RLC cell capacity at the pixel, subscriber, or mobile Mi.
Cap E UL : Uplink effective RLC cell capacity at the pixel, subscriber, or mobile Mi.
Cap A UL : Uplink application cell capacity at the pixel, subscriber, or mobile Mi.
ABTP P UL : Uplink peak RLC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
ABTP E UL : Uplink effective RLC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
ABTP A UL : Uplink application allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
PUTP P UL : Uplink peak RLC throughput per user at the pixel, subscriber, or mobile Mi.
PUTP E UL : Uplink effective RLC throughput per user at the pixel, subscriber, or mobile Mi.
PUTP A UL : Uplink application throughput per user at the pixel, subscriber, or mobile Mi.
Mi
Mi
M
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
555


Atoll LTE module includes a number of scheduling methods which can be used for scheduling and radio resource allocation
during Monte Carlo simulations. These resource allocation algorithms are explained in "Scheduling and Radio Resource
Allocation" on page 556 and the calculation of user throughputs is explained in "User Throughput Calculation" on page 565.
6.4.7.1 Scheduling and Radio Resource Allocation

Input
TX i ic
N Users Max : Maximum number of users defined for the cell TXi(ic).
p QCI : QCI priority of the service accessed by a mobile Mi.
p Service : User-defined priority of the service accessed by a mobile Mi.
TPD Min DL : Downlink minimum throughput demand for the service accessed by a mobile Mi.
TPD Min UL : Uplink minimum throughput demand for the service accessed by a mobile Mi.
TPD Max DL : Downlink maximum throughput demand for the service accessed by a mobile Mi.
TPD Max UL : Uplink maximum throughput demand for the service accessed by a mobile Mi.
G MU MIMO UL : Average number of co-scheduled MU-MIMO users in uplink for the cell TXi(ic).
G MU MIMO DL : Average number of co-scheduled MU-MIMO users in downlink for the cell TXi(ic).

M : Bearer efficiency (bits/symbol) of the bearer assigned to the mobile Mi in the uplink in "C/(I+N) and Bearer

: Bearer efficiency (bits/symbol) of the highest bearer of the service being used by the mobile Mi in the
M
TX i ic
TX i ic
Mi
M
Mi
Mi
Mi
Mi
TX ic
i
TX i ic
Mi
B DL
: Bearer efficiency (bits/symbol) of the bearer assigned to the mobile Mi in the downlink in "C/(I+N) and Bearer
i
B UL
i
DL Highest
downlink.
M
i
B UL Highest
: Bearer efficiency (bits/symbol) of the highest bearer of the service being used by the mobile Mi in the
uplink.
TX ic
i
i
BLER BDL : Downlink block error rate read from the BLER vs. CINR PDSCH graph available in the reception equipment
assigned to the terminal used by the mobile Mi.
i
i
BLER BUL : Uplink block error rate read from the BLER vs. CINR PUSCH PUCCH graph available in the reception

DL
T SCell : Secondary cell activation threshold of the reception equipment assigned to the mobile Mi.
T SCell : Secondary cell activation threshold of the reception equipment assigned to the cell TXi(ic).
f TP Scaling : Throughput scaling factor defined in the properties of the service used by the mobile Mi.
TP Offset : Throughput offset defined in the properties of the service used by the mobile Mi.
CTP P DL : Downlink peak RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
UL
Mi
Mi
Mi
page 541.
556

AT332_TRR_E0
CTP E DL : Downlink effective RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 541.
CTP P UL : Uplink peak RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 541.
Mi
CTP E UL : Uplink effective RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 541.
Mi
ABTP P UL : Uplink peak RLC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput
Mi
ABTP E UL : Uplink effective RLC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput
Max
TP UE DL : Maximum downlink throughput capacity of the UE category of the mobile Mi as calculated in "Calculation
of Downlink UE Capacity" on page 549.
Max
TP UE UL : Maximum uplink throughput capacity of the UE category of the mobile Mi as calculated in "Calculation of
Uplink UE Capacity" on page 550.
Calculations
The following calculations are described for any cell TXi(ic) containing the users Mi for which it is the best server.
Mobile Selection:
TX ic
i
The scheduler selects N Users mobiles for the scheduling and RRM process. If the Monte Carlo user distribution has generated
TX i ic
a number of users which is less than N Users Max , the scheduler keeps all the mobiles generated for the cell TXi(ic).
TX ic
TX ic
TX ic
i
i
i
N Users = Min N Users Max N Users Generated
Sel
For a cell, mobiles M i
TX ic
i
N Users are selected for RRM by the scheduler.
Calculation of Actual Minimum and Maximum Throughput Demands:

If the service maximum throughput demand downgrading is active (for more information, see the Administrator Manual), the
maximum throughput demand of each user will be downgraded as follows:
Downlink:
Uplink:
Sel
Mi
TPD Max DL
Sel
Mi
TPD Max UL
Sel
M
Sel
Sel
i
Mi
Mi
B DL
-----------------------------= Max TPD Min DL TPD Max DL
Sel
Mi
B DL Highest
Sel
Mi
Sel
Sel
Mi
Mi
B UL
= Max TPD Min UL TPD Max UL -------------------------------
Sel
Mi
B UL Highest
Then, depending on the selected target throughput of the scheduler assigned to the cell TXi(ic), the actual minimum and
maximum throughput demands can be considered as the peak RLC, effective RLC, or application throughput. Therefore:
Target Throughput = Peak RLC Throughput

M
Sel
i
Sel
i
Downlink: TPD Min DL , TPD Max DL

Sel
Sel
Mi
Mi
Mi
Uplink: TPD Min UL , Min TPD Max UL ABTP P UL
Target Throughput = Effective RLC Throughput
557

Downlink:
Sel
i
TPD Min DL
M
Sel
i
Sel
Sel
i
M
TPD Min DL
TPD Max DL
i
= --------------------------------------------- , TPD Max DL = --------------------------------------------Sel
Sel
Mi
Mi
1 BLER B DL
1 BLER BDL
Sel
Uplink:
Sel
Mi
TPD Min UL
Sel
Mi
TPD Min UL
= ---------------------------------------------,
Sel
Mi
1 BLER B UL
Sel
Mi
TPD Max UL
Mi
Mi
Min TPD Max UL ABTP P UL
= ------------------------------------------------------------------------Sel
Mi
1 BLER B UL
Target Throughput = Application Throughput

Sel
Mi
Sel
Mi
Mi
Sel
Mi
Sel
Mi
Mi
TPD Min DL + TP Offset
TPD Max DL + TP Offset
- , TPD Max DL = ----------------------------------------------------------------------------Downlink: TPD Min DL = ----------------------------------------------------------------------------Sel
Sel
Mi Mi
Mi Mi
1 BLER B DL f TP Scaling
Sel
Mi
Sel
Mi
Mi
TPD Min UL + TP Offset

-,
Uplink: TPD Min UL = ----------------------------------------------------------------------------Sel
Mi Mi
1 BLER B UL f TP Scaling
Sel
Mi
Mi
Mi
Min TPD Max UL ABTPP UL + TP Offset

Sel
Mi
TPD MaxUL = -------------------------------------------------------------------------------------------------Sel
Mi Mi
1 BLER B UL f TP Scaling
The Min() function selects the lower of the two values. This calculation is performed in order to limit the maximum uplink
throughput demand to the maximum throughput that a user can get in uplink using the allocated bandwidth (number of
frequency blocks) calculated for it in "C/(I+N) and Bearer Calculation (UL)" on page 534.
Resource Allocation for Minimum Throughput Demands:
For their minimum throughput demands, LTE-A users are only scheduled on their primary
serving cells. This is valid for carrier aggregation as well as for downlink non-coherent
joint transmission CoMP.
Sel
1. Atoll sorts the M i
Sel
Mi
TX i ic
Sel
Mi
N Users in order of decreasing effective service priority (combination of p QCI and p Service ).
The mobiles are sorted first in the order of decreasing QCI priority (as listed in the table below) and then in the order
of decreasing user-defined service priority within a QCI. For example:
QoS class identifier
QCI priority
Sel
Mi
Sel
Mi
p Service
:
:
558
Sel
Mi
p QCI
i
:

AT332_TRR_E0
Sel
Mi
Sel
i
p QCI
:
4
0
5
0
6
i
:
Sel
:
:
= 1 up to M i
:
:
Sel
Sel
i
2. Starting with M i
p Service
NULL
0
TX i ic
= N Users , Atoll allocates the downlink and uplink resources required to satisfy
each users minimum throughput demands in downlink and uplink as follows:

Sel
Mi
R Min DL
Sel
Mi
Sel
Sel
Mi
Mi
TPD Min DL
TPD Min UL
= -------------------------- and R Min UL = -------------------------Sel
Mi
Sel
Mi
CTP P DL
CTP P UL
3. Atoll stops the resource allocation in downlink or uplink,
When/If in downlink
Sel
Mi
TX i ic
R Min DL = TL DL Max , i.e., the resources available in downlink have been used up for
Sel
Mi
satisfying the minimum throughput demands of the mobiles.
When/If in uplink
Sel
Mi
TX i ic
R Min UL = TL UL Max , i.e., the resources available in uplink have been used up for
Sel
Mi

4. Mobiles which are active DL+UL must be able to get their minimum throughput demands in both UL and DL in order
to be considered connected DL+UL. If an active DL+UL mobile is only able to get its minimum throughput demand in
one direction, it is rejected, and the resources that were allocated to it in the one direction in which it was able to get
a throughput are allocated to other mobiles.
Sel
Max
Mi
TP UE DL
- or
5. Mobiles with minimum throughput demands higher than their UE capacities, i.e., R Min DL -------------------Sel
Mi
CTP P DL
Sel
Max
Mi
TP UE UL
- , are rejected due to No Service.
R Min UL -------------------Sel
Mi
CTP P UL
559

6. Mobiles which are active UL and whose minimum throughput demand in UL is higher than the uplink allocated
M
Sel
i
Sel
i
bandwidth throughput ( TPD Min UL ABTP P UL ) are rejected due to Resource Saturation.
7. If
Sel
i
TX ic
i
Sel
Mi
R Min DL TL DL Max or
Sel
i
TX ic
i
R Min UL TL UL Max , and all the minimum throughput resources demanded by
Sel
Mi
the mobiles have been allocated, Atoll goes to the next step for allocating resources to satisfy the maximum
throughput demands.
Backhaul Saturation:
If at this stage, a sites downlink or uplink effective RLC aggregate throughput exceeds its maximum downlink or uplink S1
interface throughput, respectively, mobiles are rejected one by one due to Backhaul Saturation, starting from the mobile with
the lowest priority service, among all the cells of the site in order to reach a downlink or uplink effective RLC aggregate site
throughput the sites maximum downlink or uplink S1 interface throughput.
Resource Allocation for Maximum Throughput Demands:
For each cell, the remaining cell resources available are:
TX i ic
TX i ic
Downlink: R Rem DL = TL DL Max
M
TX i ic
TX i ic
Uplink: R Rem UL = TL UL Max
Sel
Mi
R Min DL
Sel
i
Sel
Mi
R Min UL
Sel
Mi
For each mobile, the remaining throughput demands are either the maximum UE capacities or the difference between the
maximum and the minimum throughput demands, whichever is smaller:
Sel
Sel
Sel
Mi
Mi
Mi
Max
Downlink: TPD Rem DL = Min TPD Max DL TP UE DL TPD Min DL
Sel
Sel
Sel
M
M
M
i
i
Max
i
Uplink: TPD Rem UL = Min TPDMax UL TP UE UL TPD Min UL
For their maximum throughput demands, LTE-A users are scheduled separately on each of their serving cells (primary and
secondary cells for carrier aggregation / coordinated servers for downlink non-coherent joint transmission CoMP). Each users
Sel
Mi
Sel
Mi
remaining throughput demand ( TPD Rem DL and TPD Rem UL ) is distributed over each of its serving cells proportionally to
the resources available on each serving cell and to the users downlink effective RLC channel throughput or uplink effective
RLC allocated bandwidth throughput on each of its serving cell:
Sel
Mi
Downlink: TPD Rem DL
TX i ic
Server n
Mi
R Rem DL
CTP E DL
Server n
Server n
= TPD Rem DL -------------------------------------------------------------------------------------------------N
Sel
Mi
TX ic
i
RRem DL Server n CTPE DL

i
Server n
n=1
Sel
Mi
Uplink: TPD Rem UL
TX i ic
Server n
Mi
R Rem UL
ABTP E UL
Server n
Server n
= TPD Rem UL -----------------------------------------------------------------------------------------------------N
Sel
Mi
TX ic
i
RRem UL
Server n
ABTP E UL
Server n
n=1
Here, N is the effective number of serving cells found for carrier aggregation or downlink non-coherent joint
transmission CoMP as described in "Best Server Determination" on page 537.
You can add an option in the Atoll.ini file to have each users remaining throughput demand distributed over each of its serving
cells proportionally only to the resources available on each serving cell:
Sel
Mi
Downlink: TPD Rem DL
TX i ic
Sel
Mi
Server n
R Rem DL
Server n
= TPD Rem DL ----------------------------------------------------N
TX i ic
RRem DL Server n
n=1
560

AT332_TRR_E0
Uplink:
Sel
i
TPD Rem UL
Server n
M
TX ic
i
Sel
i
TPD Rem UL
R Rem UL
Server n
----------------------------------------------------N
TX ic
i
RRem UL
Server n
n=1
Here, N is the effective number of serving cells found for carrier aggregation or downlink non-coherent joint
transmission CoMP as described in "Best Server Determination" on page 537.
For more information, see the Administrator Manual.
Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell activation threshold defined in the
DL
terminal reception equipment properties ( T SCell ) are activated for aggregation in downlink. Similarly, only secondary cells
whose PDSCH C/(I+N) and PUSCH C/(I+N) are both higher than or equal to the secondary cell activation threshold defined in
DL
UL
the terminal and cell reception equipment properties ( T SCell and T SCell ), respectively, are activated for aggregation in uplink.
Within each serving cell, resource allocation for the maximum throughput demands is carried out according to the scheduler
used by that particular cell. For the remaining throughput demands of the mobiles, the following resource allocation methods
are available:
Proportional Fair:
The goal of this scheduling method is to distribute resources among users fairly in such a way that, on the average,
each user gets the highest possible throughput that it can get under the radio conditions at its location.
Sel
Let the total number of users be N M i
.
TX i ic
TX i ic
a. Each users channel throughput is increased by the multi-user diversity gain G MUG DL or G MUG UL read from the
Sel
Mi
scheduler properties for the downlink or uplink bearer ( B DL

Sel
Mi
Sel
Mi
Sel
or B UL ), Mobility M i assigned to mobile
, and the number of connected users, DL or UL, in the cell TXi(ic) in the iteration k-1.
Sel
Mi
Sel
Mi
CTP P DL = CTP P DL
Sel
Mi
TX i ic
Without MUG
Sel
Mi
TX i ic
Sel
Mi
G MUG DL and CTP P UL = CTP P UL

Max
TX i ic
TX i ic
Without MUG
G MUG UL
Sel
Mi
Max
G MUG DL = 1 if CINR PDSCH CINR MUG and G MUG UL = 1 if CINR PUSCH PUCCH CINR MUG .
If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph,
it is interpolated from the gain values available for the numbers of users just less than and just greater than the
actual number of users.
b. Atoll divides the remaining resources in the cell into equal parts for each user:
TX i ic
TX i ic
R Rem DL
R Rem UL
--------------------- and -------------------N
N
c. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel
Mi
RD Rem DL
Sel
Mi
Sel
Sel
Mi
Mi
TPD Rem DL
TPD Rem UL
= --------------------------- and RD Rem UL = --------------------------Sel
Mi
Sel
Mi
CTP P DL
CTP P UL
Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the
peak channel throughputs at the users location.
d. The resources allocated to each user by the Proportional Fair scheduling method for satisfying its maximum
throughput demands are:
Sel
Sel
TX i ic
Sel
Sel
TX i ic
Mi
Mi
Mi
Mi
R Rem DL
R Rem UL
- and R Max
R Max DL = Min RD Rem DL ------------------- UL = Min RD Rem UL ---------------------
N
N
Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from
the remaining resources of the cell, whichever is smaller.
e. Atoll stops the resource allocation in downlink or uplink,
561

When/If in downlink
Sel
i
TX ic
i
R Max DL = R Rem DL , i.e., the resources available in downlink have been used up
Sel
i
for satisfying the maximum throughput demands of the mobiles.
When/If in uplink
Sel
i
TX ic
i
R Max UL = R Rem UL , i.e., the resources available in uplink have been used up for
Sel
Mi
satisfying the maximum throughput demands of the mobiles.

f. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of
remaining users.
g. Atoll recalculates the remaining resources as follows:
TX i ic
TX i ic
R Rem DL = TL DL Max
Sel
Mi
R Min DL
Sel
Mi
TX ic
i
TX ic
i
R Rem UL = TL UL Max
Sel
Mi
R Max DL and
Sel
Mi
M
Sel
i
R Min UL
Sel
Mi
Sel
i
R Max UL
Sel
Mi
h. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied
TX i ic
TX i ic
until either R Rem DL = 0 and R Rem UL = 0 , or all the maximum throughput demands are satisfied.
Round Robin:
The goal of this scheduling method is to allocate equal resources to users fairly.
Sel
Let the total number of users be N M i
a. Atoll divides the remaining resources in the cell into equal parts for each user:
TX ic
i
TX ic
i
R Rem DL
R Rem UL
--------------------- and -------------------N
N
b. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel
Mi
Sel
Mi
Sel
Mi
Sel
Mi
TPD Rem DL
TPD Rem UL
RD Rem DL = --------------------------- and RD Rem UL = --------------------------Sel
Mi
Sel
Mi
CTP P DL
CTP P UL
c. The resources allocated to each user by the Round Robin scheduling method for satisfying its maximum
Sel
Mi
R Max DL
TX i ic
Sel
Sel
Sel
TX i ic
Mi
R Rem DL
R Rem UL
Mi
Mi
= Min RD Rem DL --------------------- and R Max UL = Min RD Rem UL ---------------------
N
N
d. Atoll stops the resource allocation in downlink or uplink,
When/If in downlink
Sel
i
TX ic
i
Sel
Mi
562

AT332_TRR_E0
When/If in uplink
Sel
i
TX ic
i
Sel
i

e. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of
remaining users.
f. Atoll recalculates the remaining resources as follows:
TX i ic
TX i ic
Sel
Mi
R Min DL
Sel
Mi
TX i ic
TX i ic
Sel
Mi
Sel
Mi
Sel
Mi
R Max DL and
R Min UL
Sel
Mi
Sel
Mi
R Max UL
Sel
Mi
g. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied
TX ic
i
TX ic
i
Proportional Demand:
The goal of this scheduling method is to allocate resources to users weighted according to their remaining throughput
demands. Therefore, the user throughputs for users with high throughput demands will be higher than those with low
throughput demands. In other words, this scheduler distributes channel throughput between users proportionally to
their demands.
a. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel
Mi
Sel
Mi
RD Rem DL
Sel
Sel
Mi
Mi
TPD Rem DL
TPD Rem UL
= --------------------------- and RD Rem UL = --------------------------Sel
Mi
Sel
Mi
CTP P DL
CTP P UL
b. Atoll calculates the amount of effective remaining resources of the cell to distribute among the users as follows:
Sel
Sel
TX i ic
Mi
TX i ic
Mi
TXi ic
TXi ic
R Eff Rem DL = Min R Rem DL

RD Rem DL and R Eff Rem UL = Min R Rem UL
RD Rem UL
Sel
Sel
Mi
Mi
c. The resources allocated to each user by the Proportional Demand scheduling method for satisfying its maximum
Sel
Mi
R Max DL
Sel
Mi
TX i ic
Mi
RDRem DL
Sel
Mi
Sel
Mi
TX i ic
RD Rem DL
RD Rem UL
- and R Max
= R Eff Rem DL --------------------------------- UL = R Eff Rem UL ---------------------------------Sel
Sel
Sel
i
Mi
RDRem UL
Sel
i
Max C/I:
The goal of this scheduling method is to achieve the maximum aggregate throughput for the cells. This is done by
allocating as much resources as needed to mobiles with high C/(I+N) conditions. As mobiles with high C/(I+N) can get
higher bearers, and therefore require less amount of resources, more mobiles can therefore be allocated resources
in the same frame, and the end-throughput for each cell will be the highest compared to other types of schedulers.
Sel
a. Atoll sorts the M i
TX i ic
N Users in order of decreasing PDSCH, or PUSCH and PUCCH C/(I+N), depending on whether
the allocation is being performed for the downlink or for the uplink.
b. Starting with the mobile with the highest rank, Atoll allocates the downlink and uplink resources required to
satisfy each users remaining throughput demands in downlink and uplink as follows:
563

Sel
i
R Max DL
M
Sel
i
Sel
Sel
i
M
TPD Rem DL
TPD Rem UL
i
= --------------------------- and R Max UL = --------------------------M
Sel
i
CTP P DL
Sel
i
CTP P UL
c. Atoll stops the resource allocation in downlink or uplink,
When/If in downlink
Sel
Mi
TX i ic
Sel
Mi
When/If in uplink
Sel
Mi
TX i ic
Sel
Mi

Spatial Multiplexing with Multi-User MIMO:
MU-MIMO lets the system/scheduler work with parallel LTE frames. Therefore, many users can be co-scheduled on the same
resources. MU-MIMO can be used if the cell supports MU-MIMO, if the calculated value for the MU-MIMO criterion is higher
TX i ic
TX i ic
than the MU-MIMO threshold T MU MIMO DL or T MU MIMO UL , and the number of antenna ports is equal to 2 or more.
MU MIMO
In both downlink and uplink, each mobile M i

resource consumptions of a mobile
MU MIMO
Mi
RC DL
MU MIMO
Mi
MU MIMO
i
TL DL
----------------------------------TX ic
i
G MU MIMO DL
M
Saturation occurs when
has a corresponding traffic load TL
and RC UL
MU MIMO UL
Mi
. However, the actual
are given by:

MU MIMO
Mi
RC UL
MU MIMO
Mi
MU MIMO
i
TL UL
-----------------------------------TX ic
i
G MU MIMO UL
M
TX i ic
= TL UL Max or
MU MIMO DL
Mi
RC DL
TX i ic
= TL DL Max .
Backhaul Capacity Limitation:

Backhaul overflow ratios are calculated for each site as follows:
Sel
Sel
Mi
Mi
R Max DL CTP E DL
Sel
M i Site
= Max 1 ------------------------------------------------------------------------------------------------------ and
Sel
Sel
Mi
Mi
Site
R Min DL CTP E DL
TP S1 DL
Sel
M i Site
Site
BHOF DL
Sel
Sel
Mi
Mi
R Max UL CTP E UL
Sel
M i Site
= Max 1 ------------------------------------------------------------------------------------------------------
Sel
Sel
Mi
Mi
Site
R Min UL CTP E UL
TP S1 UL
Sel
M i Site
Site
BHOF UL
Total Amount of Resources Assigned to Each Selected Mobile:

Sel
Atoll calculates the amounts of downlink and uplink resources allocated to each individual mobile M i
referred to as the traffic loads of the mobiles) as follows:
Sel
Sel
Mi
Downlink: TL DL
Sel
Mi
= R DL
Sel
Mi
Sel
Sel
Mi
CTP P DL
564
Sel
Mi
Mi
Mi
Max DL CTP P DL

BHOF DL
= ------------------------------------------------------------------------------------------------------------------
(which can also be

AT332_TRR_E0
Sel
Uplink:
Sel
i
TL UL
M
Sel
i
R UL
M
Sel
i
Sel
i
M
Mi
i
Max UL CTP P UL

BHOF UL
= -----------------------------------------------------------------------------------------------------------------Sel
Sel
i
CTP P UL
Output
Sel
Mi
TL DL
Sel
Mi
TL UL
Sel
Mi
= R DL
Sel
: Downlink traffic load or the amount of downlink resources allocated to the mobile M i .
Sel
Mi
Sel
= R UL : Uplink traffic load or the amount of uplink resources allocated to the mobile M i .
6.4.7.2 User Throughput Calculation

User throughputs are calculated for the percentage of resources allocated to each mobile selected by the scheduling for RRM
Sel
during the Monte Carlo simulations, M i .

any LTE-A mobile, limited by the maximum throughput supported by the UE category.
Input
Sel
Mi
R DL
Sel
: Amount of downlink resources allocated to the mobile M i
as calculated in "Scheduling and Radio Resource
Sel
Mi
Sel
R UL : Amount of uplink resources allocated to the mobile M i
Sel
Mi
Sel
CTP P DL : Downlink peak RLC channel throughput at the mobile M i
as calculated in "Throughput Calculation" on
page 541.
Sel
Mi
Sel
CTP P UL : Uplink peak RLC channel throughput at the mobile M i
page 541.
Sel
TX i ic
Mi
BLER B DL : Downlink block error rate read from the BLER vs. CINR PDSCH graph available in the reception
Sel
equipment assigned to the terminal used by the mobile M i
Sel
Mi
Mi
BLER B UL : Uplink block error rate read from the BLER vs. CINR PUSCH PUCCH graph available in the reception

Sel
Mi
Sel
f TP Scaling : Throughput scaling factor defined in the properties of the service used by the mobile M i .
TP Offset : Throughput offset defined in the properties of the service used by the mobile M i
Sel
Mi
Sel
Calculations
Downlink:
Sel
Mi
Sel
Mi
Sel
Mi
CTP P DL
Peak RLC User Throughput: UTP P DL = R DL
Mi
Mi
Mi
Effective RLC User Throughput: UTP E DL = UTP P DL 1 BLER B DL
Sel
Sel
Sel
565

Application User Throughput:
Sel
i
UTP A DL
M
Sel
i
UTP E DL
M
Sel
i
Sel
M
f TP Scaling
i
------------------------- TP Offset
100
Uplink:
Sel
Mi
Sel
Mi
Sel
Mi
CTP P UL
Peak RLC User Throughput: UTP P UL = R UL
M
M
Mi
i
i
Effective RLC User Throughput: UTP E UL = UTP P UL 1 BLER B UL
Sel
Sel
Mi
Application User Throughput: UTP A UL
Sel
Sel
Mi
Sel
Sel
Mi
Sel
Mi
f TP Scaling
- TP Offset
= UTP E UL -----------------------100
Output
M
Sel
i
Sel
UTP P DL : Downlink peak RLC user throughput at the pixel, subscriber, or mobile M i
UTP E DL : Downlink effective RLC user throughput at the pixel, subscriber, or mobile M i .
UTP A DL : Downlink application user throughput at the pixel, subscriber, or mobile M i .
UTP P UL : Uplink peak RLC user throughput at the pixel, subscriber, or mobile M i
UTP E UL : Uplink effective RLC user throughput at the pixel, subscriber, or mobile M i
Sel
Mi
UTP A UL
Sel
Mi
Sel
Sel
Mi
Sel
Sel
i
Sel
Sel
Mi
.
Sel
Sel
: Uplink application user throughput at the pixel, subscriber, or mobile M i
6.5 Automatic Planning Algorithms

The following sections describe the algorithms for:
"Automatic Neighbour Planning" on page 566.

"Automatic Inter-technology Neighbour Planning" on page 570.
"Automatic Frequency Planning Using the AFP" on page 573.
"Automatic Physical Cell ID Planning Using the AFP" on page 575.
"Automatic PRACH RSI Planning Using the AFP" on page 580.
6.5.1 Automatic Neighbour Planning

The intra-technology neighbour planning algorithm takes into account the cells of all the TBC transmitters. It means that the
cells of all the TBC transmitters of your ATL document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfil the following conditions:
They are active,

They belong to the folder on which allocation has been executed. This can be the Transmitters folder or a group of
transmitters (subfolder).
Only TBA cells are assigned neighbours.

zone.
566

AT332_TRR_E0
We assume a reference cell TXi(ic) and a candidate neighbour cell TXj(jc). When automatic allocation starts, Atoll checks the
1. The distance between both cells must be less than the user-definable maximum inter-site distance. If the distance
discarded.
When allocation is based on distance only, Atoll calculates the effective distance between the reference cell and its
candidate neighbour from the real distance between them and the azimuths of their antennas:
Dist CellA CellB = D 1 + x cos x cos
Where x = 0.3% so that the maximum variation in D does not to exceed 1%. D is stated in m.
Figure 6.3: Inter-Transmitter Distance Calculation

The formula above implies that two cells facing each other have a smaller effective distance than the actual distance.
Candidate neighbours are ranked in the order of increasing effective distance from the reference cell.
[Neighbours]
The real distance is considered when coverage conditions are used in the allocation.
2. The calculation options,
Force Co-site Cells as Neighbours: If selected, Atoll adds all the cells located on the same site as the reference cell
to the candidate neighbour list. The weight of this constraint can be defined. It is used to calculate the rank of each
neighbour, and its importance.
Force Adjacent Cells as Neighbours: If selected, Atoll adds all the cells geographically adjacent to the reference
cell to the candidate neighbour list. The weight of this constraint can be defined. It is used to calculate the rank of
each neighbour, and its importance.
Determination of Adjacent Cells: Geographically adjacent cells are determined on the basis of their best server
coverage areas. A candidate neighbour cell TXi(ic) is considered adjacent to the reference cell TXi(ic) if there exists
at least one pixel of TXj(jc)s best server coverage area where TXi(ic) is the second best server. The ranking of
adjacent neighbour cells increases with the number of such pixels. Adjacent cells are sorted in the order of
decreasing rank.
Figure 6.4: Determination of Adjacent Cells
Force Adjacent Layers as Neighbours: If selected, Atoll adds all the cells adjacent across network layers to the
reference cell to the candidate neighbour list. The weight of this constraint is always the average of the Min and
567

Max values defined for the adjacency factor. This weight is used to calculate the rank of each neighbour and its
importance. Cells are considered adjacent across layers if they belong to different layers and have a coverage
overlap of at least one pixel.
Force Symmetry: If selected, Atoll adds the reference cell to the candidate neighbour list of the its candidate
neighbour.
A symmetric neighbour relation is allowed only if the neighbour list of the reference cell is not already full. If TXj(jc)
is a neighbour of TXi(ic) but TXi(ic) is not a neighbour of TXj(jc), there can be two possibilities:
i.
The neighbour list of TXj(jc) is not full, Atoll will add TXi(ic) to the end of the list.
ii. The neighbour list of TXj(jc) is full, Atoll will not be able to add TXi(ic) to the list, so it will also remove TXj(jc)
from the neighbour list of TXi(ic).
[Neighbours]
Force Exceptional Pairs: This option enables you to force/forbid some neighbour relations. Exceptional pairs are
pairs of cells which will always or never be neighbours of each other.
If you select "Force exceptional pairs" and "Force symmetry", Atoll considers the constraints between exceptional
pairs in both directions so as to respect symmetry condition. On the other hand, if neighbourhood relationship is
forced in one direction and forbidden in the other, symmetry cannot be respected. In this case, Atoll displays a
warning in the Event viewer.
Delete Existing Neighbours: If selected, Atoll deletes all the current neighbours and carries out a new neighbour
allocation. If not selected, the existing neighbours are kept in the list.
3. If the Use Coverage Conditions check box is selected, the coverage areas of TXi(ic) and TXj(jc) must have an overlap.
Otherwise, only the distance criterion is taken into account.
The overlapping zone ( S TX ic S TX jc ) is defined as follows
i
Here S TX ic is the surface area covered by the cell TXi(ic) that comprises all the pixels where:
i
The distance to the cell TXi(ic) is less than or equal to the round-trip time distance corresponding to the cells
PRACH preamble format.
The received RSRP is greater than or equal to the cells Min RSRP: E DLRS T RSRP .
The received RSRP is within E DLRS + O Individual + M HO
TX i ic
TX ic
i
TX i ic
M HO
TX i ic
M HO
TX ic
i
TX ic
i
TX i ic
TX ic
i
TX ic
i
TX ic
i
and E DLRS + O Individual + M HO
+ M End .
is the handover margin defined for the cell TXi(ic). When a global handover start value is used,
= M Start . M Start and M End are global handover start and handover end values.
S TX jc is the surface area covered by the cell TXj(jc) that comprises all the pixels where:
j
The distance to the cell TXj(jc) is less than or equal to the round-trip time distance corresponding to the cells
PRACH preamble format.
The received RSRP is greater than or equal to the cells Min RSRP: E DLRS T RSRP .
The received RSRP with offset ( E DLRS + O Individual ) is the highest.
TX j jc
TX jc
j
568
TX j jc
TX jc
j
TX i ic
If a global value of the minimum RSRP threshold ( T RSRP ) is set in the coverage
conditions dialog box, for each cell, Atoll uses the higher of the two values, i.e., global
value and the value defined for that cell.
body loss, the terminal that has the highest difference between gain and losses, and
the shadowing margin calculated using the defined cell edge coverage probability, if
the option is selected. The service and terminal are selected such that the selection
gives the largest possible coverage areas for the cells.

AT332_TRR_E0
S TX ic S TX jc
i
j
- 100 ),
When the above conditions are met, Atoll calculates the percentage of the coverage area overlap ( -------------------------------------S TX ic
i
and compares this value with the % Min Covered Area. TXj(jc) is considered a neighbour of TXi(ic) if
S TX ic S TX jc
i
j
-------------------------------------- 100 % Min Coverage Area .
S TX ic
i

Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by decreasing
importance in order to keep the ones with high importance. If the maximum number of neighbours to be allocated to each
cell is exceeded, Atoll keeps the ones with high importance.
Neighbourhood cause
When
Importance value
Existing neighbour
Only if the Delete Existing Neighbours option is not selected

Existing importance
Exceptional pair
Only if the Force Exceptional Pairs option is selected
100 %
Co-site cell
Only if the Force Co-site Cells as Neighbours option is

selected
Adjacent layer
Only if the Force Adjacent Layers as Neighbours option is

selected
Adjacent cell
Only if the Force Adjacent Cells as Neighbours option is

selected

Only if the % Min Covered Area is exceeded
relationship
Only if the Force Neighbour Symmetry option is selected
The importance is evaluated using an Importance Function (IF), which takes into account the following factors:
transmitter.
d Di = 1 ---------d max
of antennas.
569

[Neighbours]

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
30%
Min(A)
30%
Max(A)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

Neighbourhood cause
Importance Function

Coverage
10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
Adjacent layer
(Min(A)+Max(A))/2
45%
Adjacent cells
30%+30%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Co-site cells
60%+40%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Where:
co-site neighbours, adjacent layers, adjacent cells, and neighbours allocated based on
coverage overlapping.
causes.
In the results, Atoll lists only the cells for which it finds new neighbours. Cells whose channels have the same centre frequency
are listed as intra-carrier neighbours. Otherwise, neighbour cells are listed as inter-carrier neighbours.
6.5.2 Automatic Inter-technology Neighbour Planning

The inter-technology neighbour planning algorithm takes into account all the TBC transmitters (if the other technology is
GSM) or the cells of all the TBC transmitters (for any other technology than GSM). This means that all the TBC transmitters
(GSM) or the cells of all the TBC transmitters (all other technologies) of the linked document are potential neighbours.
The cells to be allocated in the main document will be called TBA cells. They must fulfil the following conditions:
570
They are active,


AT332_TRR_E0


zone.
We assume a reference cell A and a candidate neighbour B. When automatic allocation starts, Atoll checks following
conditions:
1. The distance between reference cell and the candidate neighbour must be less than the user-definable maximum
inter-site distance. If the distance is greater than this value, the candidate neighbour is discarded.

[Neighbours]
2. The calculation options:
CDMA carriers: This option is available when an LTE network is being co-planned with a UMTS, CDMA, or TDSCDMA network. This option enables you to select the CDMA carrier(s) that you want Atoll to consider as potential
neighbours of LTE cells. You may choose one or more carriers. Atoll will allocate only the cells using the selected
carriers as neighbours.
Force co-site cells as neighbours: If selected, Atoll adds all the transmitters/cells located on the same site as the
reference cell in its candidate neighbour list. The weight of this constraint can be defined. It is used to calculate
the rank of each neighbour and its importance.
Force exceptional pairs: This option enables you to force/forbid some neighbour relations. Exceptional pairs are
Delete existing neighbours: If selected, Atoll deletes all the current neighbours and carries out a new neighbour
3. Neighbour relation criterion:
Allocation based on distance:

The allocation algorithm is based on the effective distance between the reference cell and its candidate
neighbour.
Algorithm based on coverage overlapping:

The coverage areas of the reference cell A and the candidate neighbour B must overlap ( S A S B ).
571

Two cases may exist for SA:
1st case: SA is the area where the cell A is the best serving cell, with a 0 dB margin.
This means that the reference signal energy per resource element received from A is greater than the
minimum required (Min RSRP), and is the highest one. .
2nd case: The margin is other than 0 dB. SA is the area where:
The reference signal energy per resource element received from A exceeds the minimum required (Min RSRP)
and is within a margin from the highest signal level.
Two cases may exist for SB:
1st case: SB is the area where the candidate neighbour is the best server. In this case, the margin must be set
to 0dB.
The signal level received from B exceeds the minimum required, and is the highest one.
2nd case: The margin is other than 0dB. SB is the area where:
The signal level received from B exceeds the minimum required and is within a margin from the best signal
level.
SA SB
Atoll calculates the percentage of the coverage area overlap ( ------------------ 100 ) and compares this value with the %
SA
SA SB
Min Covered Area. B is considered a neighbour of A if ------------------ 100 % Min Covered Area .
SA
Candidate neighbours are ranked in the order of decreasing coverage area overlap percentages.
The importance (%) of neighbours depends on the distance and on the reason of allocation:
For allocation based on distance:

Neighbour cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
Co-site transmitter/cell
100 %
Neighbour relation that fulfils

distance conditions
d1 ---------d max
d is the effective distance between the reference cell and the neighbour and d max is the maximum inter-site
distance.
For allocation based on coverage overlapping:

Neighbour cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
IF

IF
transmitter.
d Di = 1 ---------d max
d is the real distance between the reference cell and the neighbour and d max is the maximum inter-site distance.
572

AT332_TRR_E0

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

cause
IF

No
10%+50%{10%(Di)+90%(O)}
Yes
60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.
In the results, Atoll displays only the cells for which it finds new neighbours.
6.5.3 Automatic Frequency Planning Using the AFP

The role of an Automatic Frequency Planning (AFP) tool is to assign frequencies (channels) to cells of a network such that the
overall network performance is optimised. In other words, the interference within the network is reduced as much as possible.
Co-channel interference is the main reason for overall network quality degradation in LTE. In order to improve network
performance, the LTE AFP tries to minimise co- and adjacent channel interference as much as possible while respecting any
constraints input to it. The main constraints are the resources available for allocation, i.e., the number of frequencies with
which the AFP can work, and the relationships to take into account, i.e., interference matrices, neighbours, and distance
between transmitters.
The AFP is based on a cost function which represents the interference level in the network. The aim of the AFP is to minimise
the cost. The best, or optimum, frequency plan is the one which corresponds to the lowest cost.
The following describes the AFPs automatic planning method for carriers in LTE networks, which takes into account
interference matrices, neighbour relations, and distance between transmitters.
The AFP takes into account the cells of all the TBC transmitters. The cells to be allocated will be called TBA cells. They must
fulfil the following conditions:
They are active,

Their channel allocation status is not set to locked,
They are located inside the focus zone.
If no focus zone exists in the ATL document, Atoll takes into account all the cells whose
path loss matrix calculation radius intersects the rectangle containing the computation
zone.
6.5.3.1 Constraint and Relationship Weights

The AFP is based on a cost function which takes into account channel separation constraints based on the channel overlap
ratio as calculated in "Co- and Adjacent Channel Overlaps Calculation" on page 493. Channel separation is studied between
573

each TBA cell and its related cells. Atoll calculates the cost between each individual TBA and related cell, and then the overall
cost for the TBA cell.
Related cells of a TBA cell are:
Its neighbours, if the check box "Existing neighbours" is selected,

Default weight Neighbour = 0,5
Cells that are listed in the interference matrix of the TBA cell,
Default weight IM = 0,3
Cells within the cells (or the default) minimum reuse distance, if the check box "Reuse distance" is selected,
Default weight Dis tan ce = 0,2
The sum of the weights assigned to the above relations is 1.
You can modify these weights in your LTE document. The absolute values of the constraint weights are calculated from the
relative weights (%) defined in the Constraint Weights dialog box as follows:
% Neighbour
Neighbour = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce
% IM
IM = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce
% Dis tan ce
Dis tan ce = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce
6.5.3.2 Cost Calculation

The cost of the relation between the TBA cell and its related cell is calculated as follows:
$
TX i ic TX j jc
TX i ic TX j jc
= rO
TX i ic TX j jc
Where r O
TX i ic TX j jc
Neighbour Neighbour
TX i ic TX j jc
+ Dis tan ce Dis tan ce
TX ic TX j jc
+ i
IM
IM
is the channel overlap ratio as calculated in "Co- and Adjacent Channel Overlaps Calculation" on
page 493.
TX i ic TX j jc
TX i ic TX j jc
Neighbour
is the importance of the relationship between the TBA cell and its related neighbour cell. Neighbour
is
calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 566. For
manual neighbour planning, this value is equal to 1.
TX i ic TX j jc
IM
is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows:
TX i ic TX j jc
TX i ic TX j jc
IM
= r CCO
TX i ic TX j jc
TX i ic TX j jc
IM CC
and IM CC
TX i ic TX j jc
IM CC
TX i ic TX j jc
+ r ACO
TX i ic TX j jc
IM AC
are respectively the co- and adjacent channel interference probabilities calculated as
TX i ic TX j jc
explained in "Interference Matrix Calculation" on page 582. r CCO
TX i ic TX j jc
and r ACO
are the co- and adjacent channel
overlap ratios as calculated in "Co- and Adjacent Channel Overlaps Calculation" on page 493.
TX i ic TX j jc
Dis tan ce
is the importance of the relationship between the TBA and its related cell with respect to the distance between
TX i ic TX j jc
them. Dis tan ce
is calculated as explained in "Distance Importance Calculation" on page 582.
Atoll calculates the quality reduction factor for the TBA cell and its related cell from the cost calculated above as follows:
QRF
TX i ic TX j jc
= 1$
TX i ic TX j jc
The quality reduction factor is a measure of the cost of an individual relation.
574

AT332_TRR_E0
The total cost of the current frequency plan for any TBA cell is given as follows, considering all the cells with which the TBA
cell has relations:
TX ic
i
$ Total = 1
QRF
TX ic TX jc
i
j
TX jc
j
And, the total cost of the current frequency plan for the entire network is simply the sum of the total TBA cell costs calculated
above, i.e.,
$ Total =
TX i ic
$ Total
TX i ic
6.5.3.3 AFP Algorithm

The AFP algorithm is an iterative algorithm which:
Calculates the cost (as describe d above) of the initial frequency plan,
Tries different frequency plans in order to reduce the cost,
Memorises the different plans in order to determine the best one, i.e., the plan that provides the lowest total cost,
Stops when it is unable to improve the cost of the network, and proposes the last known best frequency plan as the
solution.
6.5.4 Automatic Physical Cell ID Planning Using the AFP

In LTE, 504 physical cell IDs are available, numbered from 0 to 503. There are as many pseudo-random sequences defined in
the 3GPP specifications. Physical cell IDs are grouped into 168 unique cell ID groups (called SSS IDs in Atoll), with each group
containing 3 unique identities (called PSS IDs in Atoll). An SSS ID is thus uniquely defined by a number in the range of 0 to 167,
and a PSS ID is defined by a number in the range of 0 to 2.
Each cells downlink reference signals transmit a pseudo-random sequence corresponding to the physical cell ID of the cell.
The SSS and PSS are transmitted over the centre six frequency blocks independent of the channel bandwidths used by cells.
Mobiles synchronise there transmission and reception frequency and time by listening first to the PSS. Once they know the
PSS ID of the cell, they listen to the SSS of the cell in order to know the SSS ID. The combination of these two IDs gives the
physical cell ID and the associated pseudo-random sequence that is transmitted over the downlink reference signals.
Once the physical cell ID and the associated pseudo-random sequence is known to the mobile, the cell is recognized by the
mobile based on the received downlink reference signals. Downlink channel quality measurements are also made on the
downlink reference signals.
As can be understood from the above description, if all the cells in the network transmit the same physical cell ID, it will be
impossible for a mobile to identify different cells. Cell search and selection will be impossible. Therefore, it is important to
intelligently allocate physical cell IDs to cells so as to allow easy recognition of cells by mobiles.
The following describes the AFPs automatic planning method for physical cell IDs in an LTE network, which takes into account
interference matrices, neighbour relations (first-order neighbours, first-order neighbours of a common LTE cell, first-order
neighbours of a common GSM or UMTS cell in 3GPP multi-RAT documents and CDMA cell in 3GPP2 multi-RAT documents, and
optionally second-order neighbours), distance between transmitters, and the frequency plan of the network.
They are active,

Their PSS ID and SSS ID statuses are both not set to locked,
zone.

The AFP is based on a cost-based function which takes into account the following constraints, in the order of priority:
1. Physical cell ID,
Assigned weight ID = 0,75
2. PSS ID,
575

Assigned weight PSS = 0,02

3. Strategy for co-site cells,
Assigned weight Strategy = 0,23
4. PCI Mod 6, for single-antenna port DL CRS),
Assigned weight CRS = 0
5. PCI Mod 30, for UL DMRS sequence groups,
Assigned weight ULDMRS = 0
6. PCI Mod (number of frequency blocks / 2), for PCFICH resource element groups,
Assigned weight PCFICH = 0
The sum of the weights assigned to the above constraints is 1.
% ID
ID = ------------------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % Strategy + % CRS + % ULDMRS + % PCFICH
% PSS
PSS = ------------------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % Strategy + % CRS + % ULDMRS + % PCFICH
% SSS
Strategy = ------------------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % Strategy + % CRS + % ULDMRS + % PCFICH
% CRS
CRS = ------------------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % Strategy + % CRS + % ULDMRS + % PCFICH
% ULDMRS
ULDMRS = ------------------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % Strategy + % CRS + % ULDMRS + % PCFICH
% PCFICH
PCFICH = ------------------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % Strategy + % CRS + % ULDMRS + % PCFICH
The above constraints are studied between each TBA cell and its related cells. Atoll calculates the cost between each
individual TBA and related cell, and then the overall cost for the TBA cell.

Assigned weight Neighbour = 0,35
TBA cells which are first-order neighbours of a common cell are also related to each other through that cell. This
relation is also taken into account,
Assigned weight Inter Neighbour = 0,15
You can choose to not take into account the physical cell ID collision between neighbours of a common cell by adding
an option in the Atoll.ini file (see the Administrator Manual). If the collision between neighbours of a common cell is
not taken into account, the weight assigned to the direct first-order neighbour relation alone is Neighbour = 0,5 and
that of the collision between neighbours of a common cell is of course Inter Neighbour = 0 .
By adding an option in the Atoll.ini file (see the Administrator Manual), second-order neighbours can also be taken
into account. In this case, the assigned weights are: Neighbour = 0,25 , 2nd Neighbour = 0,10 , and
Inter Neighbour = 0,15 .
Inter Neighbour applies to the relation between neighbours of a common cell, which can be an LTE cell, a UMTS cell
or a GSM transmitter in 3GPP multi-RAT documents or an LTE or CDMA cell in 3GPP2 multi-RAT documents.
576

AT332_TRR_E0
Figure 6.7 on page 577 depicts the different neighbour relations that may exist in LTE.
Figure 6.7: Neighbour Relations for Physical Cell ID Allocation
Assigned weight IM = 0,3
Cells within the cells (or the default) reuse distance, if the check box "Reuse distance" is selected,
Assigned weight Dis tan ce = 0,2
% Neighbour
Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% Inter Neighbour
Inter Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% 2nd Neighbour
2nd Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% IM
IM = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% Dis tan ce
Dis tan ce = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce

Atoll calculates the constraint violation levels between the TBA cell TXi(ic) and its related cell TXj(jc) as follows:
For Strategy for co-site cells = None or Same SSS ID

TX i ic TX j jc
VL 1
TX i ic TX j jc
VL 2
ID
SSS
CRS
ULDMRS
= ID p Coll + Strategy p Penalty + CRS p Coll + ULDMRS p Coll
PCFICH
+ PCFICH p Coll
PSS
= PSS p Coll
For Strategy for co-site cells = Fixed PCI step

TX i ic TX j jc
VL 1
TX ic TX jc
i
j
VL 2
ID
IDStep
= ID p Coll + Strategy p Penalty

= 0
577

Where ID , PSS , and Strategy are the weights assigned to the physical cell ID, PSS ID, and co-site strategy constraints.
1
ID
ID
p Coll is the physical cell ID collision probability given by p Coll =
1
PSS
PSS
p Coll is the PSS ID collision probability given by p Coll =
SSS
SSS
p Penalty is the SSS ID penalty given by p Penalty
=
1
TX ic
i
if ID
TX ic
i
ID
if
TX i ic
if ID PSS
TX jc
j
ID
TX j jc
= ID PSS
if ID PSS
TX i ic
TX jc
j
= ID
TX j jc
ID PSS
TX i ic
TX i ic
if R Co-site 3 AND ID SSS

TX ic
i
TX ic
i
if R Co-site 3 AND
ID SSS
TX j jc
ID SSS
TX jc
j
ID SSS
TX ic
i
R Co-site if the SSS
--------------3
Otherwise
SSS
ID planning strategy is set to "Same per site", and by p Penalty = 0 if the allocation strategy for co-site cells is set to "None".
The SSS penalty models the SSS ID allocation constraint.
IDStep
p Penalty
is
the
penalty
of
not
respecting
the
user-defined
TX ic
TX jc
i
Mod Step ID j Mod Step OR

= 1 If ID
0
Otherwise
IDStep
p Penalty
TX i ic
fixed
TX j jc
ID
ID
----------------- ----------------Step
Step
PCI
step,
given
by
TX ic
i
R Co-site .
R Co-site is the number of co-site relationships at the site.

CRS
p Coll
is
CRS
p Coll
1
=
ULDRMS
p Coll
PCFICH
p Coll
PCFICH
p Coll
the
single
antenna
TX i ic
if ID
TX i ic
if ID
downlink
TX j jc
Mod6 = ID
TX j jc
Mod6 ID
cell-specific
Mod6
reference
signal
collision
probability
given
by
Mod6
is the UL DMRS collision probability given by
ULDMRS
p Coll
1
=
TX i ic
if ID
TX i ic
if ID
TX j jc
Mod30 = ID
TX j jc
Mod30 ID
Mod30
Mod30
is the collision probability of the physical control format indicator channel resource element groups given by
TX i ic
if ID
TX i ic
if ID
TX i ic
TX j jc
TX j jc
N FB
N FB
-
Mod ---------------Mod --------------- = ID
2
2
TX ic
i
N FB
TX j jc
Mod ---------------- ID
2
TX jc
j
N FB
Mod -----------------
2
Next, Atoll calculates the importance of the neighbour relations between the TBA cell and its related cell.
TX i ic TX j jc
Neighbours
TX i ic TX j jc
= Neighbour Neighbour
TX i ic TX j jc
Where Neighbour
+ Inter Neighbour Inter Neighbour + 2nd Neighbour 2nd Neighbour

TX i ic TX j jc
is calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 566. For
Inter Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour
planning. If two cells are neighbours of a common cell and have the same physical cell ID assigned, the importance of the
physical cell ID collision is the average of their neighbour importance values with the common neighbour cell. If more than
578

AT332_TRR_E0
one pair of neighbours of the TBA cell has the same physical cell ID assigned, then the importance is the highest value among
all the averages:
TX ic TX j1c
i
j1
Inter Neighbour
TX ic TX j2c
i
j2
Neighbour
+ Neighbour
Max
=
---------------------------------------------------------------------------------
2
All Neighbour Pairs

with ID Collisions
Where TX j1 j1c and TX j2 j2c are two neighbours of the TBA cell TX i ic that have the same physical cell ID assigned. The
above applies to intra-technology as well as inter-technology neighbours in 3GPP multi-RAT and 3GPP2 multi-RAT documents.
2nd Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour
planning. If the TBA cell has the same physical cell ID assigned as one of its second-order neighbours, the importance of the
physical cell ID collision is the multiple of the importance values of the first order neighbour relations between the TBA cell
and its second order neighbour. If the TBA cell is related to its second order neighbour through more than one first order
neighbour, the importance is the highest value among all the multiples:
2nd Neighbour =
TX ic TX jc
j
i
Neighbour
All Neighbour Pairs
Max
TX j jc TX k kc
Neighbour
with ID Collisions
Where TX k kc is the second-order neighbour of TX i ic through TX j jc .

Next, Atoll calculates the importance of the interference relations between the TBA cell and its related cell.
TX ic TX jc
i
j
Interference
TX ic TX jc
i
j
IM
TX i ic TX j jc
IM
TX ic TX jc
i
j
= IM IM
TX ic TX jc
i
j
TX ic TX jc
i
j
f Overlap
TX i ic TX j jc
= r CCO
TX i ic TX j jc
and IM
TX i ic TX j jc
TX i ic TX j jc
IM CC
TX i ic TX j jc
= IM CC
TX i ic TX j jc
IM CC
and IM CC
TX i ic TX j jc
+ r ACO
TX i ic TX j jc
IM AC
if the frequency plan is taken into account
otherwise.
TX ic TX jc
i
j
explained in "Interference Matrix Calculation" on page 582. r O
TX ic TX jc
i
j
, r CCO
TX ic TX jc
i
j
, and r ACO
are the total,
co-channel, and adjacent channel overlap ratios as calculated in "Co- and Adjacent Channel Overlaps Calculation" on
page 493.
TX i ic TX j jc
Dis tan ce
them.
TX i ic TX j jc
Dis tan ce
TX i ic TX j jc
f Overlap
is
TX i ic TX j jc
= rO
calculated
as
explained
in
"Distance
Importance
TX i ic TX j jc
if the frequency plan is taken into account and f Overlap
Calculation"
on
page 582.
= 1 otherwise.
From the constraint violation levels and the importance values of the relations between the TBA and its related cell, Atoll
calculates the quality reduction factor for the pair as follows:
QRF
TX i ic TX j jc
TX i ic TX j jc
TX i ic TX j jc
TX ic TX j jc
TX i ic TX j jc
TX i ic TX j jc
TX i ic TX j jc
i
= 1 VL1
+ VL 2
+ VL 1
Neighbours f Overlap
Interference

The total cost of the current physical cell ID plan for any TBA cell is given as follows, considering all the cells with which the
TBA cell has relations:
TX ic
i
$ Total = 1
QRF
TX ic TX jc
i
j
TX j jc
And, the total cost of the current physical cell ID plan for the entire network is simply the sum of the total TBA cell costs
calculated above, i.e.,
$ Total =
TX i ic
$ Total
TX i ic
579


Calculates the cost (as described above) of the current physical cell ID plan,
Tries different physical cell IDs to cells in order to reduce the costs,
Stops when it is unable to improve the cost of the network, and proposes the last known best physical cell ID plan as
the solution.
6.5.5 Automatic PRACH RSI Planning Using the AFP

The following describes the AFPs automatic planning method for PRACH RSIs in an LTE network, which takes into account
interference matrices, neighbour relations (first-order neighbours and optionally second-order neighbours), and distance
They are active,

Their PRACH RSI status is not set to locked,
zone.

The AFP is based on a cost-based function which takes into account various relations between each TBA cell and its related
cells. Atoll calculates the cost between each individual TBA and related cell, and then the overall cost for the TBA cell.

into account. In this case, the assigned weights are: Neighbour = 0,25 and 2nd Neighbour = 0,10 .
Cells within the cells (or the default) reuse distance, if the check box "Reuse distance" is selected,
relative weights (%) defined in the Weights dialog box as follows:
% Neighbour
Neighbour = ---------------------------------------------------------------------------------------------------------------------------------% Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% 2nd Neighbour
2nd Neighbour = ---------------------------------------------------------------------------------------------------------------------------------% Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% IM
IM = ---------------------------------------------------------------------------------------------------------------------------------% Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% Dis tan ce
Dis tan ce = ---------------------------------------------------------------------------------------------------------------------------------% Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
580

AT332_TRR_E0

VL
TX ic TX jc
i
j
PRACH
p Coll
PRACH
= Min 1 p Coll
is the PRACH RSI collision probability given by
ID
p Coll
TX i ic TX j jc
N Common PRACH RSIs

.
= ----------------------------------------TX i ic
N Req PRACH RSIs
TX i ic TX j jc
TX i ic
Where N Common PRACH RSIs is the number of PRACH RSIs common between cells TXi(ic) and TXj(jc), and N Req PRACH RSIs is the
number of PRACH RSIs required by the cell TXi(ic).
TX i ic TX j jc
Neighbours
TX i ic TX j jc
TX i ic TX j jc
+ 2nd Neighbour 2nd Neighbour

TX i ic TX j jc
Where Neighbour
is calculated during automatic neighbour planning by Atoll as explained in the Technical Reference Guide. For manual
neighbour planning, this value is equal to 1.
planning. If the TBA cell has the same PRACH RSI assigned as one of its second-order neighbours, the importance of the PRACH
RSI collision is the multiple of the importance values of the first order neighbour relations between the TBA cell and its second
order neighbour. If the TBA cell is related to its second order neighbour through more than one first order neighbour, the
importance is the highest value among all the multiples:
2nd Neighbour =
TX ic TX jc
j
i
Neighbour
All Neighbour Pairs
Max
TX j jc TX k kc
Neighbour
with Collisions

TX i ic TX j jc
Interference
TX i ic TX j jc
IM
TX i ic TX j jc
IM
TX i ic TX j jc
= IM IM
TX i ic TX j jc
= r CCO
TX i ic TX j jc
TX i ic TX j jc
TX i ic TX j jc
f Overlap
and IM
IM CC
TX i ic TX j jc
TX i ic TX j jc
IM CC
TX i ic TX j jc
= IM CC
TX i ic TX j jc
and IM CC
TX i ic TX j jc
+ r ACO
TX i ic TX j jc
IM AC
otherwise.
TX i ic TX j jc
TX i ic TX j jc
, r CCO
TX i ic TX j jc
, and r ACO
are the total,
co-channel, and adjacent channel overlap ratios calculated as calculated in "Co- and Adjacent Channel Overlaps Calculation"
on page 493.
TX i ic TX j jc
Dis tan ce
them.
TX ic TX jc
i
j
Dis tan ce
TX i ic TX j jc
f Overlap
is
TX i ic TX j jc
= rO
calculated
as
explained
in
"Distance
Importance
TX i ic TX j jc
Calculation"
on
page 582.
= 1 otherwise.
QRF
TX i ic TX j jc
= 1 VL
TX i ic TX j jc
TX i ic TX j jc
Interference
TX i ic TX j jc
TX i ic TX j jc
+ Neighbours f Overlap

The total cost of the current PRACH RSI plan for any TBA cell is given as follows, considering all the cells with which the TBA
cell has relations:
581

TX ic
i
QRF
$ Total = 1
TX ic TX jc
i
j
TX jc
j
And, the total cost of the current PRACH RSI plan for the entire network is simply the sum of the total TBA cell costs calculated
above, i.e.,
$ Total =
TX i ic
$ Total
TX i ic

Calculates the cost (as described above) of the current PRACH RSI plan,
Tries different PRACH RSIs to cells in order to reduce the costs,
Stops when it is unable to improve the cost of the network, and proposes the last known best PRACH RSI plan as the
solution.
6.5.6 Appendices
6.5.6.1 Interference Matrix Calculation
The co-channel interference probability is calculated as follows:
S TX ic
i
TX ic TX jc
i
j
IM CC
TX j jc
TX i ic
n Sym
C Max + M Quality
-------------------- ---------------------------------------------------- TX ic
TX i ic
TX
ic
TX i ic
10
10
i
T i
C DLRS 10 Log 10
+ 10
2 N FB
RSRP n Sym
= --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX ic
i
The adjacent channel interference probability is calculated as follows:

S TX ic
i
TX i ic TX j jc
IM AC
TX j jc
TX i ic
TX i ic
n
C Max + M Quality + f ACS
Sym ------------------- ------------------------------------------------------------------------------ TX ic

TX i ic
TX
ic
TX i ic
10
10
i
T i
C DLRS 10 Log 10
+ 10
2 N FB
n Sym
RSRP
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX ic
i
For frequencies farther than the adjacent channel, the interference probability is 0.
TX i ic
TX i ic
Here S TX ic is the best server coverage area of the cell TXi(ic), that comprises all the pixels where E DLRS T RSRP
i
calculated in "Best Server Determination" on page 537. S TX ic
i
Condition
as
is the best server coverage area of the cell TXi(ic)
TX i ic
TX j jc
where the given condition is true. C DLRS is the received downlink reference signal level from the cell TXi(ic). C Max
TX i ic
received maximum signal level from the cell TXj(jc) calculated using the Max Power defined for this cell. n Sym
TX i ic
subcarrier noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 507 and N FB
is the
is the per-
is the total number of
frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell. M Quality is the quality
TX i ic
margin used for the interference matrices calculation. And, f ACS
is the adjacent channel suppression factor defined for the
frequency band of the cell TXi(ic).
6.5.6.2 Distance Importance Calculation

TX i ic TX j jc
The distance importance between two cells ( Dis tan ce
582

AT332_TRR_E0
TX ic TX jc
i
j
Dis tan ce
2
D Reuse
= Log --------------------------------
TX ic TX jc
j
D i

--------------------------------------------------------2
Log D Reuse
if D
TX ic TX jc
i
j
Otherwise
Where D Reuse is the minimum reuse distance, either defined for each TBA cell individually or set for all the TBA cells in the
AFP dialog box, and D
as follows:
D
D
TX ic TX jc
i
j
TX i ic TX j jc
them. d
= d
TX i ic TX j jc
TX ic TX jc
i
j
is the weighted distance between the TBA cell TXi(ic) and its related cell TXj(jc) calculated
1 + x cos cos 2
is weighted according to the azimuths of the TBA cell and its related cell with respect to the straight line joining
TX i ic TX j jc
is the distance between the two cells considering any offsets with respect to the site locations. x is set
TX i ic TX j jc
to 10 % so that the maximum variation in D

due to the azimuths does not exceed 40 %. and are calculated
from the azimuths of the two cells as shown in Figure 6.8 on page 583.
Figure 6.8: Weighted Distance Between Cells

The above formula implies that two cells facing each other will have a shorter effective distance between them than the real
distance, and two cells pointing in opposite directions will have a greater effective distance.
The importance of the distance relation is explained in Figure 6.9 on page 583. This figure shows that cells that are located
near (based on the effective distance which is weighted by the orientations of the cells) have high importance, which is
interpreted as a high cost, and cells that are located far have low importance. Cells that are further than the reuse distance
do not have any cost related to the distance relation.
Figure 6.9: Importance Based on Distance Relation
583

584
Chapter 7
3GPP Multi-RAT
Networks
"Multi-RAT Monte Carlo Simulations" on page 587
"Multi-RAT Coverage Predictions" on page 589

586

Chapter 7: 3GPP Multi-RAT Networks
AT332_TRR_E0
7 3GPP Multi-RAT Networks

This chapter describes the calculations specific to 3GPP multi-RAT documents. Multi-RAT calculations that are the same as
those in single-RAT documents can be found in:
"GSM GPRS EDGE Networks" on page 125,

"UMTS HSPA Networks" on page 209, and
"LTE Networks" on page 443.
7.1 Definitions
This table lists the input to coverage prediction and simulation calculations.
Name
Value
Unit
Description
f act
UL
Service parameter
None
f act
DL
Service parameter
None
TL DL GSM
Subcell parameter
Downlink traffic load (GSM)
7.2 Multi-RAT Monte Carlo Simulations


7.2.1 User Distribution

"Simulations Based on User Profile Traffic Maps" on page 588.

[Simulation]
RandomTotalUsers=0
In 3GPP multi-RAT documents, services can be classified under constant bit rate and variable bit rate services, which can be
provided by one or more technology. These service categories comprise the following service types in different technologies:
Constant Bit Rate Services
Variable Bit Rate Services
GSM GPRS EDGE
Circuit
Packet (Constant Bit Rate)
Packet (Max Bit Rate)
UMTS HSPA
Circuit R99
Packet HSPA (Constant Bit Rate)
Packet R99
Packet HSDPA (Best Effort)
Packet HSPA (Best Effort)
LTE
Voice
Data
587


points.
service usage characteristics described in the user profiles, i.e., the number of service sessions, the average duration of each
constant bit rate service session, or the volume of the data transfer in the uplink and the downlink in each variable bit rate
service session as explained in:
GSM and LTE: "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 477
UMTS: "Simulations Based on User Profile Traffic Maps" on page 223.
For any variable bit rate service (j), once several numbers of users with different activity statuses have been calculated
for different technologies, the final numbers of users are obtains as follows:
inactive
Number of inactive users: n j
inactive
= Average n j
inactive
nj
GSM
Number of users active on UL: n j UL = Average n j UL
GSM
Number of users active on DL: n j DL = Average n j DL
GSM
inactive
UMTS
nj
n j UL
UMTS
n j DL
UMTS
Number of users active on UL+DL: n j UL + DL = Average n j UL + DL
LTE
n j UL
LTE
n j DL
LTE
GSM
n j UL + DL
UMTS
n j UL + DL
LTE

Atoll calculates the number of active users of each service UL and DL as follows:
GSM and LTE: "Simulations Based on Sector Traffic Maps" on page 478
UMTS: "Simulations Based on Sector Traffic Maps" on page 227.
Distribution of Terminals
Terminals assigned to users depend on the percentages defined per traffic map and the technologies supported by each
terminal. For example, if the percentages of terminals are defined as follows:
Terminal A (GSM): 30 %
Terminal B (GSM+UMTS): 50 %
Terminal C (GSM+UMTS+LTE): 20 %
For users of services that can be provided by GSM, UMTS, or LTE:
Terminal A: 30 %
Terminal B: 50 %
588

AT332_TRR_E0
Terminal C: 20 %
For users of services that can be provided by UMTS or LTE:
Terminal B: 50/70 = 71.4 %
Terminal C: 20/70 = 28.6 %
For users of services that can be provided by LTE only, Terminal C will be assigned.
7.2.2 Simulation Process

Each Monte Carlo simulation is a snap-shot of the network where resource allocation is carried out. The steps of this algorithm
are listed below.
Serving cell/technology selection

For each mobile, Atoll searches for a serving cell of each supported and available technology as follows:
GSM: Atoll determines a best server based on the HCS layer/server selection algorithm. If no best server can be
found, the mobile will be considered rejected by GSM.
UMTS: Atoll determines a best server based on Ec/Io. If no best server can be found, the mobile will be considered
rejected by UMTS.
LTE: Atoll determines the best server based on RSRP or RS level and the serving cell selection method. If no best
server can be found, the mobile will be considered rejected by LTE.
Once the potential serving technologies have been identified, Atoll selects the highest priority as defined in the service
assigned to each mobile. The best server and technology assigned to each mobile remains unchanged for the rest of
the simulation.
Technology-wise Monte Carlo simulations as explained in:
GSM: "Radio Resource Management in GSM" on page 184

UMTS: "Power Control Simulation" on page 229
LTE: "Scheduling and Radio Resource Management" on page 556
7.3 Multi-RAT Coverage Predictions

Coverage predictions are calculated by determining the best server for each technology on each pixel and then determining
the selected display parameter within the best servers calculation area. Each pixel within the calculation area is considered
a non-interfering receiver. The properties of the non-interfering probe receiver are set by selecting a terminal, a mobility type,
and a service.
The resolutions of coverage predictions do not depend on the resolutions of the path loss matrices or the geographic data and
can be defined separately for each coverage prediction. Coverage predictions are generated using a bilinear interpolation
3GPP multi-RAT coverage predictions are combinations of corresponding single-RAT coverage predictions with specific
parameter settings. Multi-RAT coverage predictions include:
Effective Service Area Analysis (DL+UL)


The 3GPP multi-RAT effective service area is the combination of single-RAT effective service areas:
GSM Service Area Analysis (DL) is based on a coverage by coding scheme, as explained in "GPRS/EDGE Coverage Predictions" on page 140 or on a coverage by codec modes, as explained in "Circuit Quality Indicators Coverage Predictions" on page 149, depending on the type of service.
Radio conditions are evaluated over the HCS server area with a margin of 4 dB, on all the interfered subcells. Codec
modes and coding schemes are obtained from these radio conditions based on C/I+N without ideal link adaptation (as
explained in "Throughput Calculation Based on Interpolation Between C/N and C/(I+N)" on page 138). This implies
that a frequency plan has to be defined in order to obtain this GSM/GPRS/EDGE coverage.
UMTS Effective Service Area Analysis (EbNt) (DL+UL) is based on a combination of downlink and uplink service area
predictions, as explained in "Downlink Service Area Analysis" on page 298 and "Uplink Service Area Analysis" on
page 300. In the case of HSPA services, the coverage is based on a combination of HSDPA et HSUPA service areas as
explained in "HSDPA Prediction Study" on page 302 and "HSUPA Prediction Study" on page 307.
LTE Effective Service Area Analysis (DL+UL) is based on a combination of downlink and uplink service area predictions,
as explained in "Effective Signal Analysis Coverage Predictions" on page 472.
Two display options are available for this prediction:
589

Technologies: Each pixel displays the colour representing the visible technology having the highest priority defined in
the selected service.
Available Technologies: Pixels display the colour representing the combined areas over which a multi-technology terminal can be served. For instance, the GSM+UMTS area shows the union between the GSM and the UMTS service
areas as explained above.

The 3GPP multi-RAT throughput prediction is the combination of single-RAT throughput predictions:
GSM Packet Throughput Analysis (DL) explained in "Application Throughput Calculation" on page 139
The 3GPP multi-RAT effective RLC throughput is obtained from the maximum effective RLC throughput of the GSM
layer. The 3GPP multi-RAT application throughput from the maximum application throughput of the GSM layer.
R99 Service Area Analysis (EbNt) (DL) explained in "Downlink Service Area Analysis" on page 298 and HSDPA
Throughput Analysis (DL) explained in "HSDPA Prediction Study" on page 302
R99: The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the effective RLC
and application throughputs of the R99 layer (see "Downlink Service Area Analysis" on page 298 for more
information).
HSDPA: The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the effective
RLC and application throughputs of the HSDPA layer (see "HSDPA Prediction Study" on page 302 for more
information).
LTE Coverage by Throughput (DL) explained in "C/(I+N)-based Coverage Predictions" on page 473
The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the Effective RLC
Channel Throughput (DL) and the Application Channel Throughput (DL) (see "C/(I+N)-based Coverage Predictions" on
page 473 for more information).
Four display options are available for this prediction:
590
Effective RLC Throughput: The throughput on the RLC layer that a cell can provide to the selected terminal per pixel
taking into account possible transmission errors (BLER) for the highest priority technology.
Max Effective RLC Throughput: The maximum throughput on the RLC layer that a cell can provide to the selected
terminal per pixel taking into account possible transmission errors (BLER), considering all available technologies.
Application Throughput: The throughput on the application layer that a cell can provide to the selected terminal per
pixel taking into account possible transmission errors (BLER) for the highest priority technology.
Max Application Throughput: the maximum throughput on the application layer that a cell can provide to the selected
Chapter 8
3GPP2 Multi-RAT
Networks
"Multi-RAT Monte Carlo Simulations" on page 593
"Multi-RAT Coverage Predictions" on page 595

592

Chapter 8: 3GPP2 Multi-RAT Networks
AT332_TRR_E0
8 3GPP2 Multi-RAT Networks

This chapter describes the calculations specific to 3GPP2 multi-RAT documents. Multi-RAT calculations that are the same as
those in single-RAT documents can be found in:
"CDMA2000 Networks" on page 335, and

"LTE Networks" on page 443
8.1 Definitions
This table lists the input to coverage prediction and simulation calculations.
Name
Value
Unit
Description
f act
UL
Service parameter
None
DL
Service parameter
None
f act
8.2 Multi-RAT Monte Carlo Simulations


8.2.1 User Distribution


[Simulation]
RandomTotalUsers=0
In 3GPP2 multi-RAT documents, services can be classified under constant bit rate and variable bit rate services, which can be
provided by one or more technology. These service categories comprise the following service types in different technologies:
Constant Bit Rate Services
Variable Bit Rate Services
CDMA
Speech
1xRTT Data
1xEV-DO rev. 0
1xEV-DO rev. A (Guaranteed Bit Rate)
1xEV-DO rev. B (Guaranteed Bit Rate)
1xEV-DO rev. A (Best Effort)

1xEV-DO rev. B (Best Effort)
LTE
Voice
Data

593

points.
service usage characteristics described in the user profiles, i.e., the number of service sessions, the average duration of each
constant bit rate service session, or the volume of the data transfer in the uplink and the downlink in each variable bit rate
service session as explained in:
LTE: "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 477
CDMA: "Simulations Based on User Profile Traffic Maps" on page 357.
For any variable bit rate service (j), once several numbers of users with different activity statuses have been calculated
for different technologies, the final numbers of users are obtains as follows:
inactive
Number of inactive users: n j
inactive
= Average n j
inactive
CDMA
nj
Number of users active on UL: n j UL = Average n j UL
CDMA
Number of users active on DL: n j DL = Average n j DL
CDMA
LTE
n j UL
LTE
n j DL
LTE
Number of users active on UL+DL: n j UL + DL = Average n j UL + DL
CDMA
n j UL + DL
LTE

Atoll calculates the number of active users of each service UL and DL as follows:
LTE: "Simulations Based on Sector Traffic Maps" on page 478

CDMA: "Simulations Based on Sector Traffic Maps" on page 360.
8.2.2 Simulation Process

Each Monte Carlo simulation is a snap-shot of the network where resource allocation is carried out. The steps of this algorithm
are listed below.
Serving cell/technology selection

For each mobile, Atoll searches for a serving cell of each supported and available technology as follows:
CDMA: Atoll determines a best server based on Ec/Io. If no best server can be found, the mobile will be considered
rejected by CDMA.
LTE: Atoll determines the best server based on RSRP or RS level and the serving cell selection method. If no best
server can be found, the mobile will be considered rejected by LTE.
Once the potential serving technologies have been identified, Atoll selects the highest priority as defined in the service
assigned to each mobile. The best server and technology assigned to each mobile remains unchanged for the rest of
the simulation.
Technology-wise Monte Carlo simulations as explained in:
594
CDMA: "Simulations" on page 356

LTE: "Scheduling and Radio Resource Management" on page 556

AT332_TRR_E0
8.3 Multi-RAT Coverage Predictions

Coverage predictions are calculated by determining the best server for each technology on each pixel and then determining
the selected display parameter within the best servers calculation area. Each pixel within the calculation area is considered
a non-interfering receiver. The properties of the non-interfering probe receiver are set by selecting a terminal, a mobility type,
and a service.
The resolutions of coverage predictions do not depend on the resolutions of the path loss matrices or the geographic data and
can be defined separately for each coverage prediction. Coverage predictions are generated using a bilinear interpolation
3GPP2 multi-RAT coverage predictions are combinations of corresponding single-RAT coverage predictions with specific
parameter settings. Multi-RAT coverage predictions include:


The 3GPP2 multi-RAT effective service area is the combination of single-RAT effective service areas:
CDMA Effective Service Area Analysis (EbNt) (DL+UL) is based on a combination of downlink and uplink service area
predictions, as explained in "Downlink Service Area Analysis" on page 411 and "Uplink Service Area Analysis" on
page 415.
LTE Effective Service Area Analysis (DL+UL) is based on a combination of downlink and uplink service area predictions,
as explained in "Effective Signal Analysis Coverage Predictions" on page 472.
Two display options are available for this prediction:
Technologies: Each pixel displays the colour representing the visible technology having the highest priority defined in
the selected service.
Available Technologies: Pixels display the colour representing the combined areas over which a multi-technology terminal can be served. For instance, the CDMA+LTE area shows the union between the CDMA and the LTE service areas
as explained above.

The 3GPP2 multi-RAT throughput prediction is the combination of single-RAT throughput predictions:
Service Area Analysis (EbNt) (DL) explained in "Downlink Service Area Analysis" on page 411
The 3GPP2 multi-RAT effective RLC and application throughputs are respectively obtained from the effective RLC and
application throughputs (see "Downlink Service Area Analysis" on page 298 for more information).
LTE Coverage by Throughput (DL) explained in "C/(I+N)-based Coverage Predictions" on page 473
The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the Effective RLC
Channel Throughput (DL) and the Application Channel Throughput (DL) (see "C/(I+N)-based Coverage Predictions" on
page 473 for more information).
Four display options are available for this prediction:
Effective RLC Throughput: The throughput on the RLC layer that a cell can provide to the selected terminal per pixel
taking into account possible transmission errors (BLER) for the highest priority technology.
Max Effective RLC Throughput: The maximum throughput on the RLC layer that a cell can provide to the selected
Application Throughput: The throughput on the application layer that a cell can provide to the selected terminal per
pixel taking into account possible transmission errors (BLER) for the highest priority technology.
Max Application Throughput: the maximum throughput on the application layer that a cell can provide to the selected
595

596
Chapter 9
TD-SCDMA Networks
"Definitions and Formulas" on page 599
"Signal Level Based Calculations" on page 606
"Monte Carlo Simulations" on page 612
"TD-SCDMA Prediction Studies" on page 630
"Smart Antenna Modelling" on page 642
"N-Frequency Mode and Carrier Allocation" on page 654
"Neighbour Allocation" on page 655
"Scrambling Code Allocation" on page 660
"Automatic GSM/TD-SCDMA Neighbour Allocation" on

page 670

598

Chapter 9: TD-SCDMA Networks
AT332_TRR_E0
9 TD-SCDMA Networks
This chapter describes in detail the algorithms, calculation parameters, and processes of the coverage predictions and the
simulations available in TD-SCDMA documents. The first part of this chapter lists all the input and output parameters in the
TD-SCDMA documents, their significance, location in the Atoll GUI, and their usage. Detailed explanation of the basic coverage
predictions, which do not require simulation results, is provided in the second part. The third part describes the traffic
scenario generation and Montel Carlo simulation algorithms including smart antenna modelling and dynamic channel
allocation. The next sections are dedicated to TD-SCDMA coverage predictions which can be based on results obtained from
simulations. The last three sections describe in detail the allocation of frequencies, i.e., master and slave carriers, the
allocation of neigbours, and the allocation of scrambling codes.
9.1 Definitions and Formulas

The tables in the following subsections list the input and output parameters and formulas used in simulations and other
computations.
9.1.1 Inputs
This table lists the inputs to computations, coverage predictions, and simulations.
Name
Value
Unit
Description
R Ch
Global parameter
Mcps
Chip rate (or Spreading rate) (1.28)
Spread
Global parameter
None
Minimum spreading factor (1)
F Max
Spread
Global parameter
None
Maximum spreading factor (16)
Proc
Global parameter
None
P-CCPCH processing gain (13.8 dB)
N TS
SF
Global parameter
None
Number of timeslots per subframe

(7)
SF
Global parameter
ms
Subframe duration (5)
Frame
Global parameter
ms
Frame duration (10)
N Ch TS
GP
Global parameter
None
Number of guard period chips per

timeslot (16)
N Ch TS
Data
Global parameter
None
Number of data chips per timeslot

(704)
Midamble
Global parameter
None
Number of midamble chips per

timeslot (144)
N Ch PTS
Global parameter
None

pilot timeslot (96)
N Ch DwPTS
GP
Global parameter
None

DwPTS timeslot (32)
SYNC_DL
Global parameter
None
Number of SYNC_DL chips per

DwPTS timeslot (64)
None
Total number of chips per DwPTS

timeslot (96)
F Min
G P CCPCH
D
D
N Ch TS
GP
N Ch DwPTS
Total
N Ch DwPTS
Global parameter
Total
N Ch DwPTS
GP
SYNC_DL
= N Ch DwPTS + N Ch DwPTS
N Ch UpPTS
GP
Global parameter
None

UpPTS timeslot (32)
SYNC_UL
Global parameter
None
Number of SYNC_UL chips per UpPTS

timeslot (128)
None
Total number of chips per UpPTS

timeslot (160)
N Ch UpPTS
Total
N Ch UpPTS
Global parameter
Total
N Ch UpPTS
GP
SYNC_UL
= N Ch UpPTS + N Ch UpPTS
599

Name
Value
Unit
Description
Calculated global parameter

Data
N Ch TS
W = --------------SF
D
bps
Chip rate (140800 bps)
F Avg
MHz
Average frequency range of the

frequency band (2010)
BW
MHz
Channel bandwidth of the carriers of

a frequency band (1.6)
F IRF
Cell parameter
None
F JD
None
Joint Detection (JD) factor
TX
None
Multi-Cell Joint Detection factor
(user-defined or calculated from transmitter equipment
characteristics)
None
BTS Noise Figure
(user-defined or calculated from transmitter equipment
characteristics)
None
Transmitter loss
L Tx = L Total UL on uplink
TX
BTS parameter
None

transmitted
P TCH
Max
Cell parameter
Maximum cell traffic timeslot power
P P CCPCH
Cell parameter
P-CCPCH power on TS0
P DwPCH
Cell parameter
DwPCH power on DwPTS
P OCCH TS0
Cell parameter
Other common channel power on

TS0
TComp P CCPCH
Cell parameter
None
P-CCPCH RSCP comparative

threshold for baton handover
P Max
Cell parameter
None
Maximum difference between two

transmitted powers
Req
Cell parameter
None
Required resource units in uplink
RU DL
Req
Cell parameter
None
Required resource units in downlink
P HS PDSCH
Available
Cell parameter
HS-PDSCH power available per

downlink timeslot
P HR
Cell parameter
None
Power headroom
P HS SCCH
Cell parameter
HS-SCCH power per downlink

timeslot
N HS SCCH
Cell parameter
None
Number of HS-SCCH channels
N HS SICH
Cell parameter
None
Number of HS-SICH channels
Max
Cell parameter
None
Maximum number of HSDPA users
N HS-PDSCH Codes
Min
Cell parameter
None
Minimum number of HS-PDSCH

codes
Max
Cell parameter
None
Maximum number of HS-PDSCH

codes
Max
Cell parameter
None
Maximum number of intratechnology neighbours
Max
Cell parameter
None
Maximum number of intertechnology neighbours
TX
F MCJD
NF
TX
TX
RU UL
N HSDPA
N HS-PDSCH Codes
N Intra Neigh
N Intra Neigh
600
L Tx = L Total DL on downlink

AT332_TRR_E0
Name
Value
Unit
Description
RSCP P CCPCH
Min
The minimum P-CCPCH RSCP

required for a user to be connected
to the cell
P OCCH
Timeslot parameter
Other common channel power
DL
Timeslot parameter
Downlink traffic power
Timeslot parameter
(Simulation constraint)
None
Maximum percentage of downlink

used power
Timeslot parameter
(Simulation result)
None
Uplink load factor
Timeslot parameter
(Simulation constraint)
None
P HS PDSCH
Timeslot parameter
HS-PDSCH power available
Min
Timeslot parameter
None
Minimum number of HS-PDSCH

codes
N HS-PDSCH Codes
Max
Timeslot parameter
None
Maximum number of HS-PDSCH

codes
RU Overhead
Timeslot parameter
P TCH
Max
%P DL
X UL
Max
X UL
Available
N HS-PDSCH Codes
Overhead resource units
Body
Service parameter
None
Body loss
Act
Service parameter
None
Downlink activity factor for circuitswitched services and the A-DPCH

activity factor for HSDPA services
f UL
Act
Service parameter
None
Uplink activity factor for circuitswitched services and the A-DPCH

activity factor for HSDPA services
f DL
Eff
Service parameter
None
Downlink efficiency factor for circuitswitched services
f UL
Eff
Service parameter
None
Uplink efficiency factor for circuitswitched services
F Scaling
Service parameter
None
Application througput scaling factor
O TP
Service parameter
kbps
Application throughput offset
UL
Service parameter (packet session modelling)
None
Average number of packet calls on

the uplink during a session
DL
None
Average number of packet calls on

the downlink during a session
UL
ms
Average time between two packet

calls on the uplink
T PacketCall
DL
ms
Average time between two packet

calls on the downlink
UL
KBytes
Minimum packet call size on the

uplink
DL
KBytes
Minimum packet call size on the

downlink
UL
KBytes
Maximum packet call size on the

uplink
S Max PacketCall
DL
KBytes
Maximum packet call size on the

downlink
T Packet
UL
ms
Average time between two packets

on the uplink
DL
ms
Average time between two packets

on the downlink
f DL
N PacketCall
N PacketCall
T PacketCall
S Min PacketCall
S Min PacketCall
S Max PacketCall
T Packet
601

Name
Value
Unit
Description
UL
Bytes
Packet size on uplink
DL
Bytes
Packet size on downlink
Nom
kbps
Downlink peak throughput
Nom
kbps
Uplink peak throughput
Proc
WR99 bearer parameter (Can be calculated as ----------)

Nom
R DL
None
Downlink processing gain
Proc
WR99 bearer parameter (Can be calculated as ----------)

Nom
R UL
None
Uplink processing gain
Min
Allowed minimum downlink traffic

channel power
Max
Allowed maximum downlink traffic

channel power
N DL
TS
None
Number of downlink timelots
TS
None
Number of uplink timelots
Req
E Req
C Req
R99 bearer parameter per mobility ( ----b-
or ---
)
N t TCH UL
I TCH UL
None
Eb/Nt or C/I target on uplink
Req
E Req
C Req
R99 bearer parameter per mobility ( ----b-
or ---
)
N t TCH DL
I TCH DL
None
Eb/Nt or C/I target on downlink
Req
R99 bearer parameter per mobility
Target RSCP on uplink TCH
Req
Target RSCP on downlink TCH
Div
None
Downlink diversity gain
Div
None
Uplink diversity gain
Term
Terminal parameter
Maximum terminal power
P Min
Term
Terminal parameter
Minimum terminal power
P UpPCH
Terminal parameter
UpPCH power
Term
Terminal parameter
None
Term
Terminal parameter
None
Joint Detection (JD) factor
Term
Terminal parameter
None

Term
Terminal parameter
None
Terminal gain
Term
Terminal parameter
None
Terminal loss
TAdd P CCPCH
Mobility parameter
Required RSCP T_Add for P-CCPCH
TDrop P CCPCH
Mobility parameter
Required RSCP T_Drop for P-CCPCH
Req
Mobility parameter
Required RSCP threshold for DwPCH
Req
Mobility parameter
Required RSCP threshold for UpPCH
E Req
C Req
Mobility parameter ( ----b-
or ---
)
N t P CCPCH
I P CCPCH
None
Required quality threshold for PCCPCH
S Packet
S Packet
R DL
R UL
G DL
G UL
P TCH DL
P TCH DL
N UL
Q TCH UL
Q TCH DL
RSCP TCH UL
RSCP TCH DL
G DL
G UL
P Max
NF
F JD
G
L
RSCP DwPCH
RSCP UpPCH
Req
Q P CCPCH
602

AT332_TRR_E0
Name
Value
Unit
Description
Req
E Req
Mobility parameter ( ----c-
)
N t HS SCCH
None
Required quality threshold for HSSCCH
Req
E P CCPCH
Mobility parameter ( ----c-
)
N t HS SICH
None
Required quality threshold for PCCPCH
C
Mobility parameter ( ---
)
I DwPCH
None
Required quality threshold for

DwPCH
Clutter class parameter
None
Model standard deviation
None
P-CCPCH Eb/Nt or C/I standard

deviation
None
Downlink Eb/Nt or C/I standard

deviation
None
Uplink Eb/Nt or C/I standard

deviation
None
Indoor loss
Ortho
None
Downlink orthogonality factor
F UL
Ortho
None
Uplink orthogonality factor
Spread
Spreading angle
1.38 x 10-23
J/K
Boltzman constant
293
Ambient temperature
TX
NFTX K T BW
Term
NF Term K T BW
TX
Antenna parameter
None
None
Path loss

standard deviation
None
Model shadowing margin used in

coverage predictions
P CCPCH
Result calculated from cell edge coverage probability and P-CCPCH

Eb/Nt standard deviation
None
P-CCPCH Eb/Nt shadowing margin

used in coverage predictions
Eb Nt DL

standard deviation
None
DL Eb/Nt shadowing margin used in

Eb Nt UL

standard deviation
None
UL Eb/Nt shadowing margin used in

Q HS SCCH
Q HS SICH
Req
Req
Q DwPCH
Model
Eb/Nt
P CCPCH or
CI
P CCPCH
CI
Eb/Nt
or DL
Eb/Nt
or UL
DL
CI
UL
L Indoor
F DL
N0
N0
L Path
Model
M Shadowing
M Shadowing
M Shadowing
M Shadowing
603

Name
Value
Unit
For RSCP calculation

Model
LT
TX
Term
Body
Transmitter-terminal total loss in

Model
M Shadowing
L Path L L
L
L Indoor
= ---------------------------------------------------------------------------------------------------------------------TX
Term
G G
In UL, only carrier power is

Eb Nt UL
For P-CCPCH Eb/Nt calculation

TX
Term
attenuated by M Shadowing .
P CCPCH
Body
L Path L L
L
L Indoor M Shadowing
= ---------------------------------------------------------------------------------------------------------------------TX
Term
G G
Eb Nt P
LT
LT
None

Eb Nt DL
LT
TX
Term
Eb Nt UL
LT
Eb Nt DL
P CCPCH
extra-cell interference is not.

Eb Nt

Term
In DL, carrier power and intra-cell

interference are attenuated by
M Shadowing or M Shadowing while
Eb Nt DL
Body
L Path L L
L
= ---------------------------------------------------------------------------------------------------------------------TX
Term
G G
TX
Description
Eb Nt UL
Body
DL
Therefore, M Shadowing or
L Path L L
L
= ---------------------------------------------------------------------------------------------------------------------TX
Term
G G
P CCPCH
M Shadowing are set to 1 in DL extracell interference calculation.
9.1.2 P-CCPCH Eb/Nt and C/I Calculation

Name
Value
E b TX i ic
--- N t P CCPCH
RSCP P CCPCH
------------------------------------------------ G Proc
P CCPCH
TX i ic
N Tot DL
TX i ic
RSCP P CCPCH
-----------------------------------------------TX ic
i
N Tot DL
TX i ic
N Tot DL
TX
TX i ic
TX i ic
Term
I Intra DL + I Extra DL + I IC DL ic jc + N 0
TX ic
i
TX i ic
TX i ic
I Extra DL
I IC DL ic jc
With
TX i
None
P-CCPCH Eb/Nt for the cell TX i ic
None
P-CCPCH C/I for the cell TX i ic
Downlink total noise for the cell

TX i ic
Downlink intra-cell interference for

the cell TX i ic
Downlink extra-cell interference for

the cell TX i ic
Inter-carrier interference
Unit
Description
None
DwPCH C/I for the cell TX i ic

TX i ic
TX ic
i
RSCP P CCPCH
I Intra DL
Description
TX i ic
TX i
C
---
I P CCPCH
Unit
TX i
TX
TX ic
i
+ RSCP OCCH TS0

Ortho
1 F DL
= 0
1
Term
1 F JD
TX
and
Without Useful Signal

Total Noise
TX j ic
TX j ic
TX j jc
TX j jc
RSCPP CCPCH + RSCPOCCH TS0

ji

TX
j
--------------------------------------------------------------------------------------F IRF ic jc
9.1.3 DwPCH C/I Calculation

Name
Value
TX i
TX ic
i
RSCP DwPCH
------------------------------------------TX i ic
N Tot DL
C
---
I DwPCH
TX i ic
N Tot DL
604
TX i ic
TX i ic
TX i ic
Term

AT332_TRR_E0
Name
Value
TX ic
i
RSCP DwPCH
TX ic
i
I Intra DL
With
TX
TX
Ortho
1 F DL
= 0
1
TX
Description
Intra-cell interference for the cell

TX i ic
Extra-cell interference for the cell

TX i ic
Unit
Description
None
Downlink TCH Eb/Nt for the cell

TX i ic
None
Downlink TCH C/I for the cell TX i ic

TX i ic
Downlink intra-cell interference for

the cell TX i ic
Downlink extra-cell interference for

the cell TX i ic
Unit
Description
None
Uplink TCH Eb/Nt for the cell TX i ic
None
Uplink TCH C/I for the cell TX i ic
Uplink required power for the

terminal
i
Term
1 F JD
and

Total Noise
TX j ic
RSCPDwPCH
TX ic
i
I Extra DL
Unit
ji
TX j jc
RSCPDwPCH
I IC DL ic jc
TX
j
---------------------------------------F IRF ic jc
9.1.4 DL TCH Eb/Nt and C/I Calculation

Name
Value
TX i ic
TX i
TX ic
i
RSCP TCH DL
Div
---------------------------------------------- G Proc
DL G DL
TX i ic
N Tot DL
TX ic
RSCP TCH DL
---------------------------------------------- G Div
DL
TX i ic
N Tot DL
E b
--- N t TCH DL
TX i ic
N Tot DL
TX i ic
I Intra DL
TX i ic
I Extra DL
I IC DL ic jc
TX i ic
TX i
i
C
---
I TCH DL
TX i ic
TX i ic
Term
TX i
TXi
Ortho
Term
1 F DL 1 F JD + 1
TX ic
TX ic
i
i
RSCP TCH DL + RSCP OCCH
TX ic
j
TX ic
j
TX j jc
TX j jc
RSCPTCH DL + RSCPOCCH
ji
TX
j
--------------------------------------------------------------------------F IRF ic jc
9.1.5 UL TCH Eb/Nt and C/I Calculation

Name
Value
TX i ic
Term
TX i ic
RSCP TCH UL
Div
------------------------------------------------- G Proc
UL G UL
TX i ic
N Tot UL
i
C
---
I TCH UL
TX ic
RSCP TCH UL
------------------------------------------------- G Div
UL
TX i ic
N Tot UL
Term
P Req
Q TCH UL
Q TCH UL
Term
- or P Term
P Max --------------------------Max ------------------------TX i ic
TX i ic
E
C
b
---
-----
I TCH UL
N t TCH UL
E b
--- N t TCH UL
Term
Req
TX i ic
Req
605

9.1.6 Interference Calculation

Name
Value
TX jc
j
I C2C TX i TX j
TX j ic
TX j ic
Unit
Description
Cell to cell interference
UpPCH interference
Unit
Description
HS-SCCH power
HS-PDSCH power
HS-SICH power
TX jc
j
j
RSCPTCH DL + RSCPOCCH +--------------------------------------------------------------------------F IRF ic jc
TX
TX j
TX i
TX i ic
N0
I TS1 UL
TX i ic
X TS1 UL
---------------------------------TX ic
1 X i
TS1 UL
9.1.7 HSDPA Dynamic Power Calculations

Name
Value
TX ic
i
TX ic
TX ic
Ec
i
i
i
--- N
RSCP HS SCCH
N t HS SCCH Tot DL
----------------------------------------------------------------------------------------------------------------------------- L Model
T
TX i
TX i ic
P HS SCCH
TX i ic
TX i ic
TX
TX i ic
TX i ic
P Max DL Eff P R99 DL P HR
P HS PDSCH
TX i ic
TX ic
TX i ic
P HS SCCH
Ec
i
i
i
---
N t HS SICH N Tot UL RSCP HS SICH

-------------------------------------------------------------------------------------------------------------------------- L Model
T
Mi
Mi
P HS SICH
9.2 Signal Level Based Calculations

Two types of signal level based calculations are available in Atoll:
1. Point Analysis: Real-time calculations for profile and reception analysis using the mouse to move a probe mobile on
the map.
2. RSCP Based Coverage Predictions: Calculation of RSCP related parameters on each pixel and colouring according to
the selected display.

For the selected transmitted TXi and carrier (ic), you can study three parameters in point analysis Profile tab:
Study criteria
Formulas
Signal level ( RSCP ) in dBm
RSCP
TX i ic
TX i ic
Model
L Path M Shadowing L Indoor

TX i
Path loss ( L Path ) in dB

Total losses ( L T ) in dB
= EIRP

L T = L Path + L
TX
Model
+ L Indoor + M Shadowing G
TX
Where,
RSCP is the received signal code power for the P-CCPCH.
EIRP is the effective isotropic radiated power of the transmitter. EIRP
TX i ic
TX i ic
= P P CCPCH + G
TX i
TX i
ic is a carrier number
L Model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model
606

AT332_TRR_E0
TX
L Ant is the transmitter antenna attenuation (from antenna patterns)

Model
M Shadowing is the shadowing margin. This parameter is taken into account when the option Shadowing taken into account
is selected
L Indoor are the indoor losses, taken into account when the option "Indoor coverage" is selected
G
L
TX i
TX i
is the transmitter antenna gain

is the transmitter loss ( L
TX i
= L Total DL )
It is possible to analyse the best carrier. In this case, Atoll takes the highest P-CCPCH
power of cells to calculate the signal level received from a transmitter.
9.2.1.1 Profile Tab

TX i ic
Atoll displays either the signal level received from the selected transmitter on a carrier ( RSCP P CCPCH ), or the highest signal
level received from the selected transmitter on the best carrier.
For a selected transmitter, it is also possible to study the path loss, L Path , or the total
losses, L T . Path loss and total losses are the same on any carrier.

Analysis provided in the Reception tab is based on path loss matrices. You can study reception from TBC transmitters for which
path loss matrices have been calculated on their calculation areas.
TX i ic
For each transmitter, Atoll displays either the signal level received on a carrier, ( RSCP P CCPCH ), or the highest signal level
received on the best carrier.
Received signal level bar graphs are displayed in a decreasing signal level order. The number of bars in the graph depends on
the signal level received from the best server. Only bars for transmitters whose signal level is within a 30 dB margin from the
best server signal are displayed.
You can use a value other than 30 dB for the margin from the best server signal level, for
example a smaller value for improving the calculation speed. For more information on
defining a different value for this margin, see the Administrator Manual.
9.2.2 RSCP Based Coverage Predictions

For each TBC transmitter, TXi, Atoll determines the value of the selected parameter on each studied pixel inside the TXi
calculation area. Each pixel within the TXi calculation area is considered a probe receiver.
The study conditions to determine the service area of each TBC transmitter
The display settings to for colouring the covered pixels
Atoll uses the parameters entered in the Condition tab of the coverage study properties dialog box to determine pixels
covered by the each transmitter. Coverage prediction display resolution is independent of the path loss matrix and geographic
data resolutions, and can be different for each coverage prediction. Coverage predictions are calculated using bilinear
interpolation of multi-resolution path loss matrices (similar to the evaluation of site altitudes).
9.2.2.1 Calculation Criteria

The RSCP from a transmitter TXi and a selected carrier (ic) is given by:
RSCP
TX i ic
= EIRP
TX i ic
Model
L Path M Shadowing L Body L Indoor + G
Term
Term
Where,
607

RSCP is the received signal code power. RSCP can be calculated for P-CCPCH, DwPCH, or the downlink TCH.
EIRP is
the
TX ic
i
EIRP DwPCH
effective
TX ic
i
P DwPCH
+G
isotropic
TX
TX
, or
radiated
power
TX ic
i
EIRP DL TCH
of
the
TX ic
i
P DL TCH
+G
transmitter.
TX
TX
TX ic
i
TX ic
i
EIRP P CCPCH = P P CCPCH + G
TX
TX
ic is a carrier number
TX i

TX i

Model
is selected
L
Term
is the terminal loss
L Body is the body loss defined in the service

G
G
L
Term
TX i
TX i
is the receiver total gain

TX i
= L Total DL )
9.2.2.2 P-CCPCH RSCP Coverage Prediction

9.2.2.2.1
Coverage Condition
This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the P-CCPCH.
The coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The
best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest PCCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage
prediction is calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any
pixels covered by this transmitter. If you perform this coverage prediction for the best carrier, Atoll calculates the RSCP
considering:
the preferred carrier of the selected service, or

the carrier with the highest P-CCPCH power, if no preferred carrier is defined for the service, or
the master carrier in case of N-frequency mode compatible transmitters.
You can select the display colours according to the RSCP, or on any best server parameter.
9.2.2.2.2
Coverage Display
It is possible to colour the transmitter service areas using a unique colour per transmitter, or colour the pixels in the coverage
areas by any transmitter attribute or other criteria such as:
Best Signal Level (dBm)

TX i ic
Atoll calculates the best RSCP P CCPCH received from each transmitter TX i ic on each pixel. Where other service
areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is coloured if the RSCP level is
greater than or equal to the defined thresholds. The pixel colour depends on the RSCP level. Coverage consists of
several independent layers whose visibility in the workspace can be managed. There are as many layers as defined
thresholds. Each layer corresponds to an area where the RSCP from the best server exceeds a defined minimum
threshold.
RSCP Margin (dB)

Coverage consists of several layers with a layer per user-defined RSCP margin defined in the Display tab (Prediction
TX i ic
RSCP
properties). For each layer, area is covered if RSCP P CCPCH TAdd P CCPCH Mobility M P CCPCH . Each layer is
608

AT332_TRR_E0

TX ic
i
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the RSCP P CCPCH from
the transmitter exceeds TAdd P CCPCH defined in the mobility selected in the Conditions tab, with different cell edge
coverage probabilities. There is one coverage area per transmitter in the explorer.
9.2.2.3 Best Server P-CCPCH Coverage Prediction

This coverage prediction calculates and displays the best server RSCP for the P-CCPCH. The coverage prediction is calculated
for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The best server for the coverage prediction
is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master carrier in
case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is calculated for the selected carrier.
If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this transmitter. If you perform
this coverage prediction for the best carrier, Atoll calculates the RSCP considering:

The pixels in the TX i ic

TX i ic
RSCP P CCPCH =
TX i ic
Min
coverage area where RSCP P CCPCH Max (TAdd P CCPCH,RSCP P CCPCH)
and where
Best RSCP TXj jc will be covered and coloured according to the transmitter colour.
P CCPCH
j = All
9.2.2.4 P-CCPCH Pollution Analysis Coverage Prediction

This coverage prediction calculates and displays the number of P-CCPCH polluters. Atoll calculates the Received Signal Code
TX i ic
Power (RSCP) for the P-CCPCH for each pixel in the TX i ic coverage area where RSCP P CCPCH TAdd P CCPCH Mobility
and determines the polluting transmitters according to:
TX i ic
TX j jc
RSCP P CCPCH Best RSCP P CCPCH M
ji
Where M is the specified pollution margin.

considering:

Atoll determines the number of transmitters covering each pixel and colours the pixel according to the number of polluting
transmitters. Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as
many layers as defined thresholds. Each layer corresponds to an area where the number of servers is greater than or equal to
a defined minimum threshold.
9.2.2.5 DwPCH RSCP Coverage Prediction

9.2.2.5.1
Coverage Condition
This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the DwPCH. The coverage
prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for DwPTS. The best server
for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power,
or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is
calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered
by this transmitter. If you perform this coverage prediction for the best carrier, Atoll calculates the RSCP considering:

609

The pixels in the TX i ic

TX ic
i

TX ic
i
Min
coverage area where RSCP P CCPCH Max (TAdd P CCPCH,RSCP P CCPCH)
and where
Req
RSCP DwPCH RSCP DwPCH Mobility are covered and coloured according to the selected display parameter.
9.2.2.5.2
Coverage Display
DwPCH RSCP (dBm)

TX i ic
Atoll calculates the best RSCP DwPCH received from each transmitter TX i ic on each pixel.. Where other service
areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is coloured if
TX i ic
Req
RSCP DwPCH RSCP DwPCH Mobility . The pixel colour depends on the RSCP level. Coverage consists of several
independent layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds.
Each layer corresponds to an area where the RSCP from the best server exceeds a defined minimum threshold.
RSCP Margin (dB)

TX ic
i
Req
RSCP
properties). For each layer, area is covered if RSCP DwPCH RSCP DwPCH Mobility M DwPCH . Each layer is assigned
Cell edge coverage probability (%)

TX ic
i
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the RSCP DwPCH from
TX i ic
the transmitter TX i ic exceeds RSCP DwPCH defined in the mobility selected in the Conditions tab, with different cell
edge coverage probabilities. There is one coverage area per transmitter in the explorer.
9.2.2.6 UpPCH RSCP Coverage Prediction

9.2.2.6.1
Coverage Condition
This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the UpPCH in the uplink. The
coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for UpPTS. The best
server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH
power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction
is calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered

Atoll uses the UpPCH power of the selected terminal to calculate the RSCP from each pixel of each transmitters best server
coverage area.
TX i ic
Min
Term
Req
The pixels where RSCP P CCPCH Max (TAdd P CCPCH,RSCP P CCPCH) and where RSCP UpPCH RSCP UpPCH Mobility are
covered and coloured according to the selected display parameter.
9.2.2.6.2
Coverage Display
UpPCH RSCP (dBm)

Term
Atoll calculates the best RSCP UpPCH received from each pixel of each transmitter service area at the transmitter.
Where other service areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is coloured
Term
Req
if RSCP UpPCH RSCP UpPCH Mobility . The pixel colour depends on the RSCP level. Coverage consists of several
independent layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds.
Each layer corresponds to an area where the RSCP at the best server exceeds a defined minimum threshold.
610
RSCP Margin (dB)

AT332_TRR_E0
Term
Req
RSCP
properties). For each layer, area is covered if RSCP UpPCH RSCP UpPCH Mobility M UpPCH . Each layer is assigned a

Term
On each pixel of each transmitter service area, the coverage corresponds to the pixels from which the RSCP UpPCH at
Term
the transmitter exceeds RSCP UpPCH defined in the mobility selected in the Conditions tab, with different cell edge
coverage probabilities. There is one coverage area per transmitter in the explorer.
9.2.2.7 Baton Handover Coverage Prediction

9.2.2.7.1
Coverage Condition
This coverage prediction determines the pixels which receive RSCP from cells other than the best server high enough to
perform baton handovers. Received Signal Code Power (RSCP) is calculated for the P-CCPCH. The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The best server for the coverage
prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master
carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is calculated for the
selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this transmitter.
If you perform this coverage prediction for the best carrier, Atoll calculates the RSCP considering:
The

pixels
TX ic
i
RSCP P CCPCH
9.2.2.7.2
are
covered
and
coloured
TAdd P CCPCH Mobility and
according
TX jc
j
RSCP P CCPCH
to
the
selected
TDrop P CCPCH Mobility
display
parameters,
TX jc
j
TComp P CCPCH
where
Coverage Display
It is possible to display the potential handover areas or the number of transmitters covering each pixel.
Handover Areas
Atoll displays the pixels where there are transmitters other than the best server that satisfy the above criteria.
Coverage consists of a single layer with a defined colour whose visibility in the workspace can be managed.
Number of Potential Servers

Atoll determines the number of transmitters covering each pixel and colours the pixel according to the number of
transmitters. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the number of servers is
greater than or equal to a defined minimum threshold.
9.2.2.8 Scrambling Code Interference Analysis

This coverage prediction calculates and displays the pixels covered by two cells using the same scrambling code. Atoll
calculates the Received Signal Code Power (RSCP) for the P-CCPCH for each pixel in the TX i ic coverage area where
TX ic
i
RSCP P CCPCH TAdd P CCPCH Mobility and determines the interfering transmitters according to:
TX i ic
TX j jc
RSCP P CCPCH Best RSCP P CCPCH M
ji
Where M is the specified pollution margin.

considering:

Atoll determines whether the cells of two transmitters covering a pixel have the same scrambling code. If the pixel is
interfered, Atoll colours it according to the colour assigned to the scrambling code in the display parameters. Coverage
611

consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as
scrambling codes. Each layer corresponds to the area where the corresponding scrambling code has interference. A layer
corresponding to areas where more than one scrambling code interferes is also available.
9.3 Monte Carlo Simulations

Generating a realistic user distribution as explained in "Generating a Realistic User Distribution" on page 612.
Dynamic channel allocation and power control as explained under "Power Control Simulation" on page 617.


[Simulation]
RandomTotalUsers=0
Each user is randomly assigned a service, a terminal, and a mobility type. The activity status is determined based on the
calculations of activity probabilities using the traffic inputs.
The user activity status influences the next step of the simulation, i.e., the interference in the network. Both active and
inactive users use radio resources and generate interference.
indoor users per clutter class defined for the traffic maps.
Atoll also calculates the shadowing margin for each user based on the standard deviations defined for the clutter class of each
user.
In TD-SCDMA networks users accessing packet-switched services can transmit either on uplink or on downlink, but never on
both simultaneously. Users accessing circuit-switched services transmit on both uplink and downlink simultaneously. Circuitswitched service users, mobiles connected in uplink and downlink both, are modelled in Atoll by two mobiles generated at
the same location with one connected on the uplink and the other on the downlink. If one of these two mobiles is rejected
for some reason, the other is also rejected due to the same reason.
9.3.1.1 Simulations Based on User Profile Traffic Maps

The user profile models the behaviour of the different user categories. Each user profile contains a list of services and their
associated parameters describing how these services are accessed by the user.
612

AT332_TRR_E0
In case of user profile traffic maps composed of lines, the number of users per user
profile is calculated from the line length (L) and the user profile density (DUP) (users
per km): N Users = L D UP
The number of users is an input when a user profile traffic map is composed of points.
At any given instant, Atoll calculates the probability for a user being active in the uplink and in the downlink according to the
duration of each voice call, or the volumes of the data exchanged in the uplink and the downlink in each data session.
9.3.1.1.1
Circuit Switched Service (i)

User profile parameters for circuit switched services are:
The user terminal equipment used for the service (from the Terminals table),
The average number of calls per hour N Call ,
N call d
p o = ------------------3600
Calculation of the number of users trying to access the service i ( n i ):
n i = N Users p 0
The activity status of each user depends on the activity periods during the connection, i.e., the uplink and downlink activity
UL
DL
factors defined for the circuit switched service i, f Act and f Act .

UL
DL

UL
UL
DL
DL
DL
UL
Probability of being active on UL: p Active = f Act 1 f Act

Probability of being active on DL: p Active = f Act 1 f Act
UL + DL
UL
DL
Probability of being active both on UL and DL: p Active = f Act f Act

Number of inactive users: n i Inactive = n i p Inactive
UL
UL
Number of users active in the uplink: n i Active = n i p Active

DL
DL
Number of users active in the downlink: n i Active = n i p Active

UL + DL
UL + DL
Number of users active in the uplink and downlink both: n i Active = n i p Active
Therefore, a connected user can be either active on both links, inactive on both links, active on UL only, or active on DL only.
9.3.1.1.2
Packet Switched Service (j)

User profile parameters for packet switched services are:
The user terminal equipment used for the service (from the Terminals table),
The average number of packet sessions per hour N Sess ,
The volume (in kBytes) which is transferred on the downlink V
DL
and the uplink V
UL
during a session.
A packet session consists of several packet calls separated by a reading time. Each packet call is defined by its size and may be
divided in packets of fixed size (1500 Bytes) separated by an inter-packet arrival time.
613

Figure 9.1: Description of a Packet Session

Calculation of the average packet call size (kBytes):
UL
DL
UL
DL
V
V
S PacketCall = -------------------------------------- and S PacketCall = --------------------------------------UL
UL
DL
DL
N PacketCall f Eff
N PacketCall f Eff
UL
DL
In case of HSDPA services, f Eff and f Eff are the uplink and downlink A-DPCH activity factors, respectively.
Calculation of the average number of packets per packet call:
UL
DL
S PacketCall
S PacketCall
UL
- + 1 and N DL
- + 1
N Packet = Int ------------------------------Packet = Int ------------------------------ S UL
S DL
Packet 1024
Packet 1024
1 kBytes = 1024 Bytes.
Calculation of the average duration of inactivity within a packet call (c):

UL
UL
DL
DL
N Packet 1 T Packet
N Packet 1 T Packet
UL
DL
D Inactivity PacketCall = --------------------------------------------------------- and D Inactivity PacketCall = --------------------------------------------------------1000
1000
Calculation of the average duration of inactivity in a session (s):
UL
UL
UL
DL
DL
DL
D Inactivity Session = N PacketCall D Inactivity PacketCall and

D Inactivity Session = N PacketCall D Inactivity PacketCall
Calculation of the average duration of activity in a session (s):
UL
UL
DL
DL
N Packet S Packet 8
UL
UL
- and
D Activity Session = N PacketCall -----------------------------------------------UL
R Nom 1000
N Packet S Packet 8
DL
DL
D Activity Session = N PacketCall -----------------------------------------------DL
R Nom 1000
Therefore, the average duration of a connection in the session s is:
UL
UL
UL
DL
DL
DL
D Connection = D Activity Session + D Inactivity Session and D Connection = D Activity Session + D Inactivity Session
Calculation of the service usage duration per hour (probability of a connection):
N Sess
N Sess
UL
DL
DL
- D UL
p Connection = ----------Connection and p Connection = ------------ D Connection
3600
3600
Calculation of the probability of being connected:
614

AT332_TRR_E0
UL
DL
p Connected = 1 1 p Connection 1 p Connection

Therefore, the number of users trying to access the service j is:
n j = N Users p Connected
As Figure 9.1 on page 614 shows, there can be three possible cases when a user is connected:
a. 1st case: At a given time, packets are downloaded and uploaded.
UL
DL
p Connection p Connection
UL + DL
The probability of being connected is: p Connected = --------------------------------------------------------p Connected
b. 2nd case: At a given time, packet are uploaded only.
UL
DL
UL
The probability of being connected is: p Connected = ----------------------------------------------------------------------p Connected
c. 3rd case: At a given time, packet are downloaded only.
DL
UL
DL
The probability of being connected is: p Connected = ----------------------------------------------------------------------p Connected
Calculation of the probability of being active:
To determine the activity status of each user, the activity periods during the connection are taken into account.
f
UL
UL
DL
D Activity Session
D Activity Session
DL
= --------------------------------------------------------------------------------------- and f = --------------------------------------------------------------------------------------UL
UL
DL
DL
D Inactivity Session + D Activity Session
D Inactivity Session + D Activity Session
Therefore, we have:
a. 1st case: At a given time, packets are downloaded and uploaded.
DL
UL + DL
UL
UL + DL
UL
UL
1 f p Connected
DL
DL
1 f p Connected
DL
p Connected
The probability of the user being active on UL and inactive on DL: p1 Active = f
The probability of the user being active on DL and inactive on UL: p1 Active = f
UL + DL
The probability of the user being active on both UL and DL: p1 Active = f
UL
UL + DL
UL
DL
UL + DL
The probability of the user being inactive on both UL and DL: p1 Inactive = 1 f 1 f p Connected
b. 2nd case: At a given time, packet are uploaded only.
UL
The probability of the user being active on UL and inactive on DL: p2 Active = f
UL
UL
p Connected
UL
UL
The probability of the user being inactive on both UL and DL: p2 Inactive = 1 f p Connected
c. 3rd case: At a given time, packet are downloaded only.
DL
The probability of the user being active on DL and inactive on UL: p1 Active = f
DL
DL
p Connected
DL
DL
The probability of the user being inactive on both UL and DL: p3 Inactive = 1 f p Connected
Number of inactive users on UL and DL: n j Inactive = n j p1 Inactive + p2 Inactive + p3 Inactive
UL
UL
UL
DL
DL
DL
Number of users active on UL and inactive on DL: n j Active = n j p1 Active + p2 Active

Number of users active on DL and inactive on UL: n j Active = n j p1 Active + p3 Active
UL + DL
UL + DL
Number of users active on UL and DL: n j Active = n j p1 Active

Therefore, a connected user can be active on both links, inactive on both links, active on UL only, or active on DL only.
615

The user distribution per service, and the activity status distribution between the users
are average distributions. The service and the activity status of each user are random in
each simulation. Therefore, if you compute several simulations at once, the average
number of users per service and average numbers of inactive, active on UL, active on DL,
and active on UL and DL users, will correspond to calculated distributions. But, if you
compare each simulation, you will observe that the user distribution between services as
well as the activity status distribution between users is different in each simulation.
9.3.1.2 Simulations Based on Sector Traffic Maps

Sector traffic maps are also referred to as live traffic maps. Live traffic data from the OMC is spread over the best server
coverage areas of the transmitters included in the traffic map. Throughput demands per service, the numbers of active users
per service, or Erlangs per service are assigned to the coverage areas of each transmitter.
9.3.1.2.1
Throughputs in Uplink and Downlink

When selecting Throughputs in Uplink and Downlink, you can input the throughput demands in the uplink and downlink for
each sector and for each listed service.
Atoll calculates the number of users active in uplink and in downlink in the Txi cell using the service (NUL and NDL) as follows:
N
UL
DL
UL
DL
RS
RS
DL
= ----------- and N = ----------- for R99 circuit and packet switched services
UL
DL
R Nom
R Nom
DL
RS
= ---------for HSDPA service
DL
R Avg
UL
DL
R S and R S are the uplink and downlink throughputs for service S in the TXi cell from the traffic map.
NUL and NDL values include:
UL
Users active in uplink and inactive in downlink ( n i Active ),
Users active in downlink and inactive in uplink ( n i Active ),
And users active in both links ( n i Active ).
DL
UL + DL
UL
DL
Probability of being inactive in UL and DL: p Inactive = 1 f Act 1 f Act

UL
UL
DL
DL
DL
UL
Probability of being active in UL only: p Active = f Act 1 f Act

Probability of being active in DL only: p Active = f Act 1 f Act
UL + DL
UL
DL
Probability of being active both in UL and DL: p Active = f Act f Act

UL
DL
Where, f Act and f Act are respectively the UL and DL activity factors defined for the service i.
We have:
UL
UL + DL
UL
DL
UL + DL
DL
UL + DL
UL
DL
UL + DL
p Active + p Active n i Active + n i Active + n i Active = N UL

p Active + p Active n i Active + n i Active + n i Active = N DL
Therefore, we have:
UL + DL
UL + DL
N UL p Active
N DL p Active
UL + DL
--------------------------------------Number of users active in UL and DL both: n i Active = min --------------------------------------
UL
UL + DL
DL
+ DL
Active
UL
UL + DL
Number of users active in UL and inactive in DL: n i Active = N UL n i Active
616

AT332_TRR_E0
UL + DL
DL
Number of users active in DL and inactive in UL: n i Active = N DL n i Active

UL
UL + DL
DL
n i Active + n i Active + n i Active

Number of inactive users in UL and DL: n i Inactive = ------------------------------------------------------------------------------- p inactive
1 p inactive
9.3.1.2.2
Total Number of Users (All Activity Statuses)

When selecting Total Number of Users (All Activity Statuses), you can input the number of connected users for each sector
and for each listed service ( n i ).
UL
DL
Probability of being inactive in UL and DL: p Inactive = 1 f Act 1 f Act

UL
UL
DL
DL
DL
UL
Probability of being active in UL only: p Active = f Act 1 f Act

Probability of being active in DL only: p Active = f Act 1 f Act
UL + DL
UL
DL
Probability of being active both in UL and DL: p Active = f Act f Act

UL
DL
Where, f Act and f Act are respectively the UL and DL activity factors defined for the service i.
Number of inactive users in UL and DL: n i Inactive = n i p Inactive
UL
UL
DL
DL
Number of users active in UL and inactive in DL: n i Active = n i p Active

Number of users active in DL and inactive in UL: n i Active = n i p Active
UL + DL
UL + DL
Number of users active in UL and DL both: n i Active = n i p Active
9.3.1.2.3
Number of Users per Activity Status

When selecting Number of Users per Activity Status, you can directly input the number of inactive users ( n i Inactive ), the
UL
DL
UL + DL
number of users active in the uplink ( n i Active ), in the downlink ( n i Active ) and in the uplink and downlink ( n i Active ), for
each sector and for each service.
The activity status of users is based on an average distribution. The activity status of each
user is random in each simulation. Therefore, if you compute several simulations at once,
the average numbers of inactive, active on UL, active on DL, and active on UL and DL
users, will correspond to calculated distributions. But, if you compare each simulation,
you will observe that the activity status distribution between users is different in each
simulation.
9.3.2 Power Control Simulation

Based on CDMA air interface, a TD-SCDMA network automatically regulates itself by using uplink and downlink power control
in order to minimise interference and maximise capacity. For each user distribution, Atoll simulates these network regulation
mechanisms using an iterative algorithm and calculates network parameters such as traffic power per cell and per timeslot,
mobile terminal power, and handoff status for each terminal.
In each iteration, all the mobiles (R99 and HSDPA service users) selected during generation of the user distribution attempt
to connect to the network one by one. The process is repeated from iteration to iteration and ends when the network is
balanced, i.e., when the convergence criteria on uplink and downlink are satisfied.
The simulation algorithm also models the impact of smart antennas in the power control loop. The influence of smart
antennas is taken into account in signal quality calculations. Smart antennas improve the signal quality of each served mobile,
617

decrease the required powers and the loads of all the surrounding cells. Interference on the downlink and the uplink is
calculated on a per user. Power control is simulated over a sub-frame, i.e., 7 timeslots.
For HSDPA users, uplink and downlink power control is performed on the associated A-DCH bearer before fast link adaptation
on downlink. The steps of this algorithm are detailed below.
Figure 9.2: TD-SCDMA Power Control Algorithm
9.3.2.1 Algorithm Initialisation

At the start of each simulation, the system loads for each carrier and timeslot are reset to initial values:
Downlink traffic powers of cells P TCH DL are initialised to 0 Watts

Uplink interference powers received on all the carriers and timeslots I Intra UL and I Extra UL are initialised to 0 Watts
(i.e., no connected mobiles)
Term
Uplink required power for mobiles is set to P Min
9.3.2.2 R99 Part of the Algorithm

Req
The algorithm is described for an iteration k. Here, Xk is the value of the variable X at the iteration k. In the algorithm, all Q UL
Req
and Q DL thresholds depend on the user mobility, and are defined in the Service and Mobility parameter tables. All the
variables used in the description below are listed in "Definitions and Formulas" on page 599.
The following calculations are made for all R99 and HSDPA mobiles (Mi) using R99 bearers.
9.3.2.2.1
Determination of Mis Best Server (SBS(Mi))

This step is performed for TS0 for each station TXi containing Mi in its calculation area.
The best server for Mi is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from
the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the P-CCPCH RSCP is calculated for:
618
the preferred carrier of the service used by Mi, or

AT332_TRR_E0
The RSCP from a transmitter TXi and a selected carrier ic is given by:
TX ic
i
TX ic
i
RSCP P CCPCH = P P CCPCH + G
TX
TX
Model
L Path M Shadowing L Body L Indoor + G
in dBm
Where,
TX i

TX i

Model
is selected
L
Mi
is the los of the terminal used by Mi
Mi
L Body is the body loss defined in the service used by Mi

G
G
L
Mi
TX i
TX i
is the receiver gain of the terminal user by Mi

TX i
= L Total DL )
A cell TX i ic is considered the best server of a mobile Mi if it satisfies the following conditions:
TX i ic
Min
RSCP P CCPCH RSCP P CCPCH ,

TX i ic
RSCP P CCPCH TAdd P CCPCH Mobility ,

TX ic
i
And RSCP P CCPCH =
Best RSCP TXj jc .

P CCPCH
j = All
The best server is determined once for the whole simulation during the first iteration, i.e., k = 0, because the best server does
not change during the simulation and smart antennas do not influence this step.
Mi is considered unable to connect to the network if no best server has been selected. In this case, Mi is rejected for the reason
P-CCPCH RSCP < Min P-CCPCH RSCP. If Mi has no best server, it is not taken into account in the next steps.
9.3.2.2.2
Dynamic Channel Allocation

The dynamic channel allocation is performed once for the whole simulation during the first iteration, i.e., k = 0. The DCA
controls the mobile admission. Once a mobile has been admitted for a simulation, it remains admitted for the all the iterations
unless there are other reasons to reject it (following steps).
The aim of Dynamic Channel Allocation (DCA) is to reduce interference in order to maximise the usage of the radio resources.
In other words, the DCA tries to find the "best carrier" and the "best timeslots", which when allocated to the mobiles will
optimise the load balance between carriers.
If a preferred carrier is defined for the service requested by Mi and if it is available at TX i . BestCarrier TX i M i =
the
carrier preferred for the service. In the case of N-frequency compatible transmitters, Mi can be allocated timeslots over more
than one slave carrier.
Mi is considered unable to connect to the network if no carrier or not enough timeslots have been selected. In this case, the
mobile Mi will be rejected for the reason "RU Saturation". If the carrier and timeslot(s) selected by the DCA do not satisfy the
control of radio resource limits for DL power or UL load, then the mobile will be rejected for the reason "DL Load Saturation"
or "Admission Rejection" respectively.
619

There are four strategies for the DCA available in Atoll. These strategies are described below one by one.
1. Load
Carrier Selection by Load: The DCA determines the least loaded carrier with enough timeslots to accomodate the
service being used by each mobile Mi. The best carrier for a mobile is the one that is least loaded:
BestCarrier TX i M i = Carrier
Where, X
DCA
DCA
= X DL
Min X
DCA
TX i ic TS M i
= N Tot DL
if the mobile is connected in the downlink.
TX i ic TS M i
And, X
DCA
DCA
DCA
DCA
X UL
N Tot UL
- X DCA if the mobile is connected in the uplink.
= ----------------------------------------------TX i ic TS M i
TX i
N Tot UL
+ N0
is the load increment given by:

Mi
TX i
1 f UL 1 f JD
= ---------------------------------------------------------------------1
1 + ----------Req
Q UL
Ortho
E b
--- N t UL
Proc
= ------------------ is the uplink required signal quality. The uplink processing gain G UL calculated
Proc
G UL
Req
Req
C
Req
Where Q UL = ---
I UL
from the service parameters, if no smart antenna is used by the transmitter in the uplink.
If a smart antenna is used by the transmitter in the uplink, the smart antenna gain is taken into account in calculating
Req
Q UL .
TX i ic TS M i
N Tot UL
TX i ic TS M i
is described in "Uplink Power Control" on page 621.
N Tot DL
is described in "Downlink Power Control" on page 623.
The carrier is the same in the uplink and in the downlink for mobiles accessing circuitswitched services.
Timeslot selection by Load: From the selected carrier, Atoll selects the timeslots which are the least loaded and have
enough resource units for the service being accessed by Mi.
2. Available RUs
Carrier selection by Available RUs: The DCA determines the carrier which has the highest number of available
resource units with enough timeslots to accomodate the service being used by each mobile Mi. The best carrier for a
mobile is the one that has the highest number of resource units:
Max RUs
Timeslot selection by Available RUs: From the selected carrier, Atoll selects the timeslots which have the highest
numbers of available resource units.
3. Direction of Arrival
Carrier selection by Direction of Arrival: The DCA determines the direction of arrival of the signal from the served user
Mi and checks whether there is an interfering mobile in the same direction as Mi. Atoll searches for interfering mobiles
within the angle defined by the Angular Step. For example, if you enter an angular step of 15 degrees, Atoll searches
for interfering mobiles within 15 degrees to the right and to the left of the served user, and allocates a different carrier
than the ones used by any interfering mobiles found. The best carrier for a mobile is the one which is not interfered
by another mobile in the direction of the mobile Mi.
DoA Mi DoA Mj
In other words, the direction of arrival for the served user Mi should not be the direction of arrival of an interfering
mobile.
620

AT332_TRR_E0
Timeslot selection by Direction of Arrival: From the selected carrier, Atoll selects the timeslots which are not being
used by any other mobile Mj located in the same direction as the served user Mi.
4. Sequential
Sequential carrier selection: The DCA allocates carriers to served users Mi in a sequential order.
Sequential timeslot selection: From the selected carrier, Atoll allocates timeslots to served users Mi in a sequential
order.
At the end of the DCA, each admitted mobile has an associated carrier and timeslots. In case of N-frequency mode compatible
transmitters, an admitted mobile can have associated timeslots over more than one slave carrier.
9.3.2.2.3

For each mobile Mi, the uplink power control step calculates the uplink power required to satisfy the required quality level on
the traffic channel, which is defined for the service being accessed by Mi.
If the mobile Mi is connected (active or inactive) in the uplink and has a best server TX i ic assigned to it, Atoll calculates the
signal quality on the uplink timeslots allocated to Mi by the DCA:
TX i ic TS M i
E b
--- N t TCH UL
TX i ic TS M i
Mi
Mi
TX i ic TS M i
TX ic TS M
i
i
RSCP TCH UL
RSCP TCH UL
Div
C
- G Proc
-- G Div
or
= ------------------------------------------------------
G
= ------------------------------------------------------UL
UL
UL
TX i ic TS M i
TX i ic TS M i
I TCH UL
N Tot UL
N Tot UL
Calculation of Uplink Total Noise ( N Tot UL ):

The uplink total noise is calculated for the uplink connection between each mobile Mi and its best server TX i ic .
TX i ic TS M i
N Tot UL
TX i ic TS M i
= I Tot UL
TX i
+ N0
Where
TX i ic TS M i
I Tot UL
Mi
RSCP TCH UL TX i ic TS M i
Mi
RSCP TCH UL TX i ic TS M i
M j TX i ic TS M i
M M
j
i
1 j RSCP j
TCH UL TX i ic TS M i +
M j TX i ic TS M i
Mj Mi
TX
j
i
RSCP TCH UL TX i ic TS M i 1 F MCJD
M j TX i ic TS M i
Mi
Mi
Ortho
1 F UL
TX i
1 F JD and = 0

Total Noise
The above formula gives the value of I Tot UL for the uplink connection between Mi and TX i ic , taking into account the
interference received from other mobiles, Mj, which are located in the Mi best server coverage area, as well as located in the
coverage areas of other cells. The mobile Mi is the focus, i.e., the mobile that is listened to by the transmitter TX i ic .
The four terms comprising I Tot UL are:
The useful signal for which the received mobile is the focus (Mi).
The intra-cell interference for which the best-server is the same for the received mobile Mj and the focus Mi, TX i ic .
The intra-cell interference due to distortion in the terminal transmission.
The extra-cell interference for which the best-server for the received mobile Mj is not TX i ic .
621


M
i
RSCP TCH UL TX i ic
The uplink received signal code power is:

TX
Model
LT
P Req TX i ic TS M i
k1
TS M i = --------------------------------------------------------------Model
LT
Model
M
L Path L L L Body L Indoor M Shadowing
i
= ------------------------------------------------------------------------------------------------------------------ and P Req TX i ic TS M i is the uplink required mobile power
TX
M
i
i
G G
Mi
Mi
calculated for the timeslot allocated to Mi. If Mi is an HSDPA user, P Req TX i ic TS M i = 0,1 P Req TX i ic TS M i
Model
In L T
, G
TX
SA
= G UL and L
TX
SA
= L UL are calculated according to the smart antenna modelling method used, for
Mi
P Req TX i ic TS M i , if a smart antenna is available in the uplink. Otherwise, G
TX i
and L
TX i
are read from the main antenna
model.
Interference is updated only for active mobiles on the uplink for circuit- and packetswitched services. However, if these mobiles are rejected, they are considered in the
number of rejected mobiles.
M
Calculation of Uplink Required Power ( P Req ):

Then Atoll determines the required uplink power by:
E b
--- N t TCH UL
Mi
= P Req TX i ic TS M i
-------------------------------------k1
E b TXi ic TS M i
--- N t TCH UL
Req
Mi
Req
Mi
or P Req TX i ic TS M i
C
---
I TCH UL
= P Req TX i ic TS M i
-----------------------------------TX i ic TS M i
k1
C
-- I TCH UL
Mi
Mi
Mi
Mi
Mi
And if P Req TX i ic TS M i P Min then P Req TX i ic TS M i = P Min

Mi
Mi
If P Req TX i ic TS M i P Max
M
then the mobile Mi is rejected for the reason "Pmob > PmobMax", and
P Req TX i ic TS M i is set to 0.
M
P Min and P Max are set in the properties of the terminal used by the mobile Mi.
TX i ic TS M i
Where RSCP TCH UL
TX i
Mi
k1
= --------------------------------------------------------------Model
LT
Mi
Mi
Model
Mi
= ------------------------------------------------------------------------------------------------------------------ and P Req TX i ic TS M i
TX i
Mi
G G
power for iteration k - 1 transmitted on the timeslot allocated to Mi.
Model
LT
Model
In L T
M
, G
TX i
SA
= G UL and L
TX i
is the uplink required mobile
k1
SA
= L UL are calculated according to the smart antenna modelling method used, for
P Req TX i ic TS M i , if a smart antenna is available in the uplink. Otherwise, G
TX
and L
TX
model.
The uplink required powers for mobiles inactive in the uplink accessing circuit- or packetswitched services are calculated for information only. However, if these mobiles are
rejected, they are considered in the number of rejected mobiles.
622

AT332_TRR_E0
9.3.2.2.4

For each mobile Mi, the downlink power control step calculates the downlink power for the best server TX i ic required to
satisfy the required quality level on the traffic channel, which is defined for the service being accessed by Mi.
If the mobile Mi is connected (active or inactive) in the downlink and has a best server TX i ic assigned to it, Atoll calculates
the signal quality on the uplink timeslots allocated to Mi by the DCA:
TX
TX i ic TS M i
TX ic TS M
i
i
TX
TX ic TS M
i
i
TX i ic TS M i
RSCP TCH DL
RSCP TCH DL
Div
C
- G Proc
- G Div
= -------------------------------------------------------= -------------------------------------------------------DL G DL or ---
DL
TX ic TS M
TX ic TS M
I TCH DL
i
i
i
i
N Tot DL
N Tot DL
E b
--- N t TCH DL
Calculation of Downlink Total Noise ( N Tot DL ):

The downlink total noise is calculated for the downlink connection between each mobile Mi and its best server TX i ic .
TX i ic TS M i
N Tot DL
TX i ic TS M i
= I Tot DL
Mi
+ I IC DL ic jc + I MM M i M j + N 0
Where
TX i ic TS M i
I Tot DL
TX ic TS M
i
i
RSCP Tot DL
Mi
TX
TX i ic TS M i
RSCP Tot DL
Mj
TX i
M j TX i ic TS M i
Mj Mi
TX ic TS M i
TX
1 i RSCP i
Tot DL
RSCP Tot DL
Mj +
M TX ic TS M
j
i
i
Mj Mi
TX ic TS M
j
i
Mj
M j TX i ic TS M i
The four terms comprising I Tot DL are:
The useful signal for which the received mobile is the focus (Mi).
The intra-cell interference for which the best-server is the same for the received mobile Mj and the focus Mi, TX i ic .
The intra-cell interference due to distortion in the transmitter.
The extra-cell interference for which the best-server for the received mobile Mj is not TX i ic .
TX j jc TS M i
RSCPTot DL
Mi
All TX
j
I IC DL ic jc = ---------------------------------------------------------------F IRF ic jc
TX i
TX i
Ortho
1 F DL
Mi
1 F JD and = 0
1

Total Noise
I IC DL ic jc is the inter-carrier interference from a carrier jc to another carrier ic on the downlink, which is reduced by the
interference reduction factor F IRF ic jc defined for the pair (ic, jc).
TX i ic TS M i
RSCP Tot DL
TX i ic TS M i
With RSCP TCH DL
TX i ic TS M i
= RSCP TCH DL
TX i ic TS M i
+ RSCP OCCH
TX ic TS M
i
i
TX ic TS M
i
i
P TCH DL
TX i ic TS M i
P OCCH
k1
= ----------------------------------------- and RSCP OCCH
= -----------------------------Model
Model
LT
LT
623

TX
Model
LT
Model
TX ic TS M
i
i
= ------------------------------------------------------------------------------------------------------------------ and P TCH DL
TX
M
i
i
G G
is the downlink traffic power transmitted
k1
TX ic TS M
i
i
on the timeslot allocated to Mi during the iteration k - 1. If Mi is an HSDPA user, P TCH DL

Model
In L T
, G
TX i ic TS M i
P TCH DL
TX i
SA
= G DL and L
TX i
TX ic TS M
i
i
= 0,1 P TCH DL
SA
= L DL are calculated according to the smart antenna modelling method used, for
TX i ic TS M i
only and not for P OCCH
, if a smart antenna is available in the downlink. Otherwise, G
TX i
and L
TX i
are
read from the main antenna model.

Mj
RSCPTCH UL Mi
M
j
I MM M i M j = ------------------------------------------------is the interference from each mobile Mj transmitting in the uplink on the same
F IRF ic jc
timeslots as those on which the mobile Mi is receiving in the downlink. Mj can interfere Mi directly if and only if:
Mi Mj
The distance between Mi and Mj ( d

) is less than the Max Distance between interfering mobiles defined by the
user when starting the simulation, and
The downlink timeslot of Mi (TSMi) is the same as the uplink timeslot of Mj, (TSMj).
The interference received from the mobile Mj at the mobile Mi is calculated using either the free-space propagation model or
the Xia model.
Mj
M
P TCH UL
j
RSCP TCH UL M i = ------------------L MM
32,4 + 20 Log F Avg + 20 Log d

L MM =
49 + 30 Log F Avg + 40 Log d
If d
If d
Mi Mj
Mi Mj
3m
with F Avg being the average frequency in MHz of the
3m
frequency band used by the best server of the mobile Mi, and d is the distance between the mobiles Mi and Mj in km.
TX i ic TS M i
Calculation of Downlink Required Power ( P Req
):
Then Atoll determines the required downlink power by:

E b
--- N t TCH DL
-------------------------------------k1
E b TXi ic TS Mi
--- N t TCH DL
Req
TX i ic TS M i
P Req
TX i ic TS M i
= P Req
Req
TX i ic TS M i
or P Req
TX i ic TS M i
= P Req
TX i ic TS M i
And if P Req
TX i ic TS M i
If P Req
C
---
I TCH DL
-----------------------------------TX i ic TS M i
k1
C
---
I TCH DL
TX i ic TS M i
Min
P TCH DL Service then P Req
Min
= P TCH DL Service
TX i ic TS M i
Max
P TCH DL Service then the mobile Mi is rejected for the reason "Ptch > PtchMax", and P Req
is set
to 0.
Min
Max
P TCH DL Service and P TCH DL Service are set in the properties of the R99 bearer associated with the service used by the
mobile Mi.
TX i ic TS M i
Otherwise, the downlink traffic power is incremented P TCH DL
TX i ic TS M i
= P TCH DL
TX i ic TS M i
+ P Req
For each mobile, Atoll also calculates the downlink traffic power for the different values of the Angular Step Step .
TX i ic TS M i
RSCP TCH DL
624
TX i ic TS M i
Step
= RSCP TCH DL
SA
G DL
-----------------------SA
L DL Step

AT332_TRR_E0
Where
TX ic TS M
i
i
P Req
k1
= ----------------------------------------Model
LT
TX ic TS M
i
i
RSCP TCH DL
TX
Model
TX ic TS M
i
i
= ------------------------------------------------------------------------------------------------------------------ and P Req
TX
M
i
i
G G
- 1 transmitted on the timeslot allocated to Mi.
Model
LT
Model
In L T
, G
TX i ic TS M i
P Req
TX i
SA
= G DL and L
TX i
k1
is the downlink traffic power for iteration k
SA
, if a smart antenna is available in the downlink. Otherwise, G
TX i
and L
TX i
model.
The downlink power for mobiles inactive in the downlink accessing circuit- or packetswitched services are calculated for information only.
9.3.2.2.5
Uplink Signals Update

This step uses the uplink terminal powers calculated for each timeslot allocated to the mobiles. The Dynamic Channel
Allocation allocates timeslots and carriers to all the connected and active mobiles. The Dynamic Channel Allocation is
performed once only, during the first iteration, and the timeslot and carrier allocation remains the same for all the following
iterations of a simulation.
This step updates the received signals for all the mobiles Mi interfered in the uplink by the uplink connection between
interfering mobiles Mj and their best servers TX j ic , only if TX j ic contain Mi in their coverage areas. TX i ic is the
interfered receiver and Mi is the focus, i.e., the mobile that is listened to by the transmitter TX i ic .
TX i ic TS M i
For each mobile Mi interfered by Mj in the uplink by the connection between Mj and TX j ic , Atoll updates RSCP TCH UL
9.3.2.2.6
Downlink Signals Update

For the first iteration, i.e., k = 0, the downlink traffic powers for all the downlink timeslots are set to 0 Watts. Therefore, for
the first iteration, this step is performed for any downlink timeslot for each mobile Mi that is connected and active.
However, for the following iterations, the downlink signals update step uses the actual downlink traffic powers calculated for
each timeslot and the actual timeslots allocated to the mobiles. The Dynamic Channel Allocation allocates timeslots and
carriers to all the connected and active mobiles. The Dynamic Channel Allocation is performed once only during the first
iteration and the timeslot and carrier allocation remains the same for all the following iterations of a simulation.
Therefore, this step is performed for any downlink timeslot for each mobile Mi that is connected and active for the first
iteration, and this step is performed for all the downlink timeslots allocated to the mobile Mi on which it is connected and
active, for the following iterations since the DCA has been performed.
This step updates the received signals for all the mobiles in the TX i ic coverage area which are interfered in the downlink
by the connection between TX i ic and Mi.
TX ic TS M
i
i
For each mobile interfered by Mi, Atoll updates RSCP TCH DL
Where TX i ic is the transmitter considered and Mi is the focus, i.e., the mobile that is the target for TX i ic .
9.3.2.2.7
Control of Radio Resource Limits (Downlink Traffic Power and Uplink Load)
This step checks whether the downlink traffic powers of the downlink timeslots and the uplink loads of the uplink timeslots
of all the cells satisfy the conditions defined globally or per cell and timeslot.
Downlink Power Control:
Atoll verifies that the total R99 power transmitted by any cell on any timeslot does not exceed the effective maximum cell
power per timeslot. The effective maximum cell traffic power per timeslot is calculated as:
TX i ic TS M i
TX i ic TS M i
P Max DL Eff = P Max DL
%P Max DL
625

TX ic TS M
i
i
Where P Max DL
is the maximum cell power per timeslot defined per cell, and %P Max DL is the maximum allowed
downlink load either taken from the properties of each cell or from the simulation properties if a global value is defined.
For each transmitter TXi, carrier ic, and downlink timeslot TS M ,
i
TX i ic TS M i
P R99 DL
TX i ic TS M i
= P TCH DL
TX i ic TS M i
If P R99 DL
TX i ic TS M i
+ P OCCH
TX i ic TS M i
P Max DL Eff the mobile with the lowest service priority is rejected for the reason "DL Load Saturation".
Uplink Load Control:

Atoll verifies that the uplink load of any cell on any timeslot does not exceed the maximum uplink cell load allowed per
timeslot.
TX i ic TS M i
The maximum allowed uplink cell load, X Max UL
, is either taken from the properties of each cell or from the simulation
properties if a global value is defined.

For each transmitter TXi, carrier ic, and uplink timeslot TS M ,
i
TX i ic TS M i
If X UL
TX i ic TS M i
X Max UL
the mobile with the lowest service priority is rejected for the reason "UL Load Saturation".
The uplink load is given by:

TX i ic TS M i
X UL
TX i ic TS M i
N Tot UL
= ----------------------------------------------- if no smart antenna is used by the transmitter in the uplink.
TX ic TS M
TX
i
i
i
N Tot UL
+ N0
If a smart antenna is used by the transmitter in the uplink, the smart antenna gain is taken into account in the calculation of
uplink load.
9.3.2.3 HSDPA Part of the Algorithm

The following calculations are made for all HSDPA mobiles (Mi).
9.3.2.3.1
HSDPA Power Allocation

TX ic
i
The total transmitted power of the cell ( P Tot DL ) is the sum of the R99 transmitted power and the HSDPA powers.
TX i ic
TX i ic
TX i ic
P Tot DL = P R99 DL + P HR
TX i ic
TX i ic
+ P HS SCCH + P HS PDSCH
The HSDPA powers, i.e., the HS-SCCH and HS-PDSCH powers are calculated as follows:
HS-SCCH Power:
HS-SCCH channels are transmitted on DL traffic timeslots. The maximum number of supported HS-SCCH channels is
defined per cell. Power can be allocated to HS-SCCH statically or dynamically:
Static Allocation
The static HS-SCCH power is defined in the properties of the HSDPA cell.
Dynamic Allocation
TX ic
i
TX ic
TX ic
Ec
Req
i
i
= Q HS SCCH Mobility so that P HS SCCH P Available HS SCCH .
HS-SCCH power is calculated for -----
N t HS SCCH
TX i ic
TX i ic
TX i ic
TX i ic
Where P Available HS SCCH = P Max DL Eff P R99 DL P HR

TX i ic
TX i ic
is the power available for HS-SCCH in the cell
TX i ic
TX i ic , and P R99 DL = P TCH DL + P OCCH .

TX i ic
TX i ic
The effective maximum cell traffic power per timeslot is calculated as: P Max DL Eff = P Max DL %P Max DL .
TX i ic
P Max DL is the maximum power defined per cell, and %P Max DL is the maximum allowed downlink load either
taken from the properties of each cell or from the simulation properties if a global value is defined.
626

AT332_TRR_E0
TX i ic
TX ic
TX ic
Ec
i
i
i
---
N t HS SCCH N Tot DL RSCP HS SCCH

Model
= ---------------------------------------------------------------------------------------------------------------------------- LT
TX
i
TX ic
i
P HS SCCH
TX
TX ic
i
Where N Tot DL is the downlink total noise calculated in "Downlink Power Control" on page 623,
TX i
TX i
Ortho
1 F DL
TX i
Mi
1 F JD and = 0
1
Mi
Mi

Total Noise
Model
TX i ic
= ------------------------------------------------------------------------------------------------------------------ and P HS SCCH is the HS-SCCH power calculated for the
TX i
Mi
G G
timeslots allocated to Mi.
Model
LT
Model
In L T
,G
TX i
SA
= G DL and L
TX i
SA
TX ic
i
P HS SCCH , if a smart antenna is available in the downlink. Otherwise, G
TX
and L
TX
are read from the main
antenna model.
HS-PDSCH Power:
HS-PDSCH channels are transmitted on DL traffic timeslots. Power can be allocated to HS-PDSCH statically or
dynamically:
Static Allocation
The static HS-PDSCH power is defined in the properties of the HSDPA cell.
Dynamic Allocation
HS-PDSCH power is calculated as follows:
TX i ic
TX i ic
TX i ic
TX i ic
P HS PDSCH = P Max DL Eff P R99 DL P HR

TX ic
i
TX ic
i
TX i ic
P HS SCCH
TX ic
i
Where P R99 DL = P TCH DL + P OCCH . The effective maximum cell traffic power per timeslot is calculated as:
TX i ic
TX i ic
TX i ic
P Max DL Eff = P Max DL %P Max DL . P Max DL is the maximum power defined per cell, and %P Max DL is the
maximum allowed downlink load either taken from the properties of each cell or from the simulation properties
if a global value is defined.
The HS-SICH power is calculated as follows:
HS-SICH Power:
HS-SICH channels can be transmitted on any UL traffic timeslot. The maximum number of supported HS-SICH channels
is defined per cell. Power can be allocated to HS-SICH statically or dynamically:
Static Allocation
The static HS-SICH power is defined in the properties of the terminal used by the HSDPA mobile Mi.
Dynamic Allocation
Mi
TX i ic
E c TXi ic
Req
= Q HS SICH Mobility so that P HS SICH P Max HS SICH and
HS-SICH power is calculated for -----
N t HS SICH
M
P HS SICH P Max HS SICH .

TX i ic
Mi
P HS SICH
Where
Mi
TX ic
Ec
i
i
i
--- N
RSCP HS SICH
N t HS SICH Tot UL
- L Model
= ------------------------------------------------------------------------------------------------------------------------T
Mi
TX i ic
N Tot UL
Mi
is the uplink total noise calculated in "Uplink Power Control" on page 621,
Ortho
1 F UL
TX i
1 F JD and = 0

Total Noise
627

TX
Model
M
i
= ------------------------------------------------------------------------------------------------------------------ and P HS SICH is the HS-SICH power calculated for the
TX
M
i
i
G G
timeslots allocated to Mi.
Model
LT
Model
In L T
,G
TX
SA
= G UL and L
TX
SA
= L UL are calculated according to the smart antenna modelling method used,
Mi
for P HS SICH , if a smart antenna is available in the uplink. Otherwise, G
TX i
and L
TX i
are read from the main
antenna model.
TX i ic
The total transmitted power of the cell ( P Tot DL ) is the sum of the R99 transmitted power and the HSDPA powers.
TX i ic
TX i ic
TX i ic
P Tot DL = P R99 DL + P HR
9.3.2.3.2
TX i ic
TX i ic
+ P HS SCCH + P HS PDSCH
Connection Status and Number of HSDPA Users

HSDPA users cannot receive HS-SCCH and HS-PDSCH powers simultaneously. HS-PDSCH arrives 3 timeslots after the HS-SCCH.
HS-SICH is 9 timeslots after the HS-PDSCH. Atoll assumes that an active HSDPA user has the same probability of receiving HSSCCH and HS-PDSCH, and transmitting HS-SICH because their occurrence is equally likely. Therefore, each HSDPA user is
assigned a sub-connection status randomly. The sub-connection status can be:
HS-SCCH: HSDPA mobile that is receiving HS-SCCH power

HS-PDSCH: HSDPA mobile that is receiving traffic power
HS-SICH: HSDPA mobile that is transmitting HS-SICH power
The number of active HSDPA users belonging to each sub-connection status is 1/3rd of the total number of active HSDPA
users.
n HS SCCH is the maximum number of HS-SCCH channels and n HS SICH is the maximum number of HS-SICH channels that the
cell can manage. Each HSDPA user consumes one HS-SCCH and HS-SICH channels. Therefore, at a given instance, the number
of connected HSDPA users cannot exceed the number of HS-SCCH and HS-SICH channels per cell. The maximum number of
HSDPA users ( n Max ) corresponds to the maximum number of HSDPA users that the cell can support.
9.3.2.3.3
HSDPA Admission Control

HS-SCCH
HS-SCCH admission control is performed for active HSDPA users connected to A-DCH bearers on the downlink and having an
HS-SCCH sub-connection status. Each cell is able to manage a maximum number of HS-SCCH channels, n HS SCCH . During the
R99 part, the DCA provides a DL timeslot with one SF16 resource unit that has the downlink Ec/Nt higher than the required
quality. If no cell with such a resource unit is available, the user is rejected.
HS-SICH
HS-SICH admission control is performed for active HSDPA users connected to A-DCH bearers on the uplink and having an HSSICH sub-connection status. Each cell is able to manage a maximum number of HS-SICH channels, n HS SICH . During the R99
part, the DCA provides an UL timeslot with one SF16 resource unit that has the uplink Ec/Nt higher than the required quality.
If no cell with such a resource unit is available, the user is rejected.
HS-PDSCH
Scheduling is performed for active HSDPA users connected to A-DCH bearers on the downlink and having an HS-PDSCH subconnection status. The scheduling is performed as follows:
1. Each HS-PDSCH user is considered as the only served user. The scheduler allocates the best available HSDPA bearer to
each user. The best available HSDPA bearer is selected depending on the users Ec/Nt. If no bearer can be allocated
due to low Ec/Nt, the user is rejected for the reason "HSDPA Scheduler Saturation".
The required HS-PDSCH Ec/Nt value is read from receiver equipment properties. For each bearer, Atoll checks that the
Ec/Nt reaches the quality target. HS-PDSCH Ec/Nt is calculated by taking into account all intra and extra cells
interferences.
2. The scheduler sorts the HS-PDSCH users to whom bearers have been assigned in the order of decreasing RLC peak
throughputs. If two users have the same bearer, the user with the higher Ec/Nt has the higher rank.
3. The scheduler considers the group of HS-PDSCH users to whom bearers, HS-SCCH, and HS-SICH have been assigned.
The number of HS-PDSCH users cannot exceed the maximum number of HSDPA users ( n Max ) supported by the cell. If
628

AT332_TRR_E0
there are enough HSDPA power and resource units available in order to obtain a HSDPA bearer, the users will be
connected. Otherwise, they will be delayed and their connection status will be HSDPA Delayed.
4. Other HS-PDSCH users will be rejected for the reason "HSDPA Scheduler Saturation".
For N-frequency mode compatible transmitters, the resource units available in the master and slave carriers can be shared,
i.e., a mobile can be connected to timeslots belonging more than one carrier.
9.3.2.3.4
HSDPA Dynamic Channel Allocation

For each mobile connected to the A-DPCH bearer:
1. Atoll selects the HSDPA bearers that match to the mobile terminal and UE category parameters.
2. For each bearer supported by a mobile:
a. The scheduler searches for the best collection of "n" ordered timeslots that can provide enough resource units to
support the service, and whose Ec/Nt is better than the minimum required and enough to reach the bearers
resource unit requirements. The best is determined by applying the R99 Dynamic Channel Allocation algorithm.
b. The scheduler calculates the HS-PDSCH Ec/Nt for each timeslot of the best collection. The Ec/Nt value associated
with the mobile-bearer pair is the worst one of all selected timeslots.
c. If the scheduler is unable to find a satisfactory timeslot collection, the bearer is removed from the list of supported
bearers.
3. The mobile is connected to the supported bearer having the highest RLC peak throughput. If two bearers have the
same RLC peak throughput, the best one is the one with the highest Ec/Nt.
9.3.2.3.5
Ressource Unit Saturation

For each time slot, a minimum and maximum number of resource units for HSDPA users are defined in the cell properties.
Atoll dynamically allocates the required number of codes respecting these limitations. The minimum number of HSDPA codes
is excluded from the set of codes available for R99 users. The scheduler checks if enough codes are available for the selected
HSDPA bearer (taking into account the maximum number of HSDPA codes). If not, the scheduler allocates a lower HSDPA
bearer which needs fewer codes. If there are no more resource units available for the lowest HSDPA bearer, the user will be
delayed or rejected.
9.3.2.4 Convergence Criteria

The convergence criteria are evaluated for each iteration and can be written as follows:
Max TXi ic TS M i
100
DL = Int
P Err
All
TX
UL
TX i ic TS M i
TX i ic TS M i
N Tot UL
Max N Tot UL
k
k 1 100
= Int
-----------------------------------------------------------------------------------TX i ic TS M i
All TX i
N Tot UL
k
TX i ic TS M i
Where, P Err
is given by:
TX i ic TS M i
TX ic TS M
i
i
P Err
TX i ic TS M i
P Rec
P Rec
Max
Step k
Step k 1
=
------------------------------------------------------------------------------------------------------------- with smart antennas.
TX i ic TS M i
0 Step 360
P Rec
Step k
TX i ic TS M i
P Err
TX i ic TS M i
TX i ic TS M i
P Rec
P Rec
k
k 1 without smart antennas.
= ---------------------------------------------------------------------------------TX i ic TS M i
P Rec
k
Atoll stops the simulations in the following cases:
Convergence: Between two successive iterations, DL and UL are less than or equal to their respective thresholds
(defined when creating a simulation).
629

Example: Let us assume that the maximum number of iterations is 100, and the UL and DL convergence thresholds
are set to 5 %. If DL 5 and UL 5 between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th
iteration. The simulation has converged.
Divergence: After 30 iterations, DL and/or UL are still higher than their respective thresholds and from the 30th
iteration, DL and/or UL do not decrease during the next 15 successive iterations.
Examples: Let us assume that the maximum number of iterations is 100, and the UL and DL convergence thresholds
are set to 5 %.
a. After the 30th iteration, DL and/or DL equal 100 and do not decrease during the next 15 successive iterations.
Atoll stops the algorithm at the 46th iteration. The simulation has not converged.
b. After the 30th iteration, DL and/or UL equal 80, they start decreasing slowly until the 40th iteration (without
going under the thresholds) and then, do not change during 15 successive iterations. Atoll stops the algorithm at
the 56th iteration without converging.
Last Iteration: If DL and/or UL are still much higher than their respective thresholds after the last iteration, the
simulation has not converged. If DL and UL are lower than their respective thresholds, the simulation has reached
convergence.
9.4 TD-SCDMA Prediction Studies

For each TBC transmitter, TXi, Atoll determines the value of the selected parameter on each studied pixel inside the TXi
calculation area. Each pixel within the TXi calculation area is considered a probe receiver.
The study conditions to determine the service area of each TBC transmitter
The display settings to for colouring the covered pixels
Atoll uses the parameters entered in the Condition tab of the coverage study properties dialog box to determine pixels
covered by the each transmitter. Coverage prediction display resolution is independent of the path loss matrix and geographic
data resolutions, and can be different for each coverage prediction. Coverage predictions are calculated using bilinear
interpolation of multi-resolution path loss matrices (similar to the evaluation of site altitudes).
9.4.1 P-CCPCH Reception Analysis (Eb/Nt) or (C/I)

E
C
These coverage predictions calculate and display the Eb/Nt or C/I on the P-CCPCH, ----b-
or ---
. The
N t P CCPCH
I P CCPCH
coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The best
servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH
power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage predictions
are calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels
covered by this transmitter. If you perform these coverage predictions for the best carrier, Atoll calculates the Eb/Nt or C/I
considering:
The

pixels
in
the
TX i ic
TX ic
i
coverage
area
where
TX i ic
Min
RSCP P CCPCH Max (TAdd P CCPCH,RSCP P CCPCH) ,
TX i ic
and
E b
C
Req
Req
--- Q P CCPCH or ---
Q P CCPCH are covered and coloured according to the selected display option.
N t P CCPCH
I P CCPCH
TX i
TX i ic
TX
TX ic
i
i
TX i ic
E b TXi ic
RSCP P CCPCH
RSCP P CCPCH
Proc
C
---
= -----------------------------------------------
=
-----------------------------------------------Where -----
and
G
P
CCPCH
TX i ic
N t P CCPCH
I P CCPCH
TX ic
i
N Tot DL
N Tot DL
TX ic
i
TX i ic
P P CCPCH
RSCP P CCPCH = ---------------------LT
The downlink total noise is calculated as follows:
630

AT332_TRR_E0
TX ic
i
TX ic
i
TX ic
i
Term
N Tot DL = I Intra DL + I Extra DL + I IC DL ic jc + N 0

Where
TX ic
i
TX ic
i
I Intra DL = RSCP P CCPCH

With
TX i
TX i ic
TX i
Ortho
1 F DL
TX
TX ic
i
+ RSCP OCCH TS0

Term
1 F JD
TX j ic
TX
and = 0
1

Total Noise
TX j ic
I Extra DL =
ji
TX j jc
TX j jc

TX
j
I IC DL ic jc = --------------------------------------------------------------------------------------F IRF ic jc
TX i ic
TX i ic
RSCP OCCH TS0
P OCCH TS0
= -----------------------LT
TX i
Eb Nt
Term
L Path L L
L Body L Indoor M Shadowing
L T = ----------------------------------------------------------------------------------------------------------------------TX i
Term
G G
TX i
Term
and N 0
are defined in "Definitions and Formulas" on page 599.
Coverage Display
Eb/Nt or C/I (dB)

Atoll calculates the Eb/Nt or C/I on each pixel of the TX i ic best server coverage area. The pixel colour depends on
the Eb/Nt or C/I level. Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as thresholds defined in the Display tab (Prediction properties). For each layer, a
TX i ic
TX ic
i
Eb
C
Threshold or ---
Threshold . Each layer is assigned a colour and
pixel is covered if -----
N t P CCPCH
I P CCPCH
Eb/Nt Margin or C/I Margin (dB)

Atoll calculates the Eb/Nt or C/I margin on each pixel of the TX i ic best server coverage area. The pixel colour
depends on the Eb/Nt or C/I margin value. Coverage consists of several independent layers whose visibility in the
workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction properties).
TX ic
E b TX i ic
C i
Req
Eb Nt
Req
CI
Q P CCPCH M P CCPCH or ---
Q P CCPCH M P CCPCH .
For each layer, a pixel is covered if -----
N t P CCPCH
I P CCPCH

Atoll calculates the cell edge coverage probability on each pixel of the TX i ic best server coverage area. The pixel
colour depends on the cell edge coverage probability value. Coverage consists of several independent layers whose
visibility in the workspace can be managed. There are as many layers as thresholds defined in the Display tab
(Prediction
properties).
TX ic
i
C
---
I P CCPCH
For
each
layer,
pixel
is
covered
if
TX i ic
E b
--- N t P CCPCH
Req
Q P CCPCH
or
CECP
Req
Q P CCPCH . Each layer is assigned a colour and displayed with intersections between layers.
CECP
631

9.4.2 DwPCH Reception Analysis (C/I)

C
This coverage prediction calculates and displays the C/I on the DwPCH, ---
. The coverage prediction is calculated for a
I DwPCH
given set of a terminal type, a mobility type, a service, a carrier, and for DwPTS. The best server for the coverage prediction is
determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master carrier in
case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is calculated for the selected carrier.
If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this transmitter. If you perform
this coverage prediction for the best carrier, Atoll calculates the C/I considering:

TX ic
i
TX ic
C
i
Min
Req
The pixels in the TX i ic coverage area where RSCP P CCPCH Max (TAdd P CCPCH,RSCP P CCPCH) and ---
Q DwPCH
I DwPCH
are covered and coloured according to the selected display option.
TX i ic
TX i
TX ic
RSCP DwPCH
C i
= -----------------------------------------Where ---
I DwPCH
TX i ic
N Tot DL
TX i ic
TX i ic
RSCP DwPCH
P DwPCH
= ---------------LT
The downlink total noise is calculated as follows:

TX ic
i
TX ic
i
TX ic
i
Term

Where
TX i ic
TX i ic
I Intra DL = RSCP DwPCH

With
TX i
TX i ic
I Extra DL =
TX i
Ortho
1 F DL
TX i
Term
1 F JD
and = 0
1

Total Noise
TX j ic
RSCPDwPCH
ji
TX j jc
RSCPDwPCH
TX
j
I IC DL ic jc = ---------------------------------------F IRF ic jc
TX i
Term
Eb Nt
L Path L L
L T = ----------------------------------------------------------------------------------------------------------------------TX i
Term
G G
TX i
Term
and N 0
Coverage Display
C/I (dB)
Atoll calculates the C/I on each pixel of the TX i ic best server coverage area. The pixel colour depends on the C/I
level. Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as
632

AT332_TRR_E0
many layers as thresholds defined in the Display tab (Prediction properties). For each layer, a pixel is covered if
TX ic
i
C
---
Threshold . Each layer is assigned a colour and displayed with intersections between layers.
I DwPCH
C/I Margin (dB)

Atoll calculates the C/I margin on each pixel of the TX i ic best server coverage area. The pixel colour depends on the
C/I margin value. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as thresholds defined in the Display tab (Prediction properties). For each layer, a pixel is
TX ic
C i
Req
CI
Q DwPCH M DwPCH . Each layer is assigned a colour and displayed with intersections between
covered if ---
I DwPCH
layers.

TX ic
C i
(Prediction properties). For each layer, a pixel is covered if ---
I DwPCH
Req
Q DwPCH . Each layer is assigned a colour

CECP
and displayed with intersections between layers.
9.4.3 Downlink TCH RSCP Coverage

This coverage prediction calculates and displays the RSCP for the downlink traffic channel, RSCP TCH DL . The coverage
prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for a downlink timeslot. The
considering:

The
pixels
in
the
TX i ic
TX i ic
coverage
area
where
TX i ic
Min
RSCP P CCPCH Max (TAdd P CCPCH,RSCP P CCPCH)
and
Req
RSCP TCH DL RSCP TCH DL Service Mobility are covered and coloured according to the selected display option.
TX i ic
Where RSCP TCH DL is given by:

Max
TX i ic
P TCH DL Service
RSCP TCH DL = ------------------------------------------Model
LT
TX i
Term
Model
L Path L L
Max
= ----------------------------------------------------------------------------------------------------------------------- and P TCH DL Service is the maximum downlink traffic power
TX i
Term
G G
defined for the selected service.
Model
LT
Model
In L T
, G
TX i
SA
= G DL and L
Max
TX i
SA
P TCH DL Service , if a smart antenna is available in the downlink. Otherwise, G
TX i
and L
TX i
model.
Coverage Display
DL TCH RSCP (dBm)

Atoll calculates the DL TCH RSCP on each pixel of the TX i ic best server coverage area. The pixel colour depends on
the RSCP level. Coverage consists of several independent layers whose visibility in the workspace can be managed.
633

TX ic
i
covered if RSCP TCH DL Threshold . Each layer is assigned a colour and displayed with intersections between layers.
RSCP Margin (dB)

Atoll calculates the RSCP margin on each pixel of the TX i ic best server coverage area. The pixel colour depends on
the RSCP margin value. Coverage consists of several independent layers whose visibility in the workspace can be
TX i ic
Req
RSCP
pixel is covered if RSCP TCH DL RSCP TCH DL Service Mobility M TCH DL . Each layer is assigned a colour and

TX i ic
(Prediction properties). For each layer, a pixel is covered if RSCP TCH DL
Req
CECP
RSCP TCH DL Service Mobility .
9.4.4 Uplink TCH RSCP Coverage

This coverage prediction calculates and displays the RSCP for the uplink traffic channel, RSCP TCH UL . The coverage prediction
is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for an uplink timeslot. The best server
The

pixels
in
TX ic
i
RSCP TCH UL
TX i ic
the
coverage
area
where
TX i ic
Min
and
Req
RSCP TCH UL Service Mobility are covered and coloured according to the selected display option.
TX i ic
Where RSCP TCH UL is given by:

Term
TX ic
P Max
i
RSCP TCH UL = -------------Model
LT
TX i
Term
Model
L Path L L
Term
= ----------------------------------------------------------------------------------------------------------------------- and P Max is the maximum uplink traffic power defined for the
TX i
Term
G G
selected terminal.
Model
LT
Model
In L T
,G
TX i
SA
= G UL and L
TX i
SA
Term
= L UL are calculated according to the smart antenna modelling method used, for P Max , if
a smart antenna is available in the uplink. Otherwise, G
TX
and L
TX
are read from the main antenna model.
Coverage Display
UL TCH RSCP (dBm)

Atoll calculates the UL TCH RSCP on each pixel of the TX i ic best server coverage area. The pixel colour depends on
the RSCP level. Coverage consists of several independent layers whose visibility in the workspace can be managed.
TX i ic
covered if RSCP TCH UL Threshold . Each layer is assigned a colour and displayed with intersections between layers.
634
RSCP Margin (dB)

AT332_TRR_E0
Atoll calculates the RSCP margin on each pixel of the TX i ic best server coverage area. The pixel colour depends on
the RSCP margin value. Coverage consists of several independent layers whose visibility in the workspace can be
TX ic
i
Req
RSCP
pixel is covered if RSCP TCH UL RSCP TCH UL Service Mobility M TCH UL . Each layer is assigned a colour and

TX i ic
(Prediction properties). For each layer, a pixel is covered if RSCP TCH UL
Req
CECP
RSCP TCH UL Service Mobility .
9.4.5 Downlink Total Noise

This coverage prediction calculates and displays the total noise on the downlink, N Tot DL . The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for a downlink timeslot. The best server
by this transmitter. If you perform this coverage prediction for the best carrier, Atoll calculates the downlink noise for all the
carriers but keeps the worst case value, i.e., the most interfered carrier. You can choose to display the minimum, the
maximum, or the average total noise values from among the values calculated for all the carriers. Pixels are covered and
coloured according to the total downlink noise thresholds defined in the display options.
Total downlink noise is given by: N Tot DL =
Term
RSCP TCH DL + RSCP OCCH + N 0
All TX, c, and TS
P TCH DL
P OCCH
- and RSCP OCCH = -------------With RSCP TCH DL = ------------------Model
Model
LT
LT
TX i
Term
Model
L Path L L
= ----------------------------------------------------------------------------------------------------------------------- and P TCH DL and P TCH DL are respectively the downlink traffic
TX i
Term
G G
power and the other common control channel power for the selected timeslot.
Model
LT
Model
In L T
, G
TX i
SA
= G DL and L
TX i
SA
Max
TX i
and L
TX i
model.
9.4.6 Downlink Service Area Analysis (Eb/Nt) or (C/I)

E
C
These coverage predictions calculate and display the Eb/Nt or C/I on the downlink traffic channel, ----b-
or ---
.
N t TCH DL
I TCH DL
The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for a
downlink timeslot. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the
carrier with the highest P-CCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters.
Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier does not exist on a
transmitter, there will not be any pixels covered by this transmitter. If you perform these coverage predictions for the best
carrier, Atoll calculates the Eb/Nt or C/I considering:
The

pixels
in
the
TX i ic
coverage
area
where
TX ic
i
Min
TX ic
TX i ic
E TXi ic
C i
Req
Req
Req
RSCP TCH DL RSCP TCH DL Service Mobility , and ----b-
Q TCH DL or ---
Q TCH DL are covered and
N t TCH DL
I TCH DL
coloured according to the selected display option.
635

TX

TX ic
i
TX
TX ic
i
TX ic
i
E TXi ic
RSCP TCH DL
RSCP TCH DL
Div
C
- G Proc
---
- G Div
Where ----b-
= --------------------------------------------
G
and
= --------------------------------------------DL
DL
DL
TX ic
TX ic
N t TCH DL
I
i
i
TCH DL
N Tot DL
N Tot DL
i
Max
TX ic
P TCH DL Service
i
With RSCP TCH DL = ------------------------------------------ Eb Nt
DL
LT
TX i
Eb Nt DL
Term
L Path L L
Max
= ----------------------------------------------------------------------------------------------------------------------- and P TCH DL Service is the maximum downlink traffic
TX i
Term
G G
power defined for the selected service.
Eb Nt DL
LT
Eb Nt DL
In L T
, G
TX i
SA
= G DL and L
TX i
SA
Max
TX i
and L
TX i
model.
TX i ic
TX i ic
TX i ic
Term

Where
TX i
TX i ic
TX i ic
TX i ic
TXi
Ortho
Term
I Intra DL = 1 F DL 1 F JD + 1 RSCP TCH DL + RSCP OCCH
TX i ic
TX i ic
P OCCH
With RSCP OCCH = --------------------- Eb Nt DL
LT
TX i ic
I Extra DL =
TX j ic
TX j ic
ji
TX j jc
TX j jc
TX
j
I IC DL ic jc = --------------------------------------------------------------------------F IRF ic jc
Coverage Display
Max Eb/Nt or Max C/I (dB)

TX i ic
TX ic
i
Eb
C
Threshold or ---
Threshold . Each layer is assigned a colour and displayed
N t TCH DL
I TCH DL
Effective Eb/Nt or Effective C/I (dB)

Atoll calculates the effective Eb/Nt or C/I on each pixel of the TX i ic best server coverage area. The pixel colour
depends on the effective Eb/Nt or C/I level. Coverage consists of several independent layers whose visibility in the
For
636
each
layer,
pixel
is
covered
if
E TXi ic
Req
Min ----b-
Q TCH DL Threshold
N
t TCH DL
or

AT332_TRR_E0
C TXi ic
Req
Min ---
Q TCH DL Threshold . Each layer is assigned a colour and displayed with intersections between
I
TCH
DL
layers.

TX i ic
TX i ic
Eb
C
Req
Eb Nt
Req
CI
Q TCH DL M TCH DL or ---
Q TCH DL M TCH DL . Each
N t TCH DL
I TCH DL
Required Power (dBm)

Atoll calculates the downlink required power on each pixel of the TX i ic best server coverage area. The pixel colour
depends on the required power level. Coverage consists of several independent layers whose visibility in the
Req
Q TCH DL
Req
Req
- P Max
For each layer, a pixel is covered if P TCH DL Threshold , where P TCH DL = --------------------------TCH DL Service or
TX i ic
E b
--- N t TCH DL
Req
Q TCH DL
Req
Max
P TCH DL = ------------------------ P TCH DL Service . Each layer is assigned a colour and displayed with intersections
TX i ic
C
---
I TCH DL
between layers.
Required Power Margin (dB)

Atoll calculates the downlink required power margin on each pixel of the TX i ic best server coverage area. The pixel
colour depends on the required power margin value. Coverage consists of several independent layers whose visibility
in the workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction
properties).
For
each
layer,
pixel
is
covered
Req
Req
Max
P TCH DL P TCH DL Service M arg in ,
if
where
Req
Q TCH DL
Q TCH DL
Req
Req
Max
- P Max
P TCH DL = -------------------------- P TCH DL Service . Each layer is assigned
TCH DL Service or P TCH DL = ------------------------TX i ic
TX i ic
E b
C
------
I TCH DL
N t TCH DL

(Prediction
properties).
TX i ic
C
---
I TCH DL
For
each
layer,
pixel
is
covered
if
TX i ic
E b
--- N t TCH DL
Req
Q TCH DL
or
CECP
Req
Q TCH DL . Each layer is assigned a colour and displayed with intersections between layers.
CECP
9.4.7 Uplink Service Area Analysis (Eb/Nt) or (C/I)

E
C
These coverage predictions calculate and display the Eb/Nt or C/I on the uplink traffic channel, ----b-
or ---
.
N t TCH UL
I TCH UL
The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for an uplink
timeslot. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with
the highest P-CCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the
coverage predictions are calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will
not be any pixels covered by this transmitter. If you perform these coverage predictions for the best carrier, Atoll calculates
the Eb/Nt or C/I considering:

637

The

pixels
in
TX i ic
the
coverage
area
TX ic
i
TX i ic
TX ic
Min
where
TX ic
i
E
C
i
Req
Req
Req
RSCP TCH UL RSCP TCH UL Service Mobility , and ----b-
Q TCH UL or ---
Q TCH UL are covered and
N t TCH UL
I TCH UL
coloured according to the selected display option.
TX ic
i
TX ic
i
TX i ic
E b TXi ic
RSCP TCH UL
RSCP TCH UL
Div
C
- G Proc
-- G Div
= ------------------------------------------------
G
= ------------------------------------------------Where -----
and
UL
UL
UL
TX ic
TX ic
N t TCH UL
I TCH UL
i
i
N Tot UL
N Tot UL
Term
Term
Req
Req
Term
TX i ic
Q TCH UL
Q TCH UL
P Max
Term
Term
Term
- or P Term
- and P Req = P Max --------------------------With RSCP TCH UL = ---------------------Req = P Max ------------------------TX i ic
TX i ic
Eb Nt UL
E b
C
---LT
-- I TCH UL
N t TCH UL
Eb Nt UL
LT
TX i
Eb Nt UL
Term
L Path L L
Term
= ----------------------------------------------------------------------------------------------------------------------- and P Max is the maximum power defined for the selected
TX i
Term
G G
terminal.
Eb Nt UL
In L T
,G
TX i
SA
= G UL and L
TX i
SA
Term
= L UL are calculated according to the smart antenna modelling method used, for P Max ,
if a smart antenna is available in the uplink. Otherwise, G
TX i
and L
TX i
are read from the main antenna model.
Coverage Display
Max Eb/Nt or Max C/I (dB)

TX ic
E b TXi ic
C i
Threshold or ---
Threshold . Each layer is assigned a colour and displayed
N t TCH UL
I TCH UL
Effective Eb/Nt or Effective C/I (dB)

Atoll calculates the effective Eb/Nt or C/I on each pixel of the TX i ic best server coverage area. The pixel colour
depends on the effective Eb/Nt or C/I level. Coverage consists of several independent layers whose visibility in the
For
each
layer,
pixel
is
covered
if
E b TXi ic
Req
Min -----
Q TCH UL Threshold
N
t TCH UL
or
C TXi ic
Req
Min ---
Q TCH UL Threshold . Each layer is assigned a colour and displayed with intersections between
I
TCH UL
layers.

TX i ic
TX i ic
Eb
C
Req
Eb Nt
Req
CI
Q TCH UL M TCH UL or ---
Q TCH UL M TCH UL . Each
N t TCH UL
I TCH UL
Required Power (dBm)

Atoll calculates the uplink required power on each pixel of the TX i ic best server coverage area. The pixel colour
depends on the required power level. Coverage consists of several independent layers whose visibility in the
638

AT332_TRR_E0
Req
For each layer, a
pixel is covered if
Term
P Req Threshold , where
Q TCH UL
Term
- P Term
P Req = --------------------------Max
TX ic
i
E
----b-
N t TCH UL
or
Req
Q TCH UL
Term
- P Term
P Req = ------------------------Max . Each layer is assigned a colour and displayed with intersections between layers.
TX i ic
C
---
I TCH UL
Required Power Margin (dB)

Atoll calculates the uplink required power margin on each pixel of the TX i ic best server coverage area. The pixel
colour depends on the required power margin value. Coverage consists of several independent layers whose visibility
in the workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction
Req
Q TCH UL
Term
Term
Term
- P Term
properties). For each layer, a pixel is covered if P Req P Max M arg in , where P Req = --------------------------Max or
TX i ic
E b
--- N t TCH UL
Req
Q TCH UL
Term
- P Term
P Req = ------------------------Max . Each layer is assigned a colour and displayed with intersections between layers.
TX i ic
C
-- I TCH UL

(Prediction
properties).
TX i ic
C
---
I TCH UL
For
each
layer,
pixel
is
covered
if
TX i ic
E b
--- N t TCH UL
Req
Q TCH UL
or
CECP
Req
Q TCH UL . Each layer is assigned a colour and displayed with intersections between layers.
CECP
9.4.8 Effective Service Area Analysis (Eb/Nt) or (C/I)

These coverage predictions consist of pixels covered by the both the uplink and the downlink service areas. These coverage
Eb
C
predictions calculate the Eb/Nt or C/I on the downlink and uplink traffic channels, -----
or ---
and
N t TCH DL
I TCH DL
E b
---C
---
N t TCH UL or I TCH UL , and display the pixels where both downlink and uplink Eb/Nt or C/I are above the required quality
thresholds.
The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for all the
6 timeslots. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier
with the highest P-CCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters.
Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier does not exist on a
transmitter, there will not be any pixels covered by this transmitter. If you perform these coverage predictions for the best
carrier, Atoll calculates the Eb/Nt or C/I considering:

The pixels in the TX i ic coverage area are covered and coloured according to the selected display option if all the following
conditions are satisfied:
TX i ic
Min
RSCP TCH DL RSCP TCH DL Service Mobility
RSCP TCH UL RSCP TCH UL Service Mobility
E b
C
Req
Req
---or ---
for any of the 6 timeslots
N t TCH DL Q TCH DL
I TCH DL Q TCH DL
TX i ic
Req
TX i ic
Req
TX i ic
TX i ic
639

TX ic
i

TX ic
i
E b
C
Req
Req
--- Q TCH UL or ---
Q TCH UL for any of the 6 timeslots
N t TCH UL
I TCH UL
9.4.9 Cell to Cell Interference

This coverage prediction calculates and displays the interference received by cells receiving in uplink from other cells which
are transmitting in downlink. The timeslot configuration of each cell defines the direction of the link at any given instance.
During each subframe, the direction of the link changes twice (downlink to uplink, and then uplink to downlink). These
transitions are referred to as switching points.
The coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and a timeslot. The
considering:

The mobility, service, and terminal are used to calculate the best server coverage of the interfered cell.
Assuming that a transmitter TX j is interfering a studied transmitter TX i on a timeslot, on the same carrier ic or on another
carrier jc, the cell to cell interference is given by:
TX j jc
I C2C TX i TX j =
TX j ic
TX j ic
TX j jc
j
RSCPTCH DL + RSCPOCCH + --------------------------------------------------------------------------F IRF ic jc
TX
TX j
TX ic
j
RSCP TCH DL
Where
TX j ic
RSCP TCH DL
TX j ic
P TCH DL
= --------------------------LT
TX j ic
and
TX j
TX jc
j
RSCP TCH DL
TX j jc
TX j jc
P TCH DL
= --------------------------LT
using
smart
antenna,
and
TX j
TX j jc
P TCH DL G Ant
P TCH DL G Ant
= -------------------- ---------and RSCP TCH DL = -------------------- ---------otherwise.
TX j
TX j
LT
LT
L Ant
L Ant
TX j ic
TX j
TX j jc
TX j
TX j ic
TX j jc
P OCCH G Ant
P OCCH G Ant
- ---------- and RSCP OCCH
RSCP OCCH = --------------= ---------------- ---------otherwise.
TX
TX j
LT
LT
j
L Ant
L Ant
ITU526 5
L T = L Path
ITU526 5
L Path
TX j
TX i
L TX L RX
is the path loss calculated using the ITU526-5 propagation model without antenna loss.
is the angle for the smart antenna pattern.

TX j
L Ant is the main antenna attenuation.

TX j
G Ant is the main antenna gain.

Atoll calculates the cell to cell interference on each pixel of the TX i ic best server coverage area. The pixel colour depends
on the cell to cell interference level. Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as thresholds defined in the Display tab (Prediction properties). For each layer, a pixel is
covered if I C2C TX i TX j Threshold . Each layer is assigned a colour and displayed with intersections between layers.
640

AT332_TRR_E0
9.4.10 UpPCH Interference

UpPCH is usually carried by the UpPTS timeslot. However, if the interference on UpPTS is high, from unsynchronised DwPTS
or TS0 timeslots of other cells, it is possible to shift the UpPCH to TS1. This is called UpPCH shifting. If some cells in a network
use UpPCH shifting, you can use this coverage prediction to study the interference on the shifted UpPCH of these cells from
other cells. The interference from other cells is in this case generated by the traffic on the TS1 of interfering cells.
This coverage prediction calculates and displays the uplink interference on the TS1, I TS1 UL . The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS1. The best server for the coverage
prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master
carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is calculated for the
selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this transmitter.
If you perform this coverage prediction for the best carrier, Atoll calculates the interference for all the carriers but keeps the
worst case value, i.e., the most interfered carrier. You can choose to display the minimum, the maximum, or the average total
noise. The coverage prediction is calculated using the main antenna.
TX i ic
TX i ic
Pixels in the TX i ic coverage area where RSCP P CCPCH TAdd P CCPCH Mobility and I TS1 UL Threshold are covered
and coloured according to the selected display option.
The uplink interference on TS1 is calculated from the uplink load calculated in the simulations or manually defiend for the TS1.
TX i ic
TX i
The uplink interference on TS1 is given by: I TS1 UL = N 0
TX i ic
X TS1 UL
--------------------------------TX ic
1 X i
TS1 UL
9.4.11 HSDPA Predictions

This coverage prediction calculates and displays the peak RLC throughput or the Peak MAC throughput per pixel covered by
HSDPA cells. The coverage prediction is calculated for a given set of an HSDPA terminal type, a mobility type, an HSDPA
service, a carrier, and for all downlink timeslots. The best server for the coverage prediction is determined according to the PCCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master carrier in case of N-frequency mode
compatible transmitters. Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier
does not exist on a transmitter or if it does not support HSDPA, there will not be any pixels covered by this transmitter. If you
perform these coverage predictions for the best carrier, Atoll calculates the RLC or Peak MAC throughput considering:

The pixels in the TX i ic coverage area are covered and coloured if:
TX i ic
RSCP P CCPCH TAdd P CCPCH Mobility ,
E C
Req
--- Q HS PDSCH , and
N t HS PDSCH
E C
---is enough to select a bearer for the pixels.
N t HS PDSCH
TX i ic
TX ic
i
For more information on HSDPA bearer selection, see "HSDPA Part of the Algorithm" on page 626.
Coverage Display
It is possible to colour the pixels in the coverage areas by criteria such as:
TX i ic
Min HS-PDSCH RSCP: On each pixel, Atoll calculates RSCP HS SCCH for all timeslots and selects the lowest value.
Average HS-PDSCH RSCP: On each pixel, Atoll calculates RSCP HS SCCH for all timeslots and calculates the average of
TX i ic
these values.
TX i ic
Max HS-PDSCH RSCP: On each pixel, Atoll calculates RSCP HS SCCH for all timeslots and selects the highest value.
E C TXi ic
Min HS-PDSCH Ec/Nt: On each pixel, Atoll calculates -----
for all timeslots and selects the lowest value.
N t HS PDSCH
641

E C TXi ic
Average HS-PDSCH Ec/Nt: On each pixel, Atoll calculates -----
for all timeslots and calculates the average
N t HS PDSCH
of these values.
TX i ic
EC
Max HS-PDSCH Ec/Nt: On each pixel, Atoll calculates -----
for all timeslots and selects the highest value.
N t HS PDSCH
Peak RLC Throughput: After selecting the bearer, Atoll reads the corresponding RLC peak throughput. This is the
highest throughput that the bearer can provide on each pixel. The pixel colour depends on the peak RLC throughput.
Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as thresholds defined in the Display tab (Prediction properties). Each layer is assigned a colour and displayed
Peak MAC Throughput: Atoll displays the Peak MAC throughput ( R DL ) provided on each pixel. The Peak MAC
MAC
throughput is calculated as follows:

MAC
R DL
= S Block 500
Where, S Block is the transport block size (in kbits) of the selected HSDPA bearer; it is defined for each HSDPA bearer
in the related table. The value 500 corresponds to the number of blocks per second (there are 4 blocks per TTI and
2000
2000 TTI in one second, i.e ------------ blocks per second).
4
The pixel colour depends on the Peak MAC throughput. Coverage consists of several independent layers whose
(Prediction properties). Each layer is assigned a colour and displayed with intersections between layers.
E b TXi ic TS
Max DL A-DPCH Eb/Nt: Atoll displays the A-DPCH Eb/Nt at the receiver ( -----
) for the best server and the
N t TCH DL Max
selected timeslot. No power control is performed as in simulations. Here, Atoll determines downlink traffic channel
quality at the receiver for a maximum traffic channel power allowed for the best server.
TX i ic TS
Eb
Max UL A-DPCH Eb/Nt: Atoll displays the A-DPCH Eb/Nt at the best server ( -----
) and the selected
N t TCH UL Max
timeslot. No power control is performed as in simulations. Here, Atoll determines uplink traffic channel quality for the
maximum terminal power allowed.
TX ic
i
HS-SCCH Power: On each pixel, Atoll calculates P HS SCCH for the selected timeslot.
HS-SCCH RSCP: On each pixel, Atoll calculates RSCP HS SCCH for the selected timeslot.
E c TX i ic
HS-SCCH Ec/Nt: On each pixel, Atoll calculates -----
for the selected timeslot.
N t HS SCCH
HS-SICH Power: On each pixel, Atoll calculates P HS SICH for the selected timeslot.
HS-SICH RSCP: On each pixel, Atoll calculates RSCP HS SICH for the selected timeslot.
Ec Mi
HS-SICH Ec/Nt: On each pixel, Atoll calculates -----
N t HS SICH
HS-PDSCH RSCP: On each pixel, Atoll calculates RSCP HS PDSCH for the selected timeslot.
Ec
HS-PDSCH Ec/Nt: On each pixel, Atoll calculates -----
N t HS PDSCH
TX i ic
Mi
Mi
TX i ic
TX i ic
9.5 Smart Antenna Modelling

Atoll calculates the smart antenna gains and losses in the direction of a user during the simulations, and in the direction of
each pixel in coverage predictions. During simulations, Atoll determines the gains and losses using the smart antenna models.
In coverage predictions, Atoll determines the gains and losses from the angular distributions calculated during the simulations
for each timeslot and stored in the Cell Parameters per Timeslot table.
If a smart antenna model is only downlink or only uplink, the other direction uses the main antenna gain and losses for
calculations. Therefore,
642
If a smart antenna is available on the downlink and uplink:

AT332_TRR_E0
TX
SA
TX
SA
TX
SA
TX
SA
G UL = G UL , L UL = L UL and G DL = G DL , L DL = L DL
If a smart antenna is available on the downlink only:

TX
SA
TX
SA
TX
TX
TX
G DL = G DL , L DL = L DL and G UL = G Ant , L UL = L
= L Total UL
TX
= L Total DL
If a smart antenna is available on the uplink only:

TX
SA
TX
SA
TX
TX
TX
G UL = G UL , L UL = L UL and G DL = G Ant , L DL = L
TX
If no smart antenna equipment is defined:

TX
TX
TX
TX
G DL = G UL = G Ant , L UL = L
TX
TX
= L Total UL , and L DL = L
TX
= L Total DL
9.5.1 Modelling in Simulations

9.5.1.1 Grid of Beams Modelling
A grid-of-beams smart antenna, called GOB, consists of more than one directional antenna pattern (beam) in different
directions. Each beam of a GOB has a different azimuth so that the GOB as a whole covers an entire sector. During the
simulations, Atoll determines the most suitable beam from the GOB for each user served by the smart antenna. The most
suitable beam (best beam) is the one which provides the highest gain towards the served user:
BeamBest = Beam
H
H
V
Max G Beam L Beam L Beam
V
Where G Beam , L Beam , and L Beam are the gains, horizontal, and vertical attenuations of the beams of the GOB. In words, the
best beam is the one among all the beams of a GOB that has the highest difference between gain, and horizontal and vertical
SA
SA
SA
SA
attenuations. The gains and losses of the GOB ( G DL , G UL , L DL , and L UL ) are determined from the selected best beam.
The following example shows how Atoll calculates the GOB gains and losses.
Example:
Let us assume a GOB with 5 beams that have the same vertical patterns, and whose horizontal patterns are pointed towards
different directions as shown in the figure below:
Figure 9.3: Grid Of Beams Modelling

Let us assume that all the beams and the main antenna have the same 18 dBi gain, and the vertical attenuation at the user
location is 15 dB, which is also the same for all the beams because we assume that the vertical patterns are the same.
If the user is located at = 70 azimuth, as shown in the figure below, Atoll determines the best beam, which has the
highest gain towards , as follows:
643

Horizontal
Vertical
Attenuation (dB) Attenuation (dB)
G Beam L Beam L Beam
Total Gain
(dB)
15
18 - 60 - 15
-57
60
15
18 - 60 - 15
-57
Beam
Gain (dBi)
18
60
30
18
60
18
2.21
15
18 - 2.21- 15
0.79
-30
18
60
15
18 - 60 - 15
-57
-60
18
60
15
18 - 60 - 15
-57
Transmitter
located

azimuth
Figure 9.4: GOB Modelling - Determination of the Best Beam

In our example, the total gain of the beam at 60 is the highest. Therefore this beam is selected as the best beam.
If this beam has been selected in the downlink,
SA
SA
G DL = 18 dB and L DL = L Beam + L Beam = 17,21 dB

If this beam has been selected in the uplink,
SA
SA
G UL = 18 dB and L UL = L Beam + L Beam = 17,21 dB
9.5.1.2 Adaptive Beam Modelling

An adaptive beam smart antenna is capable of steering a given antenna pattern towards the direction of the served signal. In
Atoll, this is modelled using a single antenna pattern, called a beam because of its highly directional shape. During the
simulations, this adaptive beam is oriented in the direction of each served user in order to model the effect of the smart
antenna.
SA
SA
SA
The adaptive beam gains ( G DL and G UL ) are the antenna gains defined for the beam, and the adaptive beam losses ( L DL and
SA
L UL ) are the horizontal and vertical pattern attenuations L Beam + L Beam towards the user direction.
The following example shows how Atoll calculates the adaptive beam gains and losses.
Example:
Let us assume an adaptive beam smart antenna selected for a transmitter along with a main antenna. Let us assume that the
adaptive beam and the main antenna have the same 18 dBi gain, and the vertical attenuation at the user location is 15 dB.
If the user is located at = 60 azimuth, as shown in the figure below:
644

AT332_TRR_E0
Transmitter
located

azimuth
Figure 9.5: Adaptive Beam Modelling - Determination of the Best Beam

If the adaptive beam smart antenna is selected in the downlink, the gain and losses of the adaptive beam at are:
SA
SA
G DL = 18 dB and L DL = L Beam + L Beam = 15 dB

If the adaptive beam smart antenna is selected in the uplink, the gain and losses of the adaptive beam at are:
SA
SA
G UL = 18 dB and L UL = L Beam + L Beam = 15 dB

H
In fact, as the ideal beam steering algorithm steers the beam towards the served user, L Beam = 0 . These values are used in
interference calculation to determine the downlink interfering signal due to transmission towards the served user, as well as
for calculating the uplink interfering signals received at transmitter when decoding signal received from the served user.
9.5.1.3 Statistical Modelling

A statistical modelling approach is also available in Atoll which can be used to model the effect of smart antennas through C/
I gains. You can create smart antenna equipment in Atoll based on the statistical approach by providing C/I gains and their
cumulative probabilities for different spreading angles, Spread .
You can assign a spreading angle to each clutter class in your document. Atoll reads the clutter class in which the served user
is located to determine the spreading angle. Different clutter types have different spreading effects on the propagation of
radio waves. Urban and dense urban clutter types introduce more multipath and spread the signal at a wider angle than an
open or rural clutte type.
Once you have assigned the spreading angles to clutter classes, you can enter the C/I gains and their cumulative probabilities
for each spreading angle, in the smart antenna equipment based on the statistical model. For each smart antenna equipment
based on statistical modelling, you can set a probability threshold, TProb
SA
To find the smart antenna gain, Atoll determines the clutter class of the served user, it reads the spreading angle from the
clutter class properties, it reads the probability threshold from the smart antenna properties, and reads the smart antenna C/
I gain defined for the Probability = 1 TProb
SA
corresponding to the spreading angle.
The following example shows how Atoll calculates the statistical C/I gains and losses.
Example:
Let us assume that the served user is located at a an urban clutter class with Spread = 10 . The smart antenna equipment
SA
SA
has TProb = 80 % . Atoll will read the smart antenna C/I gain G for Prob = 20 % . If a gain for the exact probability
value of 20% is not defined, Atoll linearly interpolates the gain value from the two surrounding values.
If G
SA
Prob = 19%
= 4,6298 dB and G
SA
Prob = 20,4%
= 4,7196 dB , then G
SA
Prob = 20%
= 4,6941 dB
The smart antenna gains are the same for uplink and downlink. Their are no losses for this type of smart antenna equipment.
Negative values of C/I gains are considered as losses.
9.5.1.4 Beamforming Smart Antenna Models

See "Beamforming Smart Antenna Models" on page 43.
645

9.5.1.5 3rd Party Smart Antenna Modelling

3rd party smart antenna models can be used in Atoll to determine the gains and losses during the simulations for a given user
distribution generated. The smart antenna gains and losses are used during the simulations and the results are stored in the
Cell Parameters per Timeslot table, which can be used in coverage predictions.
9.5.2 Construction of the Geographic Distributions

During simulations, Atoll uses the smart antenna model selected for each transmitter to calculate the smart antenna gains
and losses. These values are calculated and stored for each user generated for the simulations. Therefore, these values are
calculated and are available for the given locations of the users, i.e., points, only. Atoll uses the Angular Step value that you
set when creating and running simulations to construct the geographic distributions of these results.
Once Atoll has calculated the downlink traffic power and the uplink load using the smart antenna gains and losses determined
as explained in the previous section, at the location of a given user, it calculates the same for points located at the angle equal
to that of the Angular Step of the simulations.
At the end of the simulations, Atoll has a number of points, Angular Step apart, available with the values of these results. The
geographic distribution of these results, i.e., downlink traffic power and uplink loads, is constructed by connecting the
resulting value points.
The following example explains how the geographic distribution of downlink traffic power is created. The geographic
distribution of uplink loads is constructed in the same manner.
Example:
Let us assume a smart antenna equipment using adaptive beam modelling. The angular step defined for the simulations is
Step = 30 . Therefore, the results are calculated for each point located at regular steps of 30 , i.e., 12 points. The downlink
traffic power at the served user (W) with the adaptive beam pointing in the users direction is P W . The downlink traffic
powers, using the same adaptive beam pointed towards the served user, at the 12 other points are also determined.
Figure 9.6: Construction of the Geographic Distribution of Downlink Traffic Power

The resulting geographic distribution is formed by linearly joining the obtained results.
646

AT332_TRR_E0
Figure 9.7: Geographic Distribution of Downlink Traffic Power

The accuracy of the geographic distribution depends upon the value of the angular step. A radiation pattern created at a 1
step will be much more accurate than one created at 45 , for example. But, the latter will be computed 45 times faster than
the first. The value of the Angular Step should be the best possible compromise between calculation speed and accuracy.
9.5.3 Modelling in Coverage Predictions

The results of Monte Carlo simulations, including the smart antenna results, can be stored in the Cells and in the Cell
Parameters per Timeslot tables, and can be used to carry out coverage predictions. The main results of Monte Carlo
simulations used in coverage predictions are:
If a smart antenna is used in both uplink and downlink:

Geographic distribution of UL load X
UL
DL
and DL traffic power P Traffic
If a smart antenna is used in downlink only:

DL
Geographic distribution of DL traffic power P Traffic
Without smart antenna:

UL load X
UL
DL
and DL traffic power P Traffic
The uplink load and the downlink traffic power at a given pixel are determined by calculating the angle of that pixel with
respect to the transmitter azimuth, and reading the uplink load and downlink traffic power from the geographic distribution
results. If an exact value for the angle is not available, the load and power are determined using linear interpolation for the
given angle between two available values.
For example, the figure below shows the distribution of downlink traffic power and uplink traffic load results from a
DL 315
simulation. For a pixel located at = 315 , the downlink traffic power P Traffic
DL 315
from these results. In this example, P Traffic
30 dBm , and X
UL 315
and the uplink load X
UL 315
are read
= 2,75 % .
For each pixel, Atoll determines the downlink traffic powers and the uplink loads from all the transmitters.
647

Figure 9.8: Geographic Distribution of downlink traffic power and uplink load
9.5.4 HSDPA Quality and Throughput Analysis

Fast link adaptation (or Adaptive Modulation and Coding) is used in HSDPA. The power on the HS-DSCH channel is transmitted
at a constant power while the modulation, the coding, and the number of codes are changed to adapt to the radio conditions
variations. Based on the reported channel quality indicator (CQI), the Node-B may change the modulation (QPSK and
optionally 16QAM), the coding, and the number of codes every 2 ms during a communication.
Fast link adaptation is modelled in a dedicated HSDPA coverage prediction. Let us assume each bin on the map corresponds
to a probe receiver with HSDPA capable terminal, mobility, and HSDPA service. The probe receiver on each bin is allocated the
cells HSDPA. This receiver may be using a specific carrier or all of them. The probe receiver does not create any interference.
Atoll calculates on each bin either the best pilot quality (P-CCPCH Ec/Nt) or the best HS-PDSCH quality (HS-PDSCH Ec/Nt); this
depends on the option selected in Global parameters (HSDPA part): CQI based on P-CCPCH quality or CQI based on HS-PDSCH
quality (CQI means channel quality indicator). Then, it determines the HS-PDSCH CQI, deduces the best HSDPA bearer that can
be used and selects the suitable bearer so as to comply with cell and terminal user equipment capabilities. Once the bearer
selected, Atoll finds the highest downlink throughput that can be carried at each bin and may deduce the application
throughput. Coverage area is limited by the RSCP P-CCPCH threshold.
The coverage prediction can be calculated for an HSDPA compatible terminal, an HSDPA service, a mobility, a carrier, and a
downlink timeslot. Smart antenna results are taken into account in the computation of this study.
9.5.4.1 Fast Link Adaptation Modelling

As explained above, the way of calculating the dedicated HSDPA study depends on if CQI is based on the P-CCPCH quality or
on the HS-PDSCH quality.
9.5.4.1.1
CQI Based on P-CCPCH Quality

When the option CQI based on CPICH quality is selected, Atoll proceeds as follows.
P-CCPCH Quality Calculation
Ec
corresponds to the P-CCPCH quality.
Nt
P CCPCH
Two options, available in Global Parameters, may be used to calculate Nt: option Without useful signal or option Total noise.
Therefore, we have:
TXi
BTS RSCP P CCPCH ic

Ec ---- ic
= -------------------------------------------------------------------- for the total noise option,
Nt
P CCPCH
DL
N Tot ic
And
TXi
BTS RSCP P CCPCH ic

Ec ---- ic
= ------------------------------------------------------------------------------------------------------------ for the without useful signal option.
Nt
P CCPCH
DL
TXi
N Tot ic 1 BTS RSCP P CCPCH ic
With
648

AT332_TRR_E0
DL
DL
DL
DL
Term
N Tot ic = I Intra ic + I Extra ic + I Inter Carrier ic jc + N 0

DL
I Inter Carrier ic jc is the inter-carrier interference from a carrier jc to another carrier ic on the downlink, which is reduced
by the interference reduction factor F IRF ic jc defined for the pair (ic, jc).
P P CCPCH ic
TXi
RSCP P CCPCH ic = ------------------------------LT
CI
L Path L TX L Term L Body L Indoor M Shadowing

L T = ---------------------------------------------------------------------------------------------------------------------G TX G Term
Term
BTS , and N 0
P-CCPCH CQI Determination

Let us assume the following notation: CQI P CCPCH corresponds to the P-CCPCH CQI. CQIP CCPCH is deduced from the table
Ec
. This table is defined for the terminal reception equipment and the specified mobility.
CQIP CCPCH = f ------ ic
Nt
P CCPCH
HS-PDSCH Quality Calculation
1st step: Atoll calculates the HS-PDSCH power ( P HS PDSCH ).
P HSDPA ic is the power available for HSDPA on the carrier ic. This parameter is a user-defined cell input.
Therefore, we have:
n HS SCCH is the number of HS-SCCH channels and P HS SCCH ic is the HS-SCCH power on carrier ic. It is either fixed by the
Req
Ec
user. P HS SCCH ic is controlled so as to reach the required HS-SCCH Ec/Nt ( ------ ic
). It is specified in mobility
Nt
HS SCCH
properties.
We have:
TXi
BTS RSCP HS SCCH ic

Ec ----
- for the total noise option,
Nt ic HS SCCH = ---------------------------------------------------------DL
N Tot ic
And
TXi
BTS RSCP HS SCCH ic

Ec ----= -------------------------------------------------------------------------------------------------------------------------------------------------------- for the without useful signal option.
ic
Nt
HS SCCH
DL
DL
Term
TXi
N Tot ic 1 F Ortho 1 F JD BTS RSCP HS SCCH ic
With
DL
DL
DL
DL
Term

DL
P HS SCCH ic
TXi
RSCP HS SCCH ic = ------------------------------LT
and
CI

L T = ---------------------------------------------------------------------------------------------------------------------G TX G Term
Term
BTS , F Ortho , F JD
Term
and N 0
649

Therefore,
Req
EcDL
---- ic
N Tot ic
HS SCCH
Nt
TXi
RSCP HS SCCH ic = ------------------------------------------------------------------- L T for the total noise option,
BTS
And
Req
EcDL
----
ic
N Tot ic
Nt
HS SCCH
TXi
RSCP HS SCCH ic = ------------------------------------------------------------------------------------------------------------------------------------------- L T for the without useful signal option.
Req
Ec 1 + 1 F DL 1 F Term ----
Ortho
JD
BTS
Nt ic HS SCCH
2nd step: Then, Atoll calculates the HS-PDSCH quality
Ec
Nt
HS PDSCH
Therefore, we have:
TXi
BTS RSCP HS PDSCH ic

Ec ---- ic
= ------------------------------------------------------------- for the total noise option,
Nt
HS PDSCH
DL
N Tot ic
And
TXi

Ec ----= ----------------------------------------------------------------------------------------------------------------------------------------------------------- for the without useful signal option.
ic
Nt
HS PDSCH
TXi
RSCP HS PDSCH ic
DL
DL
Term
N Tot ic 1 F Ortho 1 F JD BTS --------------------------------------------n
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5).
With
DL
DL
DL
DL
Term

DL
P HS PDSCH ic
TXi
RSCP HS PDSCH ic = ---------------------------------LT
And
CI

L T = ---------------------------------------------------------------------------------------------------------------------G TX G Term
Term
Term
and N 0
HS-PDSCH CQI Determination

The best bearer that can be used depends on the HS-PDSCH CQI. Let us assume the following notation: CQI HS PDSCH
corresponds to the HS-PDSCH CQI. Atoll deduces CQI HS PDSCH as follows:
CQI HS PDSCH = CQI P CCPCH P P CCPCH + P HS PDSCH
Bearer Selection
Knowing the HS-PDSCH CQI, Atoll finds the best bearer that can be used in the table Best Bearer=f(HS-PDSCH CQI). This table
is defined for the terminal reception equipment and the specified mobility.
Then, Atoll checks if best bearer characteristics are compliant with cell and user equipment category capabilities. Atoll selects
the bearer which is the best bearer compliant with the cell and UE category capabilities.
Bearer characteristics are provided in the HSDPA Bearer table. Assuming the best bearer = 23. Characteristics of this bearer
are:
650
Transport block size: 9719 Bytes

AT332_TRR_E0

16QAM modulation used: Yes
Peak Throughput: 4.48 Mb/s
Figure 9.9: Radio Bearers Table

Assuming user equipment category = 3. Its capabilities are:
Maximum transport block size: 7298 Bytes

Maximum number of HS-PDSCH channels used: 5
16QAM modulation used: Yes
Minimum number of TTI between two TTI used: 2
Figure 9.10: UE Categories Table

HSDPA cell capabilities are:
Maximum number of HS-PDSCH channels: 15.
The bearer 23 cannot be selected because:
The number of HS-PDSCH channels (7) exceeds the maximum number of HS-PDSCH channels the terminal can use (5),
And the transport block size (9719 Bytes) exceeds the maximum transport block size (7298 Bytes) the terminal can
carried.
In the Bearer table, Atoll searches a suitable bearer and selects the bearer index 22.
The number of HS-PDSCH channels (5) does not exceed the maximum number of HS-PDSCH channels the terminal can
use (5) and the maximum number of HS-PDSCH channels available at the cell level (15),
651

The transport block size (7168 Bytes) does not exceed the maximum transport block size (7298 Bytes) the terminal can
carried.
16QAM modulation is supported by the terminal.
HS-PDSCH Quality Update

Once the bearer selected, Atoll knows the number of HS-PDSCH channels. Therefore, when the method Without useful
signal is used, Atoll can recalculate the HS-PDSCH quality with the real number of HS-PDSCH channels (A default value of 5
was taken into account in the first HS-PDSCH quality calculation).
9.5.4.1.2
CQI Based on HS-PDSCH Quality

When the option CQI based on HS-PDSCH quality is selected, Atoll proceeds as follows.
HS-PDSCH Quality Calculation
1st step: Atoll calculates the HS-PDSCH power ( P HS PDSCH ).
P HSDPA ic is the power available for HSDPA on the carrier ic. This parameter is a user-defined cell input.
Therefore, we have:
n HS SCCH is the number of HS-SCCH channels and P HS SCCH ic is the HS-SCCH power on carrier ic fixed by the user. The HSReq
Ec
SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt ( ------ ic
) specified in mobility properties.
Nt
HS SCCH
We have:
TXi
BTS RSCP HS SCCH ic

Ec ---- ic
= ---------------------------------------------------------- for the total noise option,
Nt
HS SCCH
DL
N Tot ic
And
TXi
BTS RSCP HS SCCH ic

Ec ----
- for the without useful signal option.
Nt ic HS SCCH = -------------------------------------------------------------------------------------------------------------------------------------------------------DL
DL
Term
TXi
N Tot ic 1 F Ortho 1 F JD BTS RSCP HS SCCH ic
With
DL
DL
DL
DL
Term

DL
P HS SCCH ic
TXi
RSCP HS SCCH ic = ------------------------------LT
i
And
CI

L T = ---------------------------------------------------------------------------------------------------------------------G TX G Term
Term
Term
and N 0
Therefore,
Req
EcDL
---- ic
N Tot ic
Nt
HS SCCH
P HS SCCH ic = ------------------------------------------------------------------- L T for the total noise option,
BTS
And
652

AT332_TRR_E0
Req
EcDL
----
Nt ic HS SCCH N Tot ic
P HS SCCH ic = ------------------------------------------------------------------------------------------------------------------------------------------- L T for the without useful signal option.

Req
Ec 1 + 1 F DL 1 F Term ----
ic
Ortho
JD
BTS
Nt
HS SCCH
2nd step: Then, Atoll evaluates the HS-PDSCH quality
Ec
Nt
HS PDSCH
We have:
TXi

Ec ---- ic
= ------------------------------------------------------------- for the total noise option,
Nt
HS PDSCH
DL
N Tot ic
And
TXi

Ec ----
= ----------------------------------------------------------------------------------------------------------------------------------------------------------- for the without useful signal option.
Nt ic HS PDSCH
TXi
RSCP HS PDSCH ic
DL
DL
Term
N Tot ic 1 F Ortho 1 F JD BTS --------------------------------------------n
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5). Then, it deduces the HS-PDSCH CQI and the
bearer to be used. Once the bearer selected, Atoll exactly knows the number of HS-PDSCH channels and recalculates the HSPDSCH quality with the real number of HS-PDSCH channels.
With
DL
DL
DL
DL
Term

DL
P HS PDSCH ic
TXi
RSCP HS PDSCH ic = ---------------------------------LT
And
CI

L T = ---------------------------------------------------------------------------------------------------------------------G TX G Term
Term
Term
and N 0
HS-PDSCH CQI Determination

Let us assume the following notation: CQIHS PDSCH corresponds to the HS-PDSCH CQI. CQI HS PDSCH is deduced from the
Ec
. This table is defined for the terminal reception equipment and the specified
table CQI HS PDSCH = f ------ ic
Nt
HS PDSCH
mobility.
Bearer Selection
The bearer is selected as described in "Bearer Selection" on page 650.
9.5.4.2 Coverage Prediction Display Options

Three display options are available in the study property dialog box.
9.5.4.2.1
Colour per CQI

Atoll displays either the P-CCPCH CQI when the selected option in Global Parameters (HSDPA part) is CQI based on P-CCPCH
quality, or the HS-PDSCH CQI when considering the CQI based on HS-PDSCH quality option.
Coverage consists of several layers with a layer per CQI threshold ( CQIThreshold ). For each layer, area is covered if
CQI CQI Threshold . Each layer is assigned a colour and displayed with intersections between layers.
653

9.5.4.2.2
Colour per Peak Throughput

After selecting the bearer, Atoll reads the corresponding RLC peak throughput. This is the highest throughput that the bearer
can provide on each bin.
DL
Coverage consists of several layers with a layer per possible peak throughput ( R Peak ). For each layer, area is covered if the
peak throughput can be provided. Each layer is assigned a colour and displayed with intersections between layers.
9.5.4.2.3
Colour per HS-PDSCH Ec/Nt

Atoll displays on each bin the HS-PDSCH quality. Coverage consists of several layers with a layer per threshold. For each layer,
Ec
area is covered if ------ ic
Threshold . Each layer is assigned a colour and displayed with intersections between
Nt
HS PDSCH
layers.
9.6 N-Frequency Mode and Carrier Allocation

Transmitters that support N-frequency mode are multi carrier transmitters with a master and one or more slave carrier. You
can assign master and slave carriers to transmitters manually, or use the automatic frequency allocation in Atoll to assign
carrier types automatically.
9.6.1 Automatic Carrier Allocation

For each transmitter, Atoll determines a list of "near" transmitters. For any transmitter TXi, its "near" transmitters are
geographically located close to the transmitter, and are sorted according to their distance from it. The calculation of distance
between TXi and any other transmitter TXj is performed using the equation below:
D
TX i TX j
Where D
= d
TX i TX j
TX i TX j
1 + x cos cos 2
is the weighted distance between TXi and TXj, d
TX i TX j
is the real distance between between TXi and TXj
considering any offsets with respect to the site locations, x is set to 15 % so that the maximum variation in D
TX i TX j
due to
the azimuths does not exceed 60 %. and are calculated from the azimuths of the two cells as shown in Figure 9.11 on
page 654.
Figure 9.11: Weighted Distance Between Transmitters

The above formula implies that two transmitters facing each other will have a shorter weighted distance between them than
the real distance, and two transmitters pointing in opposite directions will have a greater weighted distance.
Allocation of All Carriers
Co-N-Frequency Allocation
654
Diff-N-Frequency Allocation

AT332_TRR_E0
Co-N-Frequency Allocation
Diff-N-Frequency Allocation
Atoll assigns the same carriers to cells of each co-site

transmitter.
Atoll assigns different carriers to cells of each co-site

transmitter.
Allocation of Master Carriers

Atoll assigns one master carrier to each transmitter TXi, such that the master carrier of TXi is different from the master carrier
of TXj, where TXj belongs to the list of "near" transmitters. The master carrier is one of the cells defined in the transmitter. All
the other cells of the transmitter are assigned the carrier-type "slave".
For transmitters that support the N-frequency mode and have master carriers properly assigned, Atoll performs the
neighbour and scrambling code allocation for the master carrier only.
9.7 Neighbour Allocation

Atoll permits the automatic allocation of intra-technology neighbours in a TD-SCDMA document. The intra-technology
neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all the cells of TBC
transmitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfill the following conditions:
They are active

Their transmitters support the N-frequency mode, and the cells are master carriers of their transmitters (neighbours
are not allocated to standalone carriers)
They satisfy the filter criteria applied to the Transmitters folder
They are located inside the focus zone

zone.
"Neighbour Allocation for All Transmitters" on page 655.


We assume that we have a reference cell A and a candidate neighbour, cell B. When automatic allocation starts, Atoll checks
1. The distance between both cells must be less than the user-defined maximum inter-site distance. If the distance
between the reference cell and the candidate neighbour is greater than this value, the candidate neighbour is
discarded.
"Appendix: Calculation of the Inter-Transmitter Distance" on page 660. Otherwise, it takes the real distance.
[Neighbours]
Force co-site cells as neighbours: This option enables you to force cells located on the same site as reference cell
in the candidate neighbour list. This constraint can be weighted among the others and ranks the neighbours
through the importance field.
Force adjacent cells as neighbours: This option enables you to force cells geographically adjacent to the reference
cell in the candidate neighbour list. This constraint can be weighted among the others and ranks the neighbours
through the importance field.
655

Adjacency criterion: Geographically adjacent cells are determined on the basis of

their best server coverages in TD-SCDMA projects. Let CellA be a candidate neighbour
cell of CellB. CellA is considered adjacent to CellB if there exists at least one pixel in
the CellB best server coverage area (and P-CCPCH RSCP of CellB > P-CCPCH RSCP
T_Add) where CellA is best server (of several cells have the same best server value) or
CellA is the second best server that enters the handover set (i.e., P-CCPCH RSCP of
CellA > P-CCPCH RSCP T_Drop and P-CCPCH RSCP of CellA > P-CCPCH RSCP of CellB T_Comp.)
When this option is selected, adjacent cells are sorted and listed from the most
adjacent to the least, depending on the above criterion. Adjacency is relative to the
number of pixels satisfying the criterion.
Force symmetry: This option enables you to force the reciprocity of a neighbourhood link. Therefore, if the
reference cell is a candidate neighbour of another cell, the later will be considered as candidate neighbour of the
reference cell.
If the neighbours list of a transmitter is full, the reference transmitter will not be added
as a neighbour of that transmitter and that transmitter will be removed from the
reference transmitters neighbours list. You can force Atoll to keep that transmitter in
the reference transmitters neighbours list by adding the following option in the Atoll.ini
file:
[Neighbours]
3. If the Use Coverage Conditions check box is selected, there must be an overlapping zone ( S A S B ) with a given cell
N-frequency handover is a baton handover. Assuming that the reference cell A and the candidate cell B are located
inside a continuous layer of cells:
SA is the area where the cell A is the best serving cell.
The P-CCPCH RSCP from the cell A is greater than the P-CCPCH RSCP T_Add.
The P-CCPCH RSCP from the cell A is greater than the P-CCPCH RSCP from all other cells.
SB is the area where the cell B can enter the handover set.
656
The P-CCPCH RSCP from the cell B is greater than the P-CCPCH RSCP T_Drop.
The P-CCPCH RSCP from the cell B is greater than the P-CCPCH RSCP from the cell A minus the P-CCPCH RSCP
T_Comp.

AT332_TRR_E0
Figure 9.12: N-frequency Neighbour Allocation

SA SB
- 100 ), which it compares with the % minimum covered
Atoll calculates the percentage of covered area ( ----------------SA
area. If this percentage is not exceeded, the candidate neighbour B is discarded.
The coverage condition can be weighted among the others and ranks the neighbours through the importance field.
: Overlapping Coverages
4. The importance of neighbours.
Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason.
Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list
if the maximum number of neighbours to be allocated to each transmitter is exceeded.
If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be
allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8 (having the highest importance
657

values) will be allocated to the reference cell. Note that maximum numbers of neighbours can be defined at the cell
level (properties dialog box or Cells table). If defined there, this value is taken into account instead of the default one
available in the dialog box.
In the Results part, Atoll provides the list of neighbours, the number of neighbours, and the maximum number of neighbours
allowed for each cell. In addition, it indicates the importance (in %) of each neighbour and the allocation reason, i.e., a
neighbour may be marked as exceptional pair, co-site, adjacent, coverage, or symmetric. For neighbours accepted for co-site,
adjacency, and coverage reasons, Atoll displays the percentage of area that satisfies the coverage conditions and the
corresponding surface area (km2), the percentage of area that satisfies the adjacency conditions and the corresponding
surface area (km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing.
You do not require simulations or coverage predictions for an automatic neighbour

allocation. For automatic neighbour allocation, Atoll automatically calculates the
missing path loss matrices.
Although no specific terminal, mobility, or service is selected for automatic neighbour
allocation, the algorithm tries to find the maximum number of neighbours by
selecting:
The service with the lowest body loss

The terminal with the highest difference between Gain and Losses. If this is the
same for all terminals, Atoll uses the terminal with the lowest noise figure.
Mobility does not impact the allocation
relationship already exists and the Delete existing neighbours option is not selected.
In this case, Atoll displays a warning message in the Event Viewer indicating that the
constraint on the forbidden neighbour will be ignored by the algorithm because the
neighbour already exists.
Symmetric neighbour relations are only added to the neighbour lists if the neighbour
lists are not already full. Thus, if the cell B is a neighbour of the cell A, but cell A is not
a neighbour of the cell B, there can be two possibilities:
1. There is space in the cell B neighbour list: cell A will be added to the list. It will be
the last one.
2. The cell B neighbour list is full: Atoll will not include cell A in the list and will remove
the symmetric relation by deleting cell B from the cell A neighbour list.
If you select Force exceptional pairs and Force symmetry options, Atoll considers the
constraints between exceptional pairs in both directions so as to respect the
symmetric relation. On the other hand, if a neighbour relation is forced in one
direction and forbidden in the other, symmetry cannot be respected. In this case,
Atoll displays a warning message in the Event Viewer.

Transmitter
TBA cells
Neighbours of TBA cells marked as exceptional pair, adjacent, or symmetric
Neighbours of TBA cells that satisfy coverage conditions

Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason and the
distance, and to quantify the neighbour importance. The neighbour importance depends on the distance from the reference
transmitter and on the neighbourhood cause (cf. table below); this value varies between 0 and 100%.
658
Neighbourhood cause
When
Importance value
Existing neighbour

Existing importance
Exceptional pair
100 %

AT332_TRR_E0
Neighbourhood cause
When
Importance value
Co-site cell

selected
Adjacent cell

selected

relationship
transmitter.
d Di = 1 ---------d max
of antennas. For information on the effective distance calculation, see "Appendix: Calculation of the Inter-Transmitter
[Neighbours]

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
30%
Min(A)
30%
Max(A)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

Neighbourhood cause
Importance Function

No
10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
No
Yes
30%+30%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Yes
Yes
60%+40%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Co-site
Adjacent
No
Where:
659

overlapping.
causes.
9.7.4 Appendix: Calculation of the Inter-Transmitter Distance


9.8 Scrambling Code Allocation

Downlink scrambling codes enable mobile to distinguish one cell from another. In TD-SCDMA, there are 128 scrambling codes
(or P-CCPCH midamble codes) distributed in 32 clusters of 4 codes each. A different DL synchronisation code, or SYNC_DL
code, is assigned to each cluster. Scrambling codes are numbered from 0 to 127, and SYNC_DL codes from 0 to 31.
Depending on the options you select for automatic allocation of scrambling and SYNC_DL codes, Atoll takes into account
either all the cells of TBC transmitters, or only cells of active and filtered transmitters located inside the computation zone.
Atoll calculates a scrambling code and a SYNC_DL code to all these cells. But, it allocates scrambling codes and SYNC_DL codes
only to TBA cells (cells to be allocated). TBA cells are the cells that fulfill the following conditions:
They are active

They satisfy the filter criteria applied to the Transmitters folder
Furthermore, if there are transmitters that support the N-frequency mode among the TBC transmitters of your network, the
scrambling code allocation also considers the master and slave carrier allocations.
zone.
660

AT332_TRR_E0

9.8.1.1 Allocation Constraints and Options
The scrambling code and SYNC_DL code allocation algorithm can take into account following constraints:
1. Neighbour relations between cells
You may consider:
Atoll can take into account inter-technology neighbour relations as constraints to

allocate different scrambling codes to the TD-SCDMA neighbours of a GSM
transmitter. In order to consider inter-technology neighbour relations in the
scrambling code allocation, you must make the Transmitters folder of the GSM.atl
document accessible in the TD-SCDMA.atl document. For information on making links
between GSM and TD-SCDMA .atl documents, see the User Manual.
2. The scrambling code reuse distance

Reuse Distance: It is a constraint on the allocation of scrambling codes. The same scarmbling code or SYNC_DL code
cannot be allocated to two sites that are not farther apart than the reuse distance. Scrambling code reuse distance
can be defined for each cell in the cell properties. If this value is not defined, Atoll uses the default reuse distance
defined in the Automatic Scrambling Code and SYNC_DL code Allocation dialog box. The reuse distance constraint is
used for clustered and distributed per cell allocation strategies.
3. The carrier for which you want to perform the automatic allocation
Carrier: You can select "All" or a specific carrier. If you select "All", Atoll allocates the same scrambling code to each
carrier of a transmitter.
4. The number of scrambling codes per SYNC_DL code
Each SYNC_DL code corresponds to a group of scrambling codes as defined in 3GPP specifications. 3GPP specifications
define 32 SYNC_DL codes with 4 corresponding scrambling codes each (SYNC_DL codes are numbered from 0 to 31).
However, it is possible to define a different value (e.g. if you set the number of scrambling codes per SYNC_DL codes
to 2, scrambling codes will be distributed among 64 SYNC_DL codes).
When the allocation is based on a Distributed strategy (Distributed per Cell or Distributed per Site), this parameter can
also be used to define the interval between the scrambling codes assigned to cells on a same site. The defined interval
is applied by adding the following lines in the Atoll.ini file:
[PSC]
ConstantStep=1
For more information about setting options in the Atoll.ini file, see the Administrator Manual.
5. Atoll can use a maximum of codes
Use a Maximum of Codes: If you choose to use a maximum of codes, Atoll will try to spread the allocated spectrum of
scrambling codes as much as possible.
6. Existing allocation
Delete All Codes: If you select this option, Atoll will delete any existing scrambling code allocation and perform a fresh
allocation. Otherwise, Atoll keeps the existing allocation.
9.8.1.2 Allocation Strategies

You can choose from the following four allocation strategies:
Clustered: The purpose of this strategy is to choose for a group of mutually constrained cells, scrambling codes among
a minimum number of clusters. In this case, Atoll will preferentially allocate all the codes within the same cluster.
Distributed per Cell: This strategy consists in using as many clusters as possible. Atoll will preferentially allocate codes
from different clusters.
661

One SYNC_DL code per site: This strategy allocates one cluster, i.e., one SYNC_DL code, per site, then one scrambling
code from the cluster to each cell of the site. When all the clusters have been allocated but there are still sites
remaining, Atoll reuses the clusters as far as possible at another site.
Same carriers must be assigned different scrambling codes. Different carriers of the same
site can be assigned the same scrambling code. Therefore, cells of a transmitter (i.e.
different carriers) are assigned the same scrambling code, if the scrambling code
domains associated with the carriers have a common cluster or enough codes in one
cluster.
Distributed per site: This strategy allocates a group of adjacent clusters, i.e., consecutive SYNC_DL codes, to each site,
then one cluster, or SYNC_DL code, to each transmitter on the site according to its azimuth, and finally one scrambling
code from each cluster to each cell of each transmitter. The number of adjacent clusters, or consecutive SYNC_DL
codes, depends on the number of transmitters per site. When all the sites have been allocated adjacent clusters, and
there are still sites remaining to be allocated, Atoll reuses the adjacent clusters as far as possible at another site.
In the Results table, Atoll only displays scrambling codes and SYNC_DL codes allocated to TBA cells.

TBA cell may be:
"First Order"),
Additional constraints are considered when:
The cell and its near cells are neighbours of a same GSM transmitter (only if the Transmitters folder of the GSM.atl
document is accessible in the TD-SCDMA.atl document),
The neighbour cells cannot share the same cluster (for the "Distributed per site" allocation strategy only).
and the cost of the scrambling code plan. During the allocation, Atoll tries to assign different scrambling codes to the TBA cell
and its near cells. If it respects all the constraints, the cost of the scrambling code plan is 0. When a cell has too many
constraints and there are not anymore scrambling codes available, Atoll breaks the constraint with the lowest cost so as to
generate the scrambling code plan with the lowest cost. For information on the cost generated by each constraint, see "Cell
9.8.1.3.1

Strategies: Clustered and Distributed per Cell
Atoll processes TBA cells according to their priority. It allocates scrambling codes starting with the highest priority cell and its
near cells, and continuing with the lowest priority cells not allocated yet and their near cells. For information on calculating
cell priority, see "Cell Priority" on page 664.
Strategy: One SYNC_DL Code per Site
Atoll assigns a cluster, i.e., a SYNC_DL code, to each site, starting with the highest priority site and its near sites, and continuing
with the lowest priority sites not allocated yet and their near sites. When all the clusters have been allocated but there are
still sites remaining, Atoll reuses the clusters at the other sites. When the Reuse Distance option is selected, the algorithm
reuses the clusters as soon as the reuse distance is exceeded. Otherwise, when the option is not selected, the algorithm tries
to assign reused clusters as spaced out as possible.
Then, Atoll allocates a scrambling code from the cluster to each cell located on the sites (codes belong to the assigned
clusters). It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet
and their near cells.
662

AT332_TRR_E0

Atoll assigns a group of adjacent clusters, i.e., SYNC_DL codes, to each site, starting with the highest priority site and its near
sites, and continuing with the lowest priority sites not allocated yet and their near sites. When all the sites have been allocated
adjacent clusters, and there are still sites remaining to be allocated, Atoll reuses the adjacent clusters at other sites. When
the Reuse Distance option is selected, the algorithm reuses the clusters as soon as the reuse distance is exceeded. Otherwise,
when the option is not selected, the algorithm tries to assign reused clusters as spaced out as possible.
Then, Atoll assigns each cluster of the group to each transmitter of the site according to the transmitter azimuth and selected
neighbourhood constraints (options "Neighbours in Other Clusters" and "Secondary Neighbours in Other Clusters"). Then,
Atoll allocates a scrambling code to each cell located on the transmitters (codes belong to the assigned clusters). It starts with
the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet and their near cells.
For information on calculating site priority, see "Site Priority" on page 667. or information on calculating cell priority, see "Cell
Determination of Groups of Adjacent Clusters
In order to determine the groups of adjacent clusters to be used, Atoll:
Defines theoretical groups of adjacent clusters, independent of the defined domain, considering the 128 scrambling
codes available and 4 codes per cluster.
Starts the distribution of clusters to groups from the cluster 0
Takes into account the maximum number of transmitters per site in order to determine the number of clusters in each
group
Determines the total number of groups
If the number of scrambling codes per cluster is set to 4 and the maximum number of transmitters per site in the network is
3, the theoretical groups of adjacent clusters will be:
Group 1
Group 2
Group 3
Group 4
Cluster 0
Cluster 3
Cluster 6
Cluster 9
Cluster 1
Cluster 4
Cluster 7
Cluster 10
Cluster 2
Cluster 5
Cluster 8
Cluster 11
...
Group 11
Cluster 30
...
Cluster 31
If no domain is assigned to cells, Atoll can use all these groups for the allocation. On the other hand, if a domain is used, Atoll
compares adjacent clusters actually available in the assigned domain with the theoretical groups and only keeps adjacent
clusters common with the theoretical groups.
If we have a domain comprising 12 clusters: clusters 1 to 8 and clusters 12 to 15. In this case, Atoll will use the following groups
of adjacent clusters:
Group 2 with cluster 3, 4 and 5

The clusters 1, 2 and 15 will not be used.

If a domain does not contain any adjacent clusters, Atoll displays a warning message in the Event Viewer.
9.8.1.3.2
In case you have a multi-carrier network and you run the scrambling code allocation on all the carriers, the allocation order
changes. It is no longer based on the cell priority but depends on the transmitter priority. All transmitters which have
constraints with the studied transmitter will be referred to as near transmitters.
In case of a "Per cell" strategy (Clustered and Distributed per cell), Atoll starts scrambling code allocation with the highest
priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their
near transmitters. The same scrambling code is assigned to each cell of the transmitter.
In case of the "One SYNC_DL code per site" strategy, Atoll assigns a cluster, i.e., a SYNC_DL code, to each site and then,
allocates a scrambling code to each transmitter. It starts with the highest priority transmitter and its near transmitters and
continues with the lowest priority transmitters not allocated yet and their near transmitters. The same scrambling code is
assigned to each cell of the transmitter.
In case of the "Distributed per site" strategy, Atoll assigns a group of adjacent clusters, i.e., SYNC_DL codes, to each site, then
a cluster to each transmitter and finally, allocates a scrambling code to each transmitter. It starts with the highest priority
transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near
transmitters. The same scrambling code is assigned to each cell of the transmitter.
For information on calculating transmitter priority, see "Transmitter Priority" on page 666.
663

alphanumeric order.

9.8.1.4.1
Cell Priority
Scrambling code allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities
assigned to cells depend upon how much constrained each cell is and the cost defined for each constraint. A cell without any
constraint has a default cost, C , equal to 0. The higher the cost on a cell, the higher the priority it has for the scrambling code
allocation process.
There are seven criteria employed to determine the cell priority. The total cost due to constraints on any cell i is defined as:
With
C i U = C i Dist + C i EP + C i N + C i N 2G + C i Cluster + C i CN
All the cost components are described below:
Scrambling Code Domain Criterion
The cost due to the domain constraint, C i Dom , depends on the number of scrambling codes available for the allocation.
The domain constraint is mandatory and cannot be broken.
When no domain is assigned to cells, 128 scrambling codes are available and we have:
C i Dom = 0
When domains of scrambling codes are assigned to cells, each unavailable scrambling code generates a cost. The higher the
number of codes available in the domain, the less will be the cost due to this criterion. The cost is given as:
C i Dom = 128 Number of scrambling codes in the domain
Distance Criterion
Cj Dist i
C i Dist =
Where
C i EP =
cEP i j
j
Where
664

AT332_TRR_E0

Ci N =
j

C j N1 i = I j c N1
Where
Because two first order neighbours must not have the same scrambling code, Atoll considers the cost created by two first
C j N1 i + C j N1 i
C j j N1 i = ----------------------------------------------------2
Where
Because two second order neighbours must not have the same scrambling code, Atoll considers the cost created by two
second order neighbours to be each other.
C k N2 i + C k N2 i
C k k N2 i = ------------------------------------------------------2
C l N3 i = Max j
c N3
Where
Because two third order neighbours must not have the same scrambling code, Atoll considers the cost created by two third
C l N3 i + C l N3 i
C l l N3 i = ----------------------------------------------------2
665


And
Close Neighbour Criterion
The constraint level of any cell i depends on the number of its close neighbour cells j. The close neighbour cost ( C i CN )
depends on two components: the importance of the neighbour relation ( I i j ) and the distance ( d i j ) relative to maximum
Max
close neighbour distance ( d CN ).
C i CN =
di j
I + 1 ----------
ij
Max
d CN
----------------------------------------- c CN
2
Where
c CN is the cost of the close neighbour constraint. This value can be defined in the Constraint Cost dialog box.
GSM Neighbour Criterion
This criterion is considered when the co-planning mode is activated (i.e. the Transmitters folder of the GSM .atl document is
made accessible in the TD-SCDMA.atl document) and inter-technology neighbours have been allocated. If the cell i is
neighbour of a GSM transmitter, the cell constraint level depends on how many cells j are neighbours of the same GSM
transmitter. The total cost due to GSM neighbour constraint is given as:
C i N 2G =
cN2G j Tx2G
j
Where
cN
2G
is the cost of the GSM neighbour constraint. This value can be defined in the Constraint Cost dialog box.
Cluster Criterion
When the "Distributed per Site" allocation strategy is used, you can consider additional constraints on allocated clusters (one
cell, its first order neighbours and its second order neighbours must be assigned scrambling codes from different clusters). In
this case, the constraint level of any cell i depends on the number of first and second order neighbours, j and k. The total cost
due to the cluster constraint is given as:
C i Cluster =
Cj N1 i cCluster + Ck N2 i cCluster
j
Where
c Cluster is the cost of the cluster constraint. This value can be defined in the Constraint Cost dialog box.
9.8.1.4.2
In case you have a multi-carrier network and you run scrambling code allocation on "all" the carriers, Atoll allots priorities to
transmitters. Priorities assigned to transmitters depend on how much constrained each transmitter is and the cost defined
for each constraint. The higher the cost on a transmitter, the higher the priority it has for the scrambling code allocation
process.
666

AT332_TRR_E0
With C Tx U =

Tx
i Tx i
9.8.1.4.3
Site Priority
In case of "Per Site" allocation strategies (One SYNC_DL code per Site and Distributed per Site), Atoll allots priorities to sites.
Priorities assigned to sites depend on how much constrained each site is and the cost defined for each constraint. The higher
the cost on a site, the higher the priority it has for the scrambling code allocation process.
Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. The cost due to constraints
on the site is given as:
With C S U =

S
Tx S Tx
Here, the domain considered for the site is the intersection of domains available for transmitters of the site. The domain
9.8.2 Scrambling Code Allocation Example

9.8.2.1 Single Carrier Network
In order to understand the differences between the different allocation strategies and the behaviour of algorithm when using
a maximum of codes or not, let us consider the following sample scenario:
Figure 9.15: Scrambling Code Allocation Example

Let Site0, Site1, Site2, and Site3 be four sites, with 3 transmitters each using carrier 0, to whom scrambling codes have to be
allocated out of 6 clusters of 4 scrambling codes. This implies that the domain of scrambling codes for the four sites is from 0
to 23 (cluster 0 to cluster 5). The reuse distance is supposed to be less than the inter-site distance. Only co-site neighbours
exist.
The following section shows the results of each combination of options with explanations where necessary.
667

9.8.2.1.1
Strategy: Clustered
Since the restrictions of neighbourhood only apply to co-sites and, in our case, the distances between sites are greater than
the reuse distance, every cell has the same priority. Allocation is performed in an alphanumeric order.
Without "Use a Maximum of Code"
With "Use a Maximum of Code"
As it is possible to use a maximum of codes, Atoll starts

allocation at the start of a different cluster at each site.
Atoll starts allocating the codes from the start of cluster 0 at
When a cluster is reused, and there are non allocated codes
each site.
left in the cluster, Atoll first allocates those codes before
reusing the already used ones.
9.8.2.1.2
Strategy: Distributed per Cell

Since the restrictions of neighbourhood only apply to co-sites and, in our case, the distances between sites are greater than
the reuse distance, every cell has the same priority. Allocation is performed in an alphanumeric order.
668
Atoll allocates codes from different clusters to each cell of

the same site. Under given constraints of neighbourhood
and reuse distance, same codes can be allocated to each
sites cells.
Atoll allocates codes from different clusters to each sites

cells. As it is possible to use a maximum of codes, Atoll
allocates the codes so that there is least repetition of codes.

AT332_TRR_E0
9.8.2.1.3
Strategy: One SYNC_DL Code per Site

Since the restrictions of neighbourhood only apply to co-sites, therefore, every site has the same priority. Cluster allocation
to sites is performed in an alphanumeric order.
9.8.2.1.4
In this strategy, a cluster of codes is limited to be used at just

one site at a time unless all codes and clusters have been
allocated and there are still sites remaining to be allocated.
In this case, Atoll reuses the cluster as far as possible at
another site.
When it is possible to use a maximum of codes, Atoll can

allocate different codes from a reused cluster at another
site.

Since the restrictions of neighbourhood only apply to co-sites, therefore, every site has the same priority. Cluster allocation
to sites is performed in an alphanumeric order.
A group of adjacent clusters is allocated to one site at a time,

unless all the codes and groups of adjacent clusters have
been allocated but there are still sites remaining to be
allocated. In this case (here only one group of adjacent
clusters 0, 1, and 2 is available), Atoll reuses the group as far
as possible at another site.
When it is possible to use a maximum of codes, Atoll can

allocate different codes from a reused groups of adjacent
clusters at another site.
9.8.2.2 Multi Carrier Network

If you have a multi carrier network, i.e., transmitters with more than one cells using different carriers, and you run scrambling
code allocation on "all" the carriers, Atoll allocates the same scrambling code to each carrier of a transmitter.
669

Let Site0, Site1, Site2, and Site3 be four sites with 3 cells using carrier 0 and 3 cells using carrier 1. Scrambling codes have to
be allocated out of 6 clusters consisted of 4 scrambling codes. This implies that the domain of scrambling codes for the four
sites is from 0 to 23 (cluster 0 to cluster 5). The reuse distance is supposed to be less than the inter-site distance. Only co-site
neighbours exist. Every site has the same priority and the cluster allocation to sites is performed in an alphanumeric order.
Atoll allocates one cluster at each site and then, one code to each transmitter. Then, the same code is given to each cell of
the transmitter.
Figure 9.16: Scrambling Code Allocation to All Carriers
9.9 Automatic GSM/TD-SCDMA Neighbour Allocation

It is possible to automatically calculate and allocate neighbours between GSM and TD-SCDMA networks. In Atoll, it is called
inter-technology neighbour allocation.
When the TD-SCDMA coverage is not continuous. In this case, the TD-SCDMA coverage is extended by TD-SCDMA to
GSM handovers.
In order to balance traffic and service distribution between both networks.
Atolls automatic inter-technology neighbour allocation algorithm takes into account both cases.
An .atl document containing the GSM network, GSM.atl, and another one containing the TD-SCDMA network, TDSCDMA.atl,
An existing link on the Transmitters folder of GSM.atl into TD-SCDMA.atl.
transmitters of GSM.atl are potential neighbours. The TD-SCDMA cells, in TD-SCDMA.atl, to be allocated neighbours are called
TBA cells which fulfill following conditions:
They are active

They satisfy the filter criteria applied to Transmitters folder
one of its subfolders.
Only TD-SCDMA TBA cells can be assigned neighbours.

The inter-transmitter distance

The maximum number of neighbours
Allocation options
The selected allocation strategy
We assume we have a TD-SCDMA reference cell, A, and a GSM candidate neighbour transmitter, B.
670

AT332_TRR_E0

1. The distance between the TD-SCDMA reference cell and the GSM neighbour must be less than the user-defined
maximum inter-site distance. If the distance between the TD-SCDMA reference cell and the GSM neighbour is greater
than this value, then the candidate neighbour is discarded.
azimuths of antennas. For information on the effective distance calculation, see "Appendix: Calculation of the InterTransmitter Distance" on page 660.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one
or more carriers. Atoll will allocate neighbours to cells using the selected carriers.
Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site
than the reference TD-SCDMA cell in the candidate neighbour list. This option is automatically selected.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you
may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.
neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by
decreasing importance in order to keep the ones with high importance. If the maximum number of neighbours to be
allocated to each cell is exceeded, Atoll keeps the ones with high importance.
As indicated in the table below, the neighbour importance depends on the distance and on the neighbourhood cause;
this value varies between 0 to 100%.
Neighbourhood cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
Co-site transmitter
100 %

d1 ---------d max
Where d is the effective distance between the TD-SCDMA reference cell and the GSM neighbour and d max is the
maximum inter-site distance. For information on the effective distance calculation, see "Appendix: Calculation of the
Inter-Transmitter Distance" on page 660.
as existing.

1. The distance between the TD-SCDMA reference cell and the GSM neighbour must be less than the user-defined
maximum inter-site distance. If the distance between the TD-SCDMA reference cell and the GSM neighbour is greater
than this value, then the candidate neighbour is discarded.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one
or more carriers. Atoll will allocate neighbours to cells using the selected carriers.
Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site
than the reference TD-SCDMA cell in the candidate neighbour list. This option is automatically selected.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you
may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.
neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
671

3. There must be an overlapping zone ( S A S B ) with a given cell edge coverage probability.
Two different cases may be considered for SA:
1st case: SA is the area where the cell A is the best serving cell of the TD-SCDMA network.
The pilot signal received from A is greater than the minimum pilot signal level and is the highest one.
The margin is set to 0 dB.
2nd case: The margin is different from 0 dB and SA is the area where:
The pilot signal level received from A exceeds the user-defined minimum pilot signal level and is within a
margin from the highest signal level.
1st case: SB is the area where the cell B is the best serving transmitter of the GSM network.
In this case, the margin must be set to 0 dB.
highest one.
2nd case: The margin is different from 0 dB and SB is the area where:
SA SB
Atoll calculates the percentage of covered area ( ------------------ 100 ) and compares this value to the % minimum covered
SA
area. If this percentage is less than the minimum, the candidate neighbour B is discarded.
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of
covered area.
Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by
decreasing importance in order to keep the ones with high importance. If the maximum number of neighbours to be
allocated to each cell is exceeded, Atoll keeps the ones with high importance.
As indicated in the table below, the neighbour importance depends on the distance and on the neighbourhood cause;
this value varies between 0 to 100%.
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
Co-site transmitter
IF

IF
transmitter.
d Di = 1 ---------d max
d is the distance between the TD-SCDMA reference cell and the GSM neighbour and d max is the maximum inter-site
distance.

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%
672

AT332_TRR_E0
cause
IF

No
10%+50%{10%(Di)+90%(O)}
Yes
60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.

9.9.1.3 Appendices
9.9.1.3.1
Delete Existing Neighbours Option

selected. If a new TBA cell i is created in TD-SCDMA.atl, you can run the automatic allocation with the Delete existing
neighbours option not selected, in order to allocate neighbours to the new cell i only.
If you change some allocation criteria (e.g., increase the maximum number of neighbours or create a new GSM TBC
transmitter) and start a new allocation without selecting the Delete existing neighbours option, Atoll examines the neighbour
list of the TBA cells and checks allocation criteria only if there is still space left in their neighbour lists. A new GSM TBC
transmitter can enter the TBA cell neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list.
673

674
Chapter 10
WiMAX BWA Networks

676

Chapter 10: WiMAX BWA Networks
AT332_TRR_E0
10 WiMAX BWA Networks

This chapter describes all the calculations performed in Atoll WiMAX documents. The first part of this chapter lists all the input
parameters in the WiMAX BWA documents, their significance, location in the Atoll GUI, and their usage. It also contains the
lists of the formulas used for the calculations.
signal levels, noise, and interference for downlink and uplink considering the effects of smart antennas, power control,
subchannelisation, MIMO etc., and the radio resource management algorithms used by the different available schedulers.
If you are new to WiMAX, you can also see the Glossary of WiMAX Terms in the User Manual for information on WiMAX terms
and concepts, especially in the context of their user in Atoll.
All the calculation algorithms in this section are described for two types of cells.
cells.
All the calculation algorithms in this section are described for two types of receivers.

10.1 Definitions
Name
Value
Unit
Description
1.38 x 10-23
J/K
Boltzmanns constant
290
Ambient temperature
n0
dBm/Hz

noise
D Frame
Global parameter
ms
Frame Duration
Choice List: 2, 2.5, 4, 5, 8, 10, 12.5, 20
r CP
None
Cyclic Prefix Ratio

Choice List: 1/4, 1/8, 1/16, 1/32
O Fixed
DL
Global parameter
SD
Fixed time-domain overhead (DL)
O Fixed
UL
Global parameter
SD
Fixed time-domain overhead (UL)
DL
Global parameter
Variable time-domain overhead (DL)
UL
Global parameter
Variable time-domain overhead (UL)
TDD
Global parameter
Ratio of the DL subframe to the

entire frame (TDD only)
O Variable
O Variable
r DL-Frame
677

Name
Value
Unit
Description
N SD DL
TDD
Global parameter
None
Number of symbol durations per

frame that corresponds to the DL
subframe (TDD only)
TDD
Global parameter
None
Number of symbol durations per

frame that corresponds to the UL
subframe (TDD only)
D TTG
TDD
Global parameter
ms
Transmit Time Guard (TDD only)
D RTG
TDD
Global parameter
ms
Receive Time Guard (TDD only)
M PC
Global parameter
dB
Uplink power control margin
CNR Min
Global parametera
dB

PZ
Permutation zone parameter
None
Number of subchannels per channel

in UL subframe
N SC DL
PZ
None
Number of subchannels per channel

in DL subframe
N SCa Total
None
Total number of subcarriers per

channel (FFT size)
N SCa Preamble
None
Number of subcarriers used by the

preamble
PZ
None
Number of used subcarriers per

channel
N SCa Data
PZ
None
Number of subcarriers per channel

used for data transfer
N SCa DC
None
Number of DC subcarriers per

channel
None
Number of pilot subcarriers per

channel
None
Number of guard subcarriers per

channel
N SD UL
N SC UL
N SCa Used
PZ
N SCa Pilot
PZ
PZ
PZ
Calculation result ( N SCa Pilot = N SCa Used N SCa Data )

Calculation result
PZ
N SCa Guard
PZ
( N SCa Guard
PZ
= N SCa Total N SCa Used N SCa DC )
PZ UL
None
Uplink permutation zone
PZ DL
None
Downlink permutation zone
QT PZ
dB
Quality threshold: Required

preamble C/N or C/(I+N) for
accessing a zone
Speed Max PZ
Km/hr
Speed limit for mobiles trying to

access a permutation zone
d Max PZ
Maximum distance from the

transmitter covered by a zone
p PZ
None
Permutation zone priority
W Channel
MHz
Channel bandwidth
First
None

frequency band
N Channel
Last
None

frequency band
F Start FB TDD
MHz
Start frequency of the TDD frequency

band
F Start FB FDD DL
MHz
DL Start frequency of the FDD

frequency band
F Start FB FDD UL
MHz
UL Start frequency of the FDD

frequency band
N Channel
678

AT332_TRR_E0
Name
Value
Unit
Description
f Sampling
None
Sampling factor
f ACS FB
dB
Adjacent Channel Suppression Factor
ICS FB
MHz
CN FB
None
Channel number step
Inter Tech
Network parameter
dB
Inter-technology interference
reduction factor
Bearer parameter
None
Bearer index
Mod B
Bearer parameter
None
CR B
Bearer parameter
None
Bearer parameter
bits/
symbol
Bearer Efficiency
TB
Bearer parameter
dB
TP BH DL
Site
Site parameter
kbps
Maximum backhaul site downlink

throughput
Site
Site parameter
kbps
Maximum backhaul site uplink

throughput

dB
N Ant TX
None
Number of antennas used for MIMO

in transmission
N Ant RX
None

in reception
TX
Antenna parameter
dB
TX

dB
Transmitter loss
TX
None
Array
dB
Array gain offset
Combining
dB
Power combining gain offset
G SA
dB
Diversity gain (cross-polarisation)
N Channel
Cell parameter
None
P Preamble
Cell parameter
dBm
Preamble power
dB
Traffic power reduction
dB
Pilot power reduction
dB
Idle pilot power reduction
f IRF
TP BH UL
nf
G
L
TX
E SA
G SA
G SA
Div
Cell parameter P Traffic = P Preamble P Traffic in dB

P Traffic
Ratio
P Traffic
= 10
P Traffic
------------------------10
in %
Cell parameter P Pilot = P Preamble P Pilot in dB

P Pilot
Ratio
P Pilot
= 10
P Pilot
-------------------10
in %
Cell parameter P Idle Pilot = P Preamble P Idle Pilot in dB

P Idle Pilot
Ratio
P Idle Pilot
= 10
P Idle Pilot
-----------------------------------10
in %
TL DL
Cell parameter
TL UL
Cell parameter
Uplink traffic load
679

Name
Value
Unit
Description
TL DL Max
Cell parameter
TL UL Max
Cell parameter
NR UL
Cell parameter
dB
Uplink noise rise
NRUL Seg
Cell parameter
dB
Segmented zone uplink noise rise
N Users Max
Cell parameter
None
N Users DL
Cell parameter
None

cell in downlink
N Users UL
Cell parameter
None

cell in uplink
SU DL
Cell parameter
Downlink segmentation usage ratio
AU DL
Cell parameter
Downlink AAS usage ratio
T AMS
Cell parameter
dB
Adaptive MIMO switch threshold
T MU MIMO
Cell parameter
dB
Multi-user MIMO threshold
PI
Cell parameter
None
Preamble index
T Preamble
Cell parameter
dB
Preamble C/N threshold
D Reuse
Cell parameter
Channel and preamble index reuse

distance
G MU MIMO
Cell parameter
None
Uplink MU-MIMO gain
Inter Tech
Cell parameter
dB
Inter Tech
Cell parameter
dB
ZPBDL
Cell parameter
None
Downlink zone permbase
ZPB UL
Cell parameter
None
Uplink zone permbase
TX i ic
None
Downlink multi-user diversity gain

(MUG)
TX i ic
None
Uplink multi-user diversity gain

(MUG)
dB
Maximum C/(I+N) above which no

MUG gain is applied
G SU MIMO
Max
Cell WiMAX equipment parameter
None
Maximum SU-MIMO gain
G Div UL
Cell WiMAX equipment parameter
dB
Uplink STTD/MRC, SU-MIMO or MUMIMO diversity gain
f Bias
QoS
Scheduler parameter
QoS class bias factor
QoS
Service parameter
None
QoS class of the service
Service parameter
None
Service priority
B DL Highest
Service parameter
None

the downlink
B UL Highest
Service parameter
None

the uplink
B DL Lowest
Service parameter
None

the downlink
B UL Lowest
Service parameter
None

the uplink
UL
Service parameter
NR DL
NR UL
G MUG DL
G MUG UL
Max
CINR MUG
f Act
680

AT332_TRR_E0
Name
Value
Unit
Description
f Act
DL
Service parameter
TPD Min UL
Service parameter
kbps

uplink
TPD Min DL
Service parameter
kbps

downlink
TPD Max UL
Service parameter
kbps

uplink
TPD Max DL
Service parameter
kbps

downlink
UL
Service parameter
kbps

uplink
TP Average
DL
Service parameter
kbps

downlink
TP Offset
Service parameter
kbps
Throughput offset
f TP Scaling
Service parameter
Scaling factor
L Body
Service parameter
dB
Body loss
N SC UL
Min
Service parameter
None
Minimum number of subchannels
P Min
Terminal parameter
dBm
P Max
Terminal parameter
dBm
nf
Terminal parameter
dB
Terminal parameter
dB
Terminal parameter
dB
Terminal loss
N Ant TX
Terminal parameter
None

in transmission
N Ant RX
Terminal parameter
None

in reception
G SU MIMO
Max
Terminal WiMAX equipment parameter
None
G Div DL
dB
Downlink STTD/MRC or SU-MIMO

diversity gain
G Div Preamble
dB
Preamble diversity gain
UL
Clutter parameter
dB
G Div
DL
Clutter parameter
dB
f SU MIMO
Clutter parameter
None
SU-MIMO gain factor
L Indoor
Clutter parameter
dB
Indoor loss
L Path
dB
Path loss
TP Average
G Div
ICP DL
Network parameter
None

M Shadowing Model

standard deviation
dB
M Shadowing C I

dB
681

a.


Name
TX i ic
F Start
Value
TX i ic
TX ic
TX i ic
TX i ic
TX jc
TX i ic TX j jc
TX ic
MHz

MHz

MHz
None
MHz

None

overlap ratio
MHz

None

overlap ratio
None
None
FDD TDD overlap ratio
None
Total overlap ratio
TX i ic TX j jc
TX i ic TX j jc
TX jc
TX ic
TX jc
TX ic
TX ic
j
i
j
i
i
TX i ic TX j jc
W ACO
L
---------------------------------TX i ic
W Channel
TX i ic TX j jc
L
TX i ic TX j jc
TX jc
j
TX ic
i
Min F End F End
TX ic
TX jc
TX ic
i
j
i
TX i ic TX j jc
W ACO
H
---------------------------------TX i ic
W Channel
TX i ic TX j jc
r ACO
TX ic
W CCO
----------------------------------TX i ic
W Channel
r CCO
W ACO
TX jc
j
i
j
i
TX i ic TX j jc
r ACO
TX i ic
F Start + W Channel
W CCO
Description
TX ic
i
i
F Start FB + W Channel + ICS FB
F End
W ACO
TXi ic N First TXi ic

Channel
Channel
N
-------------------------------------------------------
TX ic
CN FB
Unit
TX i ic TX j jc
TX i ic TX j jc
r ACO
r ACO
TX i ic TX j jc
+ r ACO
TDD
TX i ic TX j jc
r FDD TDD
r DL Frame
----------------------- if interferer uses a TDD frequency band and victim uses
100
an FDD frequency band, 1 otherwise
TX ic
f
ACS FB-
--------------------------TX ic TX j jc
10
i
j
i
j
r
r i
+ r ACO
10
FDD TDD
CCO
TX i ic TX j jc
rO
TX i ic
TX j jc
TX ic
f ACS FB
TX ic
TX ic TX jc TX ic TX jc --------------------------- TX ic TX jc W i
10
j
i
j
i
j
Channel
r i
--------------------+ r ACO
10
TX jc
CCO
r FDD TDD
j
W Channel
TX i ic
TX j jc
682

AT332_TRR_E0
10.2.2 Preamble Signal Level Calculation

Name
TX ic
i
C Preamble
Value
Unit
Description
dBm
Received preamble signal level
dBm
Preamble EIRP of a cell
dB
Path loss
dB
Total losses
Value
Unit
Description
TX i ic
TX ic
N SCa Preamble Preamble
i
--------------------------------- f Segment
n 0 + 10 Log F Sampling
TX i ic
N SCa Total
dBm
Preamble thermal noise for a cell
1
--3
None
Preamble segmenting factor
dBm
Preamble noise for a cell
Unit
Description
dBm
Total interference generated by an

interfering cell
dB

dB

segmentation
None
Preamble subcarrier collision

probability
interference
Unit
Description
dB
Preamble C/N for a cell
TX ic
i
EIRP Preamble L Path M Shadowing Model L Indoor + G

L
Mi
Mi
L Ant L Body
TX i ic
TX i ic
TX i
TX i

TX ic
i
P Preamble
+G
TX
TX
TX
+ 10 Log E SA +
TX
L Path
L Total
Mi
Without smart antennas: P Preamble + G

EIRP Preamble
Combining
G SA
Div
G SA
L Model + L Ant
L Path + L
Mi
TX i
TX i
+L
Mi
Mi
Mi
+ L Ant + L Body
10.2.3 Preamble Noise Calculation

Name
TX i ic
n 0 Preamble
Preamble
f Segment
TX i ic
TX i ic
n 0 Preamble + nf
n Preamble
Mi
10.2.4 Preamble Interference Calculation

Name
TX j jc
I Preamble
Value
TX j jc
TX i ic TX j jc
C Preamble + f O
TX i ic TX j jc
TX i ic TX j jc
TX i ic TX j jc
10 Log r O
TX ic TX jc
i
j
10 Log p Collision
fO
TX i ic TX j jc
TX i ic TX j jc
f Seg Preamble
p Collision
Inter Tech
+ f Seg Preamble + I DL
TX i ic
1 if N Seg
TX j jc
= N Seg
TX i ic
and 0 if N Seg
TX j jc
N Seg
TX
k
P DL Rec
--------------------------------------
F TX i ic TX k
TX k ICP DL
Inter Tech
I DL
10.2.5 Preamble C/N Calculation

Name
TX i ic
CNR Preamble
Value
TX i ic
TX i ic
Mi
DL
C Preamble n Preamble + G Div Preamble + G Div
683

10.2.6 Preamble C/(I+N) Calculation

Name
Value
Unit
Description
TX ic
TX i ic
CINR Preamble
TXj jc
n
Preamble
IPreamble
-----------------------------
TX i ic
Inter Tech
10
---------------------------- Inter Tech
C Preamble 10 Log
+ 10
+ NR DL
10
I
10
DL
dB
All TXj jc
Preamble C/(I+N) for a cell
Mi
DL
+ G Div Preamble + G Div

TX ic
TX ic
i
I + N Preamble
TXj jc
n Preamble
IPreamble
-----------------------------
Inter Tech
10
-
-------------------------- + NR Inter Tech
dBm
10 Log
+I
+ 10
10
DL
10
DL
All TXj jc
Preamble Total Noise (I+N) for a cell
10.2.7 Traffic and Pilot Signal Level Calculation (DL)

Name
TX i ic
C Traffic
TX i ic
C Pilot
TX i ic
EIRP Traffic
TX i ic
EIRP Pilot
TX i ic
P Traffic
TX i ic
P Pilot
Value
TX i ic
EIRP Traffic L Path M Shadowing Model L Indoor + G

Mi
Mi
dBm
Received traffic signal level
dBm
Received pilot signal level
Mi
Mi
TX i ic
Mi
Description
L Ant L Body
EIRP Pilot L Path M Shadowing Model L Indoor + G

L
Unit
Mi
Mi
Mi
L Ant L Body
TX i ic
TX i
TX i ic
TX i
P Traffic + G
P Pilot + G
Array
+ G SA
Array
+ G SA
+ G SA
+ G SA
Div
TX i
dBm
Traffic EIRP of a cell
Div
TX i
dBm
Pilot EIRP of a cell
Combining
+ G SA L
Combining
+ G SA L
TX i ic
TX i ic
dBm
Traffic transmission power of a cell
TX i ic
TX i ic
dBm
Pilot transmission power of a cell
P Preamble P Traffic
P Preamble P Pilot
10.2.8 Traffic and Pilot Noise Calculation (DL)

Name
Value
Unit
Description
dBm
Thermal noise for a cell
None
Downlink segmenting factor
dBm
Downlink noise for a cell
Mi
TX i ic
n 0 DL
PZ DL
TXi ic
N SCa Used
n 0 + 10 Log F Sampling ------------------------
TX i ic
N
SCa Total
With Segmentation:
Mi
PZ DL
TXi ic
N SCa Used
n 0 + 10 Log F Sampling ------------------------ f Segment DL

TX i ic
N SCa Total
f Segment DL
TX i ic
n DL
684
3 PSG + 2 SSG
1 without and --------------------------------------------- with downlink segmentation
15
TX i ic
n 0 DL + nf
Mi

AT332_TRR_E0
10.2.9 Traffic and Pilot Interference Calculation (DL)

Name
Value
TX jc
Unit
Description
dBm
Total interference generated by an

interfering cell
TX jc
j
I j
I
Non AAS
Idle
----------------------------------------------
10
10
Monte Carlo Simulations: 10 Log 10

+ 10
TX jc
I j
AAS
-----------------10
without smart antennas, or 10 Log 10
with smart
TX j jc
I Total
antennas
TX jc
TX jc
TX jc
j
j
I j
I Idle
I AAS
Non AAS
----------------------------------------------------------------
10
10
10
Coverage Predictions: 10 Log 10
+ 10
+ 10
Monte Carlo Simulations:

TX j jc

TX j jc
Mi
Mi
Mi
Mi
L Ant L Body
Coverage Predictions:
I Traffic
dBm
TX j jc
Traffic interference power of an

interfering cell
EIRP Traffic L Path M Shadowing Model + M Shadowing C I L Indoor

+G
Mi
Mi
Mi
Mi
L Ant L Body
TX jc
j
EIRP Pilot
TX j jc
L Ant L Body
dBm
Pilot interfering power of an

interfering cell
TX j
dBm
Traffic EIRP of an interfering cell
TX j
dBm
Pilot EIRP of an interfering cell
dBm
Interference from the loaded part of

the frame transmitted using the
transmitter antenna of an interfering
cell
I Pilot
TX jc
j
EIRP Pilot L Path M Shadowing Model + M Shadowing C I L Indoor

+G
Mi
Mi
Mi
Mi
L Ant L Body
TX j jc
TX j jc
TX j
TX j jc
TX j
P Traffic + G
EIRP Traffic
TX j jc
P Pilot + G
EIRP Pilot
L
L
TX jc
TX j jc
I Non AAS
I j
TX j jc
Traffic
------------------ TX jc
TX j jc
N SCa Data
j
10
-+
10 Log TL DL
----------------------- 1 AU DL
10
TX j jc
N SCa Used
10
TX j jc
I Pilot
------------------10
TX j jc
N SCa Data
1 ------------------------
TX jc
j
N SCa Used
TX j jc
EIRP AAS
TX j jc
TX j jc
EIRP AAS
+G
TX jc
j
Mi
Mi
Mi
Mi
L Ant L Body
I AAS
EIRP AAS
Mi
dBm
Interference power of an interfering

cell transmitted using smart antenna
dBm
Traffic EIRP of an interfering cell

using smart antenna
L Path M Shadowing Model + M Shadowing C I L Indoor
Mi
Mi
Mi
L Ant L Body
TX jc
j
P Traffic + G
TX
TX
685

Name
Value
TX jc
j
I Idle Pilot
TX jc
j
EIRP Idle Pilot
TX j jc
L Path L Indoor + G
TX j jc
P Idle Pilot + G
EIRP Idle Pilot
TX j
i
L Ant
i
L Body
TX j
Unit
Description
dBm
Interference from empty part of the

frame transmitted using the
cell
dBm
Idle pilot EIRP of an interfering cell
dBm
Interference from the empty part of

the frame transmitted using the
cell
dB

TX jc
TX j jc
I Idle
I j
TX j jc
Idle Pilot-
----------------------------TX jc
N
10
j
SCa Data
10 Log 1 TL DL 10
1 ------------------------
TX j jc
N SCa Used
TX i ic TX j jc
TX i ic TX j jc
10 Log r O
TX ic TX jc
i
j
i
j
10 Log p Collision DL
dB

downlink segmentation
TX k
P DL Rec
--------------------------------------
F TX i ic TX k
TX ICP DL
k
interference
Unit
Description
dB
Traffic C/N for a cell
dB
Pilot C/N for a cell
fO
TX ic TX jc
f Seg DL
Inter Tech
I DL
10.2.10 Traffic and Pilot C/N Calculation (DL)

Name
Value
TX i ic
TX i ic
C Traffic n DL
TX i ic
CNR Traffic
TX i ic
Mi
DL
With MIMO: CNR Traffic + G Div DL + G Div

TX i ic
TX i ic
C Pilot n DL
TX i ic
CNR Pilot
TX i ic
Mi
With MIMO: CNR Pilot + G Div DL +
DL
G Div
10.2.11 Traffic and Pilot C/(I+N) Calculation (DL)

Name
Value
TXj jc
IDL
-----------------C Traffic 10 Log

10
10
All TX j jc
TX i ic
TX i ic
CINR Traffic
TX i ic
Unit
TX i ic
n DL
+ I Inter Tech + -------------------10
DL
10
Mi
+ NR Inter Tech
DL
dB
Description
Traffic C/(I+N) for a cell
DL
With MIMO: CINR Traffic + G Div DL + G Div

TX jc
TX i ic
TX i ic
CINR Pilot
C Pilot
I j
DL
-----------------10
10
10 Log
All TXj jc
TX ic
i
TX ic
n DL
---------------------
Inter
Tech
Inter
Tech
10
+I
+ NR
+10
DL
DL
dB
Pilot C/(I+N) for a cell
DL
With MIMO: CINR Pilot + G Div DL + G Div

TX jc
TX i ic
I + N DL
686
TX ic
I j
n DL
DL
------------------
---------------------
10 Log
DL
DL
All TXj jc
dBm
Traffic Total Noise (I+N) for a cell

AT332_TRR_E0
10.2.12 Traffic Signal Level Calculation (UL)

Name
Value
M
i
C UL
Unit
Description
dBm
Received uplink signal level
dBm
Uplink EIRP of a user equipment
Unit
Description
dBm
dBm
Uplink noise for a cell
Unit
Description
dBm
Uplink interference received at a cell
dB

dB

load
i
j
10 Log p Collision UL
db

uplink segmentation
SC Com
-----------------TX i ic
SC
None
Uplink segmentation collision

probability
EIRP UL L Path M Shadowing Model L Indoor + G

L
TX i
Mi
Mi
With P
Mi
Mi
L Ant L Body
P
EIRP UL
TX
Mi
+G
Mi
Mi
Mi
= P Max without power control and P
Mi
Mi
= P Eff after
power control
10.2.13 Traffic Noise Calculation (UL)

Name
Value
M
TX i ic
n 0 UL
PZ UL
TXi ic
N SCa Used
n 0 + 10 Log F Sampling ------------------------
TX i ic
N
SCa Total
TX i ic
TX i ic
n 0 UL + nf
n UL
TX i ic
10.2.14 Traffic Interference Calculation (UL)

Name
M
I UL
Value
M
TX ic TX jc
i
j
C UL + f O
TX ic TX jc
i
j
+ f TL UL + f Seg UL
TX i ic TX j jc
TX i ic TX j jc
10 Log r O
fO
j
10 Log TL UL
Mj
f TL UL
TX ic TX jc
TX ic TX jc
i
j
f Seg UL
TX i ic TX j jc
p Collision UL
TX i ic
NR UL
TX i ic
NRUL Seg
TX i ic
I + N UL
TX i ic
IMj
n UL
UL
non-seg M i
---------------------
TX ic
10
------------------------------------------ + NR Inter Tech n i dB
10
10 Log
10
+ 10
UL
UL
All M j
All TX j jc
TX i ic
IMj
n UL
UL
seg M i
---------------------
TX i ic
Inter Tech
10
--------------------------------10 Log
+ NR UL
n UL
10
10
+ 10
All M j
All
TX
jc
TX i ic
NR UL
TX i ic
+ n UL
TX i ic
TX i ic
or NR UL Seg + n UL
dB
Segmented zone uplink noise at a

cell without smart antennas
dBm
Total Noise (I+N) for a cell
dB
Uplink noise at a cell with smart

antenna
NR UL
I UL + n I
--------------------------------2
n I
Non-segmented zone uplink noise at

a cell without smart antennas
687

Name
TX ic
i
I + N UL
Value
Unit
Description
dBm
Total Noise (I+N) for a cell in case of

smart antennas
Unit
Description
dB
Uplink C/N at a cell
Unit
Description
dB
Uplink C/(I+N) at a cell
Unit
Description
Hz
Sampling frequency
F Sampling 10
-------------------------------------TX i ic
N SCa Total
kHz
Inter-subcarrier distance
1
------------------TX i ic
F
ms
Useful symbol duration
ms
ms
Symbol duration
D Frame D TTG D RTG
ms
Used frame duration
D Used
Frame
Floor ---------------- TXi ic
D Symbol
SD
Frame duration in terms of symbol

durations
SD
Downlink subframe duration in

terms of symbol durations
I UL + n I
10.2.15 Traffic C/N Calculation (UL)

Name
Value
TX i ic
Mi
C UL n UL
Mi
CNR UL
Mi
TX i ic
With MIMO: CNR UL + G Div UL +
UL
G Div
10.2.16 Traffic C/(I+N) Calculation (UL)

Name
Value
TX i ic
Mi
Without smart antennas: CNR UL NR UL

Mi
CINR UL
Mi
TX i ic
Mi
With smart antennas: CNR UL NR UL

Mi
TX i ic
or CNR UL NR UL Seg
TX i ic
UL
With MIMO: CINR UL + G Div UL + G Div
10.2.17 Calculation of Total Cell Resources

Name
Value
TX ic
i
F Sampling
W Channel 10
- 8000
Floor f Sampling ----------------------------------8000
TX i ic
TX i ic
TX i ic
TX i ic
D Sym Useful
TX i ic
Used
D Frame
TX i ic
N SD Used Frame
TX i ic
TX i ic
r CP
-------------F
D CP
D Symbol
TX i ic
TX i ic
D Sym Useful + D CP
TDD
TDD
If DL:UL ratio is defined in percentage:

TX ic
TX ic
i
N SD DL Subframe
i
TDD
DL
RoundUp N SD Used Frame r DL Frame O Fixed
If DL:UL ratio is defined in fraction:

TDD
TXi ic
N SD DL
DL
RoundUp N SD Used Frame ----------------------------------------- O Fixed
TDD
TDD
N SD DL + N SD UL
688

AT332_TRR_E0
Name
TX ic
i
R DL
TX ic
i
N Sym DL Subframe
Value
Unit
Description
i
DL
PZ
O Variable
TXi ic
DL
Floor N SD DL Subframe N SCa Data 1 --------------------
100
Symbols
Total downlink cell resources, i.e.,

the number of symbols in the
downlink subframe
SD
Uplink subframe duration in terms of

symbol durations
Symbols
Total uplink cell resources, i.e., the

number of symbols in the uplink
subframe
If DL:UL ratio is defined in percentage:

TX ic
i
TDD
UL
RoundDown N SD Used Frame 1 r DL Frame O Fixed
TX ic
i
If DL:UL ratio is defined in fraction:
N SD UL Subframe
TDD
TX i ic
N SD UL
UL
RoundDown N SD Used Frame ----------------------------------------- O Fixed
TDD
TDD
N SD DL + N SD UL
TX i ic
R UL
TX i ic
N Sym UL Subframe
Mi
UL
PZ
O Variable
TXi ic
UL
Floor N SD UL Subframe N SCa Data 1 -------------------
100
10.2.18 Channel Throughput, Cell Capacity, Allocated Bandwidth

Throughput, and Per-user Throughput Calculation
Name
Value
TX ic
i
R DL
Unit
Description
kbps
Downlink peak MAC channel

throughput
kbps
Downlink effective MAC channel

throughput
kbps

throughput
kbps
Downlink peak MAC cell capacity
kbps
Downlink effective MAC cell capacity
kbps
kbps
Downlink peak MAC throughput per

user
M
i
B DL
--------------------------------D Frame
TX ic
i
R DL
Mi
TX ic
i
DL
For proportional fair schedulers: -------------------------------- G MUG DL
D Frame
Mi
CTP P DL
With downlink segmentation: CTP P DL f Segment DL

With MIMO (SU-MIMO):
Mi
B DL
Max
Mi
1 + f SU MIMO G SU MIMO 1
B DL
With MIMO (AMS):
Mi
B DL
TX i ic
Max
Mi
B DL
TX i ic
if CNR Preamble T AMS

M
CTP E DL
Mi
CTP A DL
Mi
Cap P DL
M
Cap E DL
Mi
Cap A DL
TX i ic
i
i

Mi
Mi
f TP Scaling
- TP Offset
CTP E DL -----------------------100
Mi
TX i ic
Mi
CTP P DL TL DL Max
M
i
i
Cap P DL 1 BLER BDL

Mi
Mi
f TP Scaling
- TP Offset
Cap E DL -----------------------100
Mi
TX i ic
or CINR Preamble T AMS
Mi
i
PUTP P DL
Cap P DL
----------------------TX i ic
N Users DL
689

Name
Value
M
i
PUTP E DL
Mi
kbps
Downlink effective MAC throughput

per user
kbps

user
kbps
Uplink peak MAC channel

throughput
kbps
Uplink effective MAC channel

throughput
kbps

throughput
kbps
Uplink peak MAC cell capacity
kbps
Uplink effective MAC cell capacity
kbps
kbps
Uplink peak MAC allocated

kbps
Uplink effective MAC allocated

Mi
Mi
f TP Scaling
- TPOffset
PUTP E DL -----------------------100
Mi
PUTP A DL
Description
Cap E DL
----------------------TX ic
i
N Users DL
Unit
TX i ic
R UL
Mi
UL
--------------------------------D Frame
TX i ic
R UL
Mi
TX ic
i
UL
For proportional fair schedulers: --------------------------------- G MUG UL
D Frame
With MIMO (SU-MIMO):

Mi
CTP P UL
i
B UL
Max
i
B UL
With MIMO (AMS):

B
Max
Mi
UL
TX i ic
Mi
UL
TX i ic
if CNR Preamble T AMS
TX i ic
TX i ic
With MIMO (MU-MIMO) in uplink throughput coverage predictions:

TX i ic
R UL
Mi
TX ic
i
UL
-------------------------------- G MU MIMO
D Frame
M
Mi
i
i
Mi
Mi
f TP Scaling
- TP Offset
CTP E UL -----------------------100
CTP E UL
CTP A UL
M
Cap P UL
TX ic
i
CTP P UL TL UL Max
M
Mi
i
i
Mi
Mi
f TP Scaling
- TP Offset
Cap E UL -----------------------100
Cap E UL
Cap A UL
Mi
ABTP P UL
Mi
N SC UL
CTP P UL ----------------Mi
Mi
PZ UL
N SC
i
i
ABTP P UL 1 BLER B UL
Mi
Mi
f TP Scaling
ABTP E UL ------------------------- TP Offset
100
kbps
Uplink application allocated

Cap Mi
M
P UL
- ABTP P i UL
Min ----------------------TX i ic
N Users UL
kbps
Uplink peak MAC throughput per

user
Cap Mi
M
E UL
- ABTP E i UL
Min ----------------------TX i ic
N Users UL
kbps
Uplink effective MAC throughput per

user
ABTP A UL
i
PUTP P UL
i
PUTP E UL
690
Mi
ABTPE UL
Mi
Mi

AT332_TRR_E0
Name
M
Value
M
PUTP A UL
i
PUTP E UL
Unit
Description
kbps

user
M
f TP Scaling
i
------------------------- TP Offset
100

Name
Value
Unit
Description
Sel
Mi
R Min DL
TPD Min DL
---------------------------
None

demand in downlink
None

demand in uplink
None

throughput demands
R Min UL
None

demands
Sel
Mi
Sel
Mi
kbps

mobile in downlink
Sel
Mi
Sel
Mi
kbps

mobile in uplink
TX i ic
kbps
Downlink peak channel throughput

with multi-user diversity gain
(Proportional Fair)
TX ic
i
kbps
Uplink peak channel throughput with

multi-user diversity gain
(Proportional Fair)
None

mobile in downlink
None

mobile in uplink
Sel
Mi
Sel
Mi
CTP P DL
Sel
Mi
TPD Min UL
---------------------------
Sel
M
i
R Min UL
Sel
Mi
CTP P UL
TX i ic
R Rem DL
TX i ic
R Rem UL
TX i ic
TL DL Max
Sel
Mi
TX i ic
TL DL Max
Sel
Mi
TPD Rem DL
Sel
Mi
TPD Rem UL
Sel
Mi
CTP P DL
Sel
Mi
CTP P UL
Sel
Mi
RD Rem DL
Sel
Mi
R Min DL
Sel
Mi
Sel
i
TPD Max DL TPD Min DL

TPD Max UL TPD Min UL
Sel
Mi
CTP P DL
M
Without MUG
G MUG DL
Without MUG
G MUG UL
Sel
i
CTP P UL
Sel
Mi
TPD Rem DL
---------------------------Sel
Mi
CTP P DL
Sel
i
RD Rem UL
Sel
Mi
TPD Rem UL
---------------------------Sel
Mi
CTP P UL
691

Name
Value
Unit
Description
None

demand in downlink
None

demand in uplink
TX ic
i
Sel
R Rem DL
Mi
-
Proportional Fair: Min RD Rem DL -------------------N
TX ic
i
R Eff Rem DL
Sel
Mi
RD Rem DL
----------------------------------
Sel
Mi
RD Rem DL
Sel
Mi
Sel
i
R Max DL
M
TX i ic
Sel
R QoS DL
Mi
-
Biased (QoS Class): Min RD Rem DL ------------------N QoS
Sel
Mi
TPD Rem DL
Max Aggregate Throughput: --------------------------Sel
Mi
CTP P DL
TX i ic
Sel
Mi
R Rem DL
Round Robin: Min RD Rem DL ---------------------
N
TX i ic
Sel
Mi
R Rem UL
-
Proportional Fair: Min RD Rem UL -------------------N
TX ic
i
R Eff Rem UL
Sel
Mi
RD Rem UL
----------------------------------
Sel
Mi
RD Rem UL
Sel
Mi
Sel
M
i
R Max UL
TX i ic
Sel
Mi
R QoS UL
-
Biased (QoS Class): Min RD Rem UL ------------------N QoS
Sel
Mi
TPD Rem UL
Max Aggregate Throughput: --------------------------Sel
Mi
CTP P UL
TX ic
i
Sel
Mi
R Rem UL
-
Round Robin: Min RD Rem UL -------------------N
TX i ic
Sel
Mi
TXi ic
Min R Rem DL
RD Rem DL
Sel
M
i
None
Effective remaining downlink

resources in a cell
TX i ic
Sel
Mi
TX i ic
Min R Rem UL
RD Rem UL
Sel
Mi
None
Effective remaining uplink resources

in a cell
None
QoS class bias

(Biased (QoS Class))
None

throughput demands for a QoS class
R Eff Rem DL
R Eff Rem UL
QoS
Sel
Mi
Sel
Mi
Sel
Mi
R Max ErtPS
f Bias
R Max rtPS
R Max nrtPS
- = -------------------------1 + --------= -------------------------= -------------------------Sel
Sel
Sel
100
Mi
Mi
Mi
R Max rtPS
R Max nrtPS
R Max BE
r
TX ic
i
R QoS DL
1 QoS
N QoS ---
TX ic

i
R Rem DL ------------------------------------------------------r
1 QoS
N QoS ---

All QoS
692

AT332_TRR_E0
Name
Value
Unit
Description
TX ic
i
R QoS UL
1 QoS
N QoS ---
TX ic
i
R Rem UL ------------------------------------------------------r
1 QoS
N QoS ---

None

throughput demands for a QoS class
Sel
Sel
Mi
Mi
R Max DL CTP E DL
Sel
M i Site
Max 1 -------------------------------------------------------------------------------------------------------
Sel
Sel
Mi
Mi
Site
TP
CTP
Min DL
BH DL
E DL
Sel
M i Site
None

downlink
Sel
Sel
Mi
Mi
CTP
Max UL
UL
Sel
M i Site
Max 1 -------------------------------------------------------------------------------------------------------
Sel
Sel
Mi
Mi
Site
R Min UL CTP E UL
TP BH UL
Sel
M i Site
None
None

in downlink
None

in uplink
Unit
Description
kbps
Downlink peak MAC user throughput
kbps
Downlink effective MAC user

throughput
kbps

throughput
kbps
Uplink peak MAC user throughput
kbps
Uplink effective MAC user

throughput
kbps
All QoS
Site
BHOF DL
Site
BHOF UL
Sel
Sel
Mi
TL DL
Sel
Mi
= R DL
Sel
Mi
Sel
Mi
Mi
Mi
Max DL CTP P DL

BHOF DL
-----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P DL
Sel
Sel
i
TL UL
M
Sel
i
R UL
M
Sel
Mi
Sel
Mi
M
Mi
i
Max UL CTP P UL

BHOF UL
-----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P UL

Name
Value
Sel
Mi
Sel
Mi
UTP P DL
Sel
i
UTP E DL
M
Sel
Mi
UTP A DL
Sel
Mi
UTP P UL
Sel
Mi
UTP E UL
Sel
Mi
UTP A UL
R DL
Sel
Mi
CTP P DL
Sel
Sel
Mi
i
Sel
i
UTP E DL
M
Sel
Mi
Sel
M
f TP Scaling
i
- TP Offset
-----------------------100
Sel
Mi
R UL
Sel
Mi
CTP P UL
Sel
Sel
Mi
Mi
UTP P UL 1 BLER B UL
Sel
Mi
Sel
Mi
Sel
Mi
f TP Scaling
- TP Offset
UTP E UL -----------------------100
693


algorithm described in "Preamble Signal Level Calculation" on page 711.
TX ic
i
Preamble signal level C Preamble
Path loss L Path
,G
, L Ant , and L Body are not used in the calculations performed for the profile view.

Analysis provided in the reception view is based on path loss matrices. So, you can display received signal levels from the cells
for which calculated path loss matrices are available. For each cell, Atoll displays the received preamble, pilot, or traffic signal
level or C/N.
Reception level bar graphs show the signal levels or C/N in decreasing order. The maximum number of bars in the graph
depends on the preamble signal level of the best server. The bar graph displays cells whose received preamble signal levels
are higher than their preamble C/N thresholds and are within a 30 dB margin from the highest preamble signal level.
You can use a value other than 30 dB for the margin from the highest preamble signal level, for example a smaller value for
Manual.

Atoll displays the best server preamble, pilot, or traffic signal level, and interference from other cells.
Interference level bar graphs show the interference levels on different channels in decreasing order. The maximum number
of bars in the graph depends on the highest interference level on the studied channel. The bar graph displays cells whose C/
N are higher than the minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the
highest interference level on the studied channel.
You can use a value other than 30 dB for the margin from the highest interference level in the studied channel, for example a
smaller value for improving the calculation speed. For more information on defining a different value for this margin, see the
10.3.1.4 Details View

Analysis provided in the details view is based on path loss matrices. So, you can display the received signal level from the best
server and interfering signal levels from other cells for which calculated path loss matrices are available. For each cell, Atoll
displays the best server preamble signal level and interference from other cells.
The results for the best server (first row) are displayed using bold italic characters. Other cells are listed in the decreasing order
of preamble signal level. All the cells from which the received preamble signal level is higher than their preamble C/N
thresholds are listed in the table. As well, interference values are listed for all the cells whose C/N are higher than the
minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the highest interference
level on the preamble.
You can use a value other than 30 dB for the margin from the highest interference level on the preamble, for example a
smaller value for improving the calculation speed. For more information on defining a different value for this margin, see the
694

AT332_TRR_E0

10.3.2.1 Preamble Signal Level Coverage Predictions
The following coverage predictions are based on the received preamble signal levels:
Overlapping Zones
For these calculations, Atoll calculates the received preamble signal level. Then, Atoll determines the selected display
parameter on each pixel inside the cells calculation area. Each pixel within the calculation area of TXi(ic) is considered a noninterfering receiver. For these calculations, the best server calculation is always based on preamble signal level.
L
Mi
,G
Mi
Mi
Mi
, L Ant , and L Body are not considered in the calculations performed for the preamble signal level based coverage
predictions.
For more information on preamble signal level calculations, see "Preamble Signal Level Calculation" on page 711


All Servers
TX ic
TX ic
TX ic
i
i
i
MinimumThreshold C Preamble or L Total or L Path MaximumThreshold

TX ic
TX ic
TX ic
i
i
i
AND
TX i ic
TX j jc
C Preamble Best C Preamble M
ji
If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is the highest.
If M = 2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is either the highest or
within a 2 dB margin from the highest.
If M = -2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is 2 dB higher than the
received preamble signal levels from the cells which are 2nd best servers.

TX ic
TX ic
TX ic
i
i
i
AND
695

TX ic
TX jc
nd
i
j
C Preamble 2 Best C Preamble M
ji
values.
If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is the second highest.
If M = 2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is either the second highest
or within a 2 dB margin from the second highest.
received preamble signal levels from the cells which are 3rd best servers.

such as:

level.
Path Loss (dB)
Total Losses (dB)
highest preamble signal level) and evaluates the path loss from this cell.
highest preamble signal level) and evaluates the total losses from this cell.
of these cells).

The following coverage predictions are based on the received preamble, traffic, or pilot signal levels and noise, and take into
account the receiver characteristics ( L
Mi
,G
Mi
Mi
Mi
, L Ant , and L Body ) when calculating the required parameter:

For these calculations, Atoll calculates the received signal level or C/N level at each pixel for the channel type being studied,
i.e., preamble, traffic, or pilot. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. The
properties of the non-interfering probe receiver are set by selecting a terminal, a mobility type, and a service.
"Preamble Signal Level Calculation" on page 711.

"Traffic and Pilot Signal Level Calculation (DL)" on page 719.
"Traffic Signal Level Calculation (UL)" on page 733
For more information on permutation zone selection, see "Permutation Zone Selection" on page 718.
"Preamble C/N Calculation" on page 715.

"Traffic and Pilot C/N Calculation (DL)" on page 729
"Traffic C/N Calculation (UL)" on page 738.
696


AT332_TRR_E0

page 717.
options:
Preamble Signal Level (DL) (dBm)

Pilot Signal Level (DL) (dBm)
Traffic Signal Level (DL) (dBm)
Preamble C/N Level (DL) (dB)
Pilot C/N Level (DL) (dB)
Traffic C/N Level (DL) (dB)
Permutation Zone (DL)
Segment
options:
Signal Level (UL) (dBm)

C/N Level (UL) (dB)
Permutation Zone (UL)


Mi
,G
Mi
Mi
Mi
For more information on C/(I+N), (I+N), and bearer calculations, see:
"Preamble C/(I+N) Calculation" on page 716.

"Traffic and Pilot C/(I+N) and Bearer Calculation (DL)" on page 731.
"Traffic C/(I+N) and Bearer Calculation (UL)" on page 740.
"Noise Rise Calculation (UL)" on page 737
For more information on thoughput calculations, see:
"Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-User Throughput Calculation" on
page 747.

697


page 717.
options:
Preamble C/(I+N) Level (DL) (dB)

Preamble Total Noise (I+N) (DL) (dBm)
Traffic C/(I+N) Level (DL) (dB)
Traffic Total Noise (I+N) (DL) (dBm)
Pilot C/(I+N) Level (DL) (dB)
options:
Bearer (DL)
Service
options:
Peak MAC Channel Throughput (DL) (kbps)

Effective MAC Channel Throughput (DL) (kbps)
Peak MAC Cell Capacity (DL) (kbps)
Effective MAC Cell Capacity (DL) (kbps)
Peak MAC Throughput per User (DL) (kbps)
Effective MAC Throughput per User (DL) (kbps)
display options:
Quality indicators available in the document (Quality Indicators table): Atoll calculates the downlink traffic C/(I+N)
levels received from the best serving cells at each pixel of their coverage areas. From the C/(I+N), Atoll determines the
best bearer available on each pixel. Then, for the calculated C/(I+N) and bearer, it determines the value of the selected
quality indicator from the quality graphs defined in the WiMAX equipment of the selected terminal.
options:
C/(I+N) Level (UL) (dB)

Total Noise (I+N) (UL) (dBm)
Allocated Bandwidth (UL) (No. of Subchannels)
C/(I+N) Level for 1 Subchannel (UL) (dB)
options:
Bearer (UL)
Service
options:
698
Peak MAC Channel Throughput (UL) (kbps)

Effective MAC Channel Throughput (UL) (kbps)
Peak MAC Cell Capacity (UL) (kbps)
Effective MAC Cell Capacity (UL) (kbps)
Peak MAC Allocated Bandwidth Throughput (UL) (kbps)
Effective MAC Allocated Bandwidth Throughput (UL) (kbps)

AT332_TRR_E0
Application Allocated Bandwidth Throughput (UL) (kbps)

Peak MAC Throughput per User (UL) (kbps)
Effective MAC Throughput per User (UL) (kbps)
display options:
Quality indicators available in the document (Quality Indicators table): Atoll calculates the uplink traffic C/(I+N) levels
received at the best serving cells from each pixel of their coverage areas. From the C/(I+N), Atoll determines the best
quality indicator from the quality graphs defined in the WiMAX equipment of the best serving cell.
10.3.2.4 Cell Identifier Collision Zones Coverage Prediction

The Cell Identifier Collision Zones coverage prediction is based on the received preamble signal levels. Atoll calculates the
received preamble signal level then Atoll determines the selected display parameter on each pixel inside the cells calculation
area. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. L
,G
, L Ant , and L Body are
not considered in the calculations.

For more information on preamble signal level calculations, see "Preamble Signal Level Calculation" on page 711


areas to display. It is possible to determine the coverage area based on the best signal level. The coverage area of each cell
TXi(ic) corresponds to the pixels where:
TX i ic
TX i ic
TX i ic
TX i ic
TX j jc
MinimumThreshold C Preamble or L Total or L Path MaximumThreshold AND C Preamble Best C Preamble M
ji
If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is the highest.
If M = 2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is either the highest or within
a 2 dB margin from the highest.
received preamble signal levels from the cells which are 2nd best servers.

It is possible to display the coverage predictions with colours per cell or:
Number of interferers
Number of interferers per cell

699

station.

"Permutation Zone Selection" on page 718.
"Traffic Signal Level Calculation (UL)" on page 733.
"Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-User Throughput Calculation" on
page 747.



[Simulation]
RandomTotalUsers=0
10.3.4.1.1

700

AT332_TRR_E0
points.
Voice Service (v)
N Call D Call
UL
DL
UL
DL

UL
UL
DL

DL
DL
UL

UL + DL
UL
DL
UL
UL

DL
DL

UL + DL
UL + DL
Therefore, a user can be either active on both links, inactive on both links, active on UL only, or active on DL only.
Data Service (d)
UL
DL
TP Average
DL
and the uplink

UL
and the uplink V
UL
TP Average
UL
during a session.
for the service d.

DL
N Session V 8
N Session V 8
DL
= ------------------------------------------ and f = -----------------------------------------UL
DL
TP Average 3600
TP Average 3600
UL
DL
701


UL

DL
UL
DL
1 f
DL
UL
1 f
UL + DL
UL
DL

UL
UL

DL
DL

UL + DL
UL + DL
UL
UL + DL
DL

10.3.4.1.2

Sector traffic maps per sector are also referred to as live traffic maps. Live traffic data from the OMC is spread over the best
server coverage areas of the transmitters included in the traffic map. Either throughput demands per service or the number
of active users per service are assigned to the coverage areas of each transmitter.

N
UL
UL
DL
TP Cell
TP Cell
DL
= ---------------------- and N = ---------------------UL
DL
TP Average
TP Average
UL
DL
UL
DL

UL

DL
UL
UL
accurately determine the number of active users in the uplink ( n Active
and N
DL

DL
UL + DL
), in the downlink ( n Active ), and both ( n Active ).
UL
DL

702

AT332_TRR_E0
UL
DL

UL
UL
DL

DL
DL
UL

UL + DL
UL
DL
We have: N
UL
UL + DL
UL
DL
UL + DL
DL
UL
UL + DL
DL
UL + DL
N p Active
N p Active
UL + DL
Number of users active in the uplink and downlink both: n Active = Min -------------------------------------- -------------------------------------- or
UL
UL + DL
DL
+ DL
Active
UL + DL
UL
DL
f Act N
DL
UL
f Act
UL
UL
DL

UL
UL + DL
n Active
DL
UL + DL
n Active
UL + DL
DL

nv
- p Inactive
them.

WiMAX cells include intelligent schedulers and radio resource management features for regulating network traffic loads,
optimising spectral efficiency, and satisfying the QoS demands of the users. Each Monte Carlo simulation in the Atoll WiMAX
BWA module is a snap-shot of the network with resource allocation carried out over a duration of 1 second. The number of
WiMAX frames in 1 second depends on the selected frame duration, D Frame . The steps of this algorithm are listed below.
TX i ic
TX i ic
TX i ic
TX i ic
Cell transmission powers and reductions ( P Preamble , P Traffic , P Pilot , and P Idle Pilot ) are set to the values
defined by the user.
Mi
Cell loads ( TL DL
TX i ic
TX i ic
, TL UL
TX i ic
, NRUL
TX i ic
TX i ic
, NR UL Seg , SU DL
TX i ic
, and AU DL
) are set to their current values in the
Cells table.
3. Determines the best servers for all the mobiles generated for the simulation as explained in "Best Server
703

Figure 10.1: WiMAX Simulation Algorithm

4. Determines the mobiles which are within the service areas of their best serving cells as explained in "Service Area
5. Determines the permutation zone assigned to each mobile as explained in "Permutation Zone Selection" on page 718.
6. Determines the downlink and uplink traffic C/(I+N) and bearers for each of these mobiles as explained in "Traffic and
Pilot C/(I+N) and Bearer Calculation (DL)" on page 731 and "Traffic C/(I+N) and Bearer Calculation (UL)" on page 740
respectively.
7. Determines the channel throughputs at the mobile as explained in "Channel Throughput, Cell Capacity, Allocated
Bandwidth Throughput, and Per-User Throughput Calculation" on page 747.
mobile according to the QoS and throughput demands of each mobile using the selected scheduler as explained in
"Scheduling and Radio Resource Allocation" on page 752.
as follows:
Calculation of Traffic Loads:
Atoll calculates the traffic loads for all the cells TXi(ic).
TX i ic
TL DL
Mi
Mi
704
TX i ic
RDL and TLUL
Mi
RUL
Mi

AT332_TRR_E0
TX ic
i
For uplink MU-MIMO, TL UL
MU MIMO
i
RC UL
MU MIMO
M
i
Calculation of Uplink Noise Rise:

For each victim cell TXi(ic), the uplink noise rise is calculated and updated by considering each interfering mobile Mj
as explained in "Noise Rise Calculation (UL)" on page 737.
Calculation of Downlink Segmentation Usage:
Atoll calculates the segmentation usages for all the cells as follows:
Mi
Mi
TX ic
i
PZ
Mi
R DL
Mi
PZ DL = Seg
= Seg
DL
= -----------------------------------------------------------TX i ic
TL DL
SU DL
Where
Mi
R DL
Mi
PZ DL = Seg
is the sum of the percentages of the downlink cell resources allocated to mobiles
i
PZ DL = Seg
served by the downlink segmented permutation zone.

Calculation of Downlink AAS Usage:
Atoll calculates the downlink AAS usages for all the cells as follows:
AAS
Mi
TX i ic
AAS
= ------------------------------TX i ic
TL DL
AU DL
Where
Mi
R DL
Mi
R DL
AAS
is the sum of the percentages of the downlink cell resources allocated to mobiles served by the
AAS
smart antennas.
Calculation of Uplink MU-MIMO Gain:
Atoll calculates the uplink MU-MIMO gain for all the cells as follows:
M
TX ic
i
G MU MIMO
MU MIMO
i
R UL
MU MIMO
Mi
------------------------------------------------------------MU MIMO
Mi
RC UL
MU MIMO
Mi
M
Where
MU MIMO
i
R UL
is the sum of the percentages of the uplink cell resources allocated to MU-MIMO
MU MIMO
Mi
mobiles and
MU MIMO
Mi
RC UL
is the sum of the real resource consumption of MU-MIMO mobiles.
MU MIMO
Mi
TX i ic
TL DL
TX ic
i
Max TL DL
All TX ic
i
TX i ic
TL DL
k 1
705

TX ic
i
TL UL
TX i ic
Req
TX ic
i
TL UL
TX ic
i
Max NR UL
All TX ic
i
TX i ic
If TL DL
TX ic
i
Max TL UL
All TX ic
i
TX ic
i
NR UL
, TL UL
k 1
TX ic
i
NR UL
k 1
TX i ic
Req
, and NR UL
Req

TX i ic
TL DL
TX i ic
TL DL
TX i ic
Req
AND TL UL
TX i ic
TL UL
TX i ic
Req
AND NR UL
TX i ic
NR UL
Req
when creating the simulation, if:
TX ic
i
TL DL
TX ic
i
TL DL
TX ic
i
Req
OR TL UL
TX ic
i
TL UL
TX ic
i
Req
OR NR UL
TX ic
i
NR UL
Req
Simulation Results
Downlink traffic loads

Uplink traffic loads
Uplink noise rise received at the main antenna
Segmented zone uplink noise rise received at the main antenna
Angular distributions of downlink traffic power density for cells with smart antennas
Angular distributions of uplink noise rise for cells with smart antennas
Downlink AAS usage
Downlink segmentation usage
Uplink MU-MIMO capacity gain

due to:
No Coverage: If the mobile does not have any best serving cell (step 3.) or if the mobile is not within the service area
of its best server (step 4.).
No Service: If the mobile is not able to access a bearer in the direction of its activity (step 6.), i.e., UL, DL, or DL+UL.
Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 8.)
uplink, the minimum uplink throughput demand is higher than the uplink allocated bandwidth throughput (step 8.)
Backhaul Saturation: If allocating resources to a mobile makes the effective MAC aggregate site throughputs exceed
the maximum backhaul throughputs defined for the site. This condition is only verified if the simulation was created



A WiMAX network can consist of cells that use different channel bandwidths. Therefore, the start and end frequencies of all
the channels may not exactly coincide. Channel bandwidths of cells can overlap each other with different ratios.
706

AT332_TRR_E0

TX i ic
TX i ic
TX i ic
TX i ic
N Channel + 1 .


Input
TX i ic
TX j jc
F Start FB and F Start FB : Start frequency of the frequency band assigned to the cells TXi(ic) and TXj(jc).
F Start FB can be the start frequency of a TDD frequency band ( F Start FB TDD ), or the uplink or the downlink start
frequency of an FDD frequency band ( F Start FB FDD UL or F Start FB FDD DL ).
First TX ic
i
N Channel
TX i ic
First TX jc
j
and N Channel
TX j jc
N Channel and N Channel : Channel numbers assigned to cells TXi(ic) and TXj(jc).
For FDD networks, Atoll considers that the same channel number is assigned to a cell in the downlink and uplink, i.e.,
the channel number you assign to a cell is considered for uplink and downlink both.
TX i ic
TX j jc
ICS FB
CN FB
TX i ic
TX j jc
and ICS FB
TX i ic
: Inter-channel spacing of the frequency bands assigned to cells TXi(ic) and TXj(jc).
TX j jc
and CN FB
: Channel number step of the frequency bands assigned to cells TXi(ic) and TXj(jc).
707

Calculations
For cell TXi(ic):
TX ic
i
TX ic
i
TX ic
i
TX i ic
F End
TX i ic
N TXi ic N First TX i ic
Channel
Channel
-
------------------------------------------------------TX i ic
CN FB
TX ic
i
F Start = F Start FB + W Channel + ICS FB

TX i ic
For cell TXj(jc):

TX j jc
F Start
TX j jc
F End
TX j jc
TX jc
TX j jc
Channel
Channel
-------------------------------------------------------
TX i ic

CN FB
TX jc
j
j
= F Start FB + W Channel + ICS FB
TX j jc
Output
TX ic
i
TX jc
j
F End
TX i ic
TX j jc
and F End

Input
TX i ic
TX j jc
F Start
TX i ic
TX j jc
F End
and F End
TX ic
i
Calculations
TX i ic
TX j jc
F Start F End
TX i ic
AND F End
TX j jc
F Start

TX i ic TX j jc
W CCO
TX jc
TX ic
TX jc
TX ic
j
i
j
i
= Min F End F End Max F Start F Start

TX i ic TX j jc
r CCO
TX i ic TX j jc
W CCO
= ---------------------------------TX i ic
W Channel
Output
708
TX i ic TX j jc
r CCO

AT332_TRR_E0

Input
TX ic
i
TX jc
j
F Start
F End
TX i ic
TX j jc
and F End
TX i ic
Calculations
TX i ic
TX i ic
TX j jc
TX i ic
TX j jc
AND F Start F Start

TX i ic
F End
TX j jc
F End
TX i ic
AND F End
TX i ic
TX j jc
+ W Channel F Start

TX i ic TX j jc
W ACO
TX jc
TX ic
TX jc
TX ic
TX ic
j
i
j
i
i

TX i ic TX j jc
W ACO
L
= ---------------------------------TX i ic
W Channel
TX ic TX jc
i
j
r ACO
L

TX i ic TX j jc
W ACO
TX jc
TX ic
j
i
= Min F End F End
TX ic
TX jc
TX ic
i
j
i

TX i ic TX j jc
W ACO
H
= ---------------------------------TX i ic
W Channel
TX i ic TX j jc
r ACO

TX i ic TX j jc
r ACO
TX i ic TX j jc
= r ACO
TX i ic TX j jc
+ r ACO
Output
TX ic TX jc
i
j
r ACO
10.4.1.4 FDD TDD Overlap Ratio Calculation

There are many different interference scenarios possible in a WiMAX network depending on the type of duplexing used by
the cells of the network. The most common interference scenarios are FDD-only and TDD-only interferences. However, coexisting FDD and TDD cells may also exist and interfere each other. Atoll models the co-existence of FDD and TDD cells in a
network by determining the FDD TDD overlap ratio as follows:
709

Input
TDD
r DL Frame : Downlink subframe ratio defined in the global network settings.
Calculations
The FDD TDD overlap ratio is calculated as follows depending on the frequency bands assigned to the cells TXi(ic) and TXj(jc):
Frequency Band
TX i ic TX j jc
Overlap Ratio r FDD TDD
TXi(ic)
TXj(jc)
TDD
TDD
TDD
FDD
FDD
TDD
r DL Frame
----------------------100
FDD
FDD
TDD
Output
TX ic TX jc
i
j
r FDD TDD
: FDD TDD overlap ratio between the cells TXi(ic) and TXj(jc).

Input
TX i ic TX j jc
r CCO
r ACO
TX i ic TX j jc
TX i ic TX j jc
r FDD TDD
: FDD TDD overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "FDD TDD Overlap Ratio

TX i ic
f ACS FB : Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic).
W Channel and W Channel : Bandwidths of the channels assigned to the cells TXi(ic) and TXj(jc).
TX ic
i
TX jc
j
Calculations
TX i ic TX j jc
rO
TX ic
f ACS FB
----------------------------
TX ic TX j jc
10
j
i
j
r i
r i
+ r ACO
10
FDD TDD
CCO
TX i ic
TX j jc
TX i ic
TX j jc
TX ic
f ACS FB
TX i ic
----------------------------
TX i ic TX j jc W
j
i
j
10
Channel
r i
+ r ACO
10
r FDD TDD
--------------------TX j jc
CCO
W Channel
TX i ic
W Channel
TX j jc
W Channel
TX j jc
TX j jc
710

AT332_TRR_E0
TX ic
i
W Channel
TX jc
j
W Channel
TX jc
j
TX jc
j
Output
TX i ic TX j jc
rO
10.4.2 Preamble Signal Level and Quality Calculations

These calculations include the calculation of the received preamble signal level, and the noise and interference on the
preamble. The following sections also describe how the received preamble signal level, the noise and interference, C/N, and
C/(I+N) ratios are calculated in Atoll:

"Preamble Noise Calculation" on page 712.
"Preamble C/N Calculation" on page 715.
"Preamble Interference Calculation" on page 714.
10.4.2.1 Preamble Signal Level Calculation

Input
TX i ic
P Preamble : Preamble transmission power of the cell TXi(ic).
G SA
Div
G SA
TX i
Combining
TX i
TX
: Smart antenna diversity gain (for cross-polarised smart antennas) defined per clutter class.
: Transmitter antenna gain for the antenna used by the transmitter TXi.
i
= L Total DL ).
TX
TX
is selected.
Mi
Mi
Mi
Mi
Mi
Mi
711

, G
, L Ant , and L Body are not used in the calculations performed for the point
analysis tools profile tab and the preamble signal level based coverage predictions.
Calculations
The received preamble signal level (dBm) from any cell TXi(ic) is calculated for a pixel, subscriber, or mobile Mi as follows:
TX i ic
TX i ic
C Preamble = EIRP Preamble L Path M Shadowing Model L Indoor + G
Mi
Mi
Mi
Mi
L Ant L Body
Where EIRP is the effective isotropic radiated power of the cell calculated as follows:
TX ic
i
TX ic
i
Without smart antennas: EIRP Preamble = P Preamble + G
With smart antennas: EIRP Preamble = P Preamble + G
TX i ic
TX i ic
TX i
TX
L
TX i
TX
i
TX i
Combining
Div
+ G SA

TX i

Furthermore, the total losses between the cell and the pixel, subscriber, or mobile Mi can be calculated as follows:
L Total = L Path + L
TX
TX
+L
+ L Ant + L Body
If you wish to exclude the the energy corresponding to the cyclic prefix part of the total
symbol duration from the useful signal level, you must add the following lines in the
Atoll.ini file:
[WiMAX]
ExcludeCPFromUsefulPower = 1
TX i ic
When this option is active, the cyclic prefix energy is excluded from C Preamble . In other
TX ic
i
words, the factor 10 Log 1 r CP
TX ic
is added to C i
Preamble .
Independant of the option, interference levels are calculated for the total symbol
durations, i.e., the energy of the useful symbol duration and the cyclic prefix energy.
Output
TX i ic
C Preamble : Received preamble signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
L Path : Path loss between the cell TXi(ic) and the pixel, subscriber, or mobile Mi.
L Total : Total losses between the cell TXi(ic) and the pixel, subscriber, or mobile Mi.
10.4.2.2 Preamble Noise Calculation

For determining the preamble C/N and C/(I+N), Atoll calculates the preamble noise over the bandwidth used by the cell. The
used bandwidth depends on the number of subcarriers used by the preamble.The number of subcarriers used by the
preamble can be different from the number of subcarriers used by the permutation zones.
The preamble noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on
the temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the
used bandwidth.
Input
712
N SCa Preamble : Number of subcarriers used by the preamble defined for the frame configuration of the cell TXi(ic).
N SCa Total : Total number of subcarriers defined for the frame configuration of the cell TXi(ic).
TX i ic
TX i ic

AT332_TRR_E0
TX ic
i
F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on page 744.
nf
Calculations
n 0 = 10 Log K T 1000 = 174 dBm/Hz
The thermal noise over the preamble for a cell is calculated as:
TX i ic
n 0 Preamble
TX ic
i
TX ic
N SCa Preamble Preamble
i
--------------------------------- f Segment
= n 0 + 10 Log F Sampling
TX ic
i
N SCa Total
Effect of Segmentation:
The preamble is segmented and one of the three preamble carrier sets is used for transmission. Each preamble carrier
set uses 1/3rd of the total number of preamble subcarriers. The power transmitted over the preamble has higher
spectral density than the power transmitted over the entire channel bandwidth. This power concentration due to
segmentation on the C/N and C/(I+N) results in an increase in the coverage footprint of the preamble. Hence, the
Preamble
thermal noise at the pixel, subscriber, or mobile Mi covered by the preamble is reduced by a factor of f Segment = 1
--- .
3
The following table shows the different types of subcarriers and their numbers for preamble transmission in WiMAX.
N SCa Total
128
512
1024
2048
Guard Subcarriers
DC
Subcarrier
N SCa Preamble
All
1 (54)
107
1 (54)
35
0.3271
None
36
0.3364
None
36
0.3364
All
1 (214)
428
None
143
0.3341
1 (214)
142
0.3318
None
143
0.3341
All
1 (426)
851
1 (426)
283
0.3325
None
284
0.3337
None
284
0.3337
All
1 (852)
1703
1 (852)
567
0.3329
None
568
0.3335
None
568
0.3335
Segment
Left
10
42
86
172
Right
10
41
86
172
Total
20
83
172
344
Preamble
f Segment
The preamble noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel,
TX i ic
TX i ic
n Preamble = n 0 Preamble + nf
Mi
Output
TX i ic
n Preamble : Preamble noise for the cell TXi(ic).
713

10.4.2.3 Preamble Interference Calculation

the preamble signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the channels
used by the cells TXi(ic) and TXj(jc) and which preamble carrier sets are used by the two cells.
Input
TX j jc
C Preamble : Preamble signal level received from an interfering cell TXj(jc) as calculated in "Preamble Signal Level
Calculation" on page 711 at the pixel, subscriber, or mobile Mi covered by the cell TXi(ic).

In Monte Carlo simulations, interfering signal levels already include M Shadowing Model , as explained in "Preamble
Signal Level Calculation" on page 683.
information, see "Shadow Fading Model" on page 90). As the received interfering signal levels already include
TX jc
j
TX jc
j
C Preamble = C Preamble + M Shadowing C I

is selected.
TX i ic TX j jc
rO
N Seg
TX i ic
TX j jc
and N Seg
: Segment numbers assigned to the cells TXi(ic) and TXj(jc) calculated from their respective
TX i ic
TX j jc
preamble indexes ( n Preamble and n Preamble ) as follows:
Inter Tech
f IRF
n Preamble
N Seg
0 to 31, 96, 99, 102, 105, 108, 111
32 to 63, 97, 100, 103, 106, 109, 112
64 to 95, 98, 101, 104, 107, 110, 113
: Inter-technology interference reduction factor.
Calculations
The received preamble interference (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows:
TX j jc
TX j jc
TX i ic TX j jc
I Preamble = C Preamble + f O
TX ic TX jc
i
j
Where f O
TX i ic TX j jc
Inter Tech
+ f Seg Preamble + I DL
is the interference reduction factor due to channel overlap between the cells TXi(ic) and TXj(jc),
TX i ic TX j jc
fO
TX i ic TX j jc
= 10 Log r O
TX i ic TX j jc
f Seg Preamble is the interference reduction factor due to preamble segmentation, calculated as follows:
TX i ic TX j jc
TX i ic TX j jc
f Seg Preamble = 10 Log p Collision
714

AT332_TRR_E0
TX ic TX jc
i
j
The probability of preamble subcarrier collision p Collision

TX ic
i
TX jc
j
= N Seg
1 if N Seg
TX ic
i
between the cells TXi(ic) and TXj(jc) is 0 if N Seg
TX jc
j
N Seg
and
.
TX jc
j
TX jc
j
In case of smart antennas, C Preamble in I Preamble already includes the effect of the
TX
number of antenna elements ( E SA ). If you wish to include the effect of the number of
antennas in case of MIMO, you must add the following lines in the Atoll.ini file:
[WiMAX]
MultiAntennaInterference
When the multi-antenna interference option is active, and TXj(jc) does not have a smart
antenna
equipment
assigned,
the
interference
is
incremented
by
TX jc
j
+ 10 Log N Ant TX .
TX j jc
Where N Ant TX is the number of MIMO transmission (downlink) antennas defined for
the cell TXj(jc).

TX k
P DL Rec
---------------------------------------
=
F TX i ic TX k
TX k ICP DL
Inter Tech
I DL
TX
F TX i ic TX k
ICPDL

TX k
Output
TX j jc
I Preamble : Preamble interference received from any interfering cell TXj(jc) at the pixel, subscriber, or mobile Mi
covered by a cell TXi(ic).
Inter Tech
I DL
10.4.2.4 Preamble C/N Calculation

Input
TX i ic
C Preamble : Received preamble signal level from the cell TXi(ic) as calculated in "Preamble Signal Level Calculation" on
page 711.
TX i ic
n Preamble : Preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on page 712.
G Div Preamble : Preamble diversity gain defined in the WiMAX equipment of the terminal used by the pixel, subscriber,
Mi
or mobile Mi.
DL
located.
Calculations
The preamble C/N for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
715

TX ic
i
TX ic
i
TX ic
i

M
DL
CNR Preamble = C Preamble n Preamble + G Div Preamble + G Div

The preamble diversity gain is applied to the preamble C/N when the cell and the terminal both support any form of MIMO in
downlink. The additional downlink diversity gain defined per clutter is also applied.
Output
TX i ic
CNR Preamble : Preamble C/N from the cell TXi(ic) at any pixel, subscriber, or mobile Mi.
10.4.2.5 Preamble C/(I+N) Calculation

The carrier signal to interference and noise ratio is calculated in three steps. First Atoll calculates the received preamble signal
level from the studied cell (as explained in "Preamble Signal Level Calculation" on page 711) at the pixel, subscriber or mobile
under study. Next, Atoll calculates the interference received at the same studied pixel, subscriber, or mobile from all the
interfering cells (as explained in "Preamble Interference Calculation" on page 714). Interference from each cell is weighted
according to the co- and adjacent channel overlap between the studied and the interfering cells, and the probabilities of
subcarrier collision. Finally, Atoll takes the ratio of the preamble signal level, and the sum of the total interference from all
interfering cells and the noise (as calculated in "Preamble Noise Calculation" on page 712).
Input
TX i ic
C Preamble : Preamble signal level received from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in
TX i ic
n Preamble : Preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on page 712.
I Preamble : Preamble interference received from any cell TXj(jc) at a pixel, subscriber, or mobile Mi covered by a cell
TX j jc
TXi(ic) as calculated in "Preamble Interference Calculation" on page 714.

Inter Tech
NRDL
G Div Preamble : Preamble diversity gain defined in the WiMAX equipment of the terminal used by the pixel, subscriber,
or mobile Mi.
DL
located.
Inter Tech
I DL
: Downlink inter-technology interference as calculated in "Preamble Interference Calculation" on page 714.
Calculations
The preamble C/(I+N) for a cell TXi(ic) is calculated as follows at any pixel, subscriber, or mobile Mi:
TX ic
TX i ic
CINR Preamble
TXj jc
n Preamble
IPreamble
-----------------------------
TX i ic
M
Inter Tech
DL
10
-
+ NR Inter Tech + G i
--------------------------= C Preamble 10 Log
+ I DL
+ 10
10
DL
Div Preamble + G Div
10
All TXj jc
The preamble diversity gain is applied to the preamble C/(I+N) when the cell and the terminal both support any form of MIMO.
The additional downlink diversity gain defined per clutter is also applied.
The preamble total noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX ic
TX i ic
I + N Preamble
716
TXj jc
n Preamble
I Preamble
-----------------------------
Inter Tech
10
-
+ NR Inter Tech
--------------------------= 10 Log
+I
+ 10
10
DL
10
DL
All TXj jc

AT332_TRR_E0
Output
TX ic
i
CINR Preamble : Preamble C/(I+N) from the cell TXi(ic) at a pixel, subscriber, or mobile Mi.
I + N Preamble : Preamble total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered
TX ic
i
by a cell TXi(ic).

In WiMAX, best server refers to a cell ("serving transmitter"-"reference cell" pair) from which a pixel, subscriber, or mobile Mi
gets the highest preamble signal level or preamble C/(I+N). This calculation also determines whether the pixel, subscriber, or
mobile Mi is within the coverage area of any transmitter or not.
Input
TX i ic
C Preamble : Preamble signal level received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in
"Preamble Signal Level Calculation" on page 711 using the terminal and service parameters ( L
, G
, L Ant , and
Mi
L Body ) of Mi. "Preamble C/(I+N) Calculation" on page 716
TX i ic
CINR Preamble : Preamble C/(I+N) received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in
Calculations
The best server of any pixel, subscriber, or mobile Mi, BS M , is the cell from which the received preamble signal level or C/
i
(I+N) is the highest among all the cells. The best server is determined as follows:
BSM = TX i ic
i
TX i ic
TX i ic
C
C
= Best
Preamble All TX ic Preamble
or BS M = TX i ic
i
CINR
TX i ic
TX i ic
CINR
= Best
Preamble All TX ic
Preamble
Here ic is the cell of the transmitter TXi with the highest preamble power. However, if more than one cell of the same
transmitter covers the pixel, subscriber, or mobile, the final reference cell ic might be different from the initial cell ic (the one
with the highest power) depending on the serving cell selection method:
Random: In coverage prediction calculations and in calculations on subsriber lists, the cell of the highest priority layer
is selected as the serving (reference) cell. In Monte Carlo simulations, a random cell is selected as the serving
(reference) cell.
Distributive: In coverage prediction calculations and in calculations on subsriber lists, the cell of the highest priority
layer is selected as the serving (reference) cell. In Monte Carlo simulations, mobiles are distributed among cell layers
one by one, i.e., if more than one cell layer covers a set of mobiles, the first mobile is assigned to the highest priority
layer, the 2nd mobile to the second highest priority layer, and so on.
When using either the Random or the Distributive cell selection method, the reference cell once assigned to a mobile
does not change during Monte Carlo simulations.
Output

i
10.4.4 Service Area Calculation

In WiMAX, a pixel, subscriber, or mobile Mi can be covered by a cell (as calculated in "Best Server Determination" on page 717)
but can be outside the service area. A pixel, subscriber, or mobile Mi is said to be within the service area of its best serving cell
TXi(ic) if the preamble C/N from the cell at the pixel, subscriber, or mobile is greater than or equal to the preamble C/N
threshold defined for the cell.
Input
TX i ic
CNR Preamble : Preamble C/N from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Preamble C/N
717

TX ic
i
T Preamble : Preamble C/N threshold defined for the cell TXi(ic).
Calculations
A pixel, subscriber, or mobile Mi is within the service area of its best serving cell TXi(ic) if:
TX i ic
TX i ic
CNR Preamble T Preamble

Output
True: If the calculation criterion is satisfied.

False: Otherwise.
10.4.5 Permutation Zone Selection

In order to be able to calculate the traffic C/(I+N) and the throughputs, a permutation zone is assigned to each pixel,
subscriber, or mobile Mi located within the service area (as calculated in "Service Area Calculation" on page 717) of its best
serving cell. The permutation zone assigned to Mi is one which covers Mi in terms of distance and preamble C/N or C/(I+N),
and accepts user speeds equal to or higher than Mis speed selected for the calculation.
A pixel, subscriber, or mobile Mi which is unable to get a permutation zone is considered to be outside the service area.
Input
TX i ic
d Max PZ : Maximum distance covered by a permutation zone of a cell TXi(ic).
QT PZ
TX i ic
: Minimum preamble C/N or C/(I+N) required at the pixel, subscriber, or mobile Mi to connect to a
permutation zone of a cell TXi(ic).

TX i ic
Speed Max PZ : Maximum speed supported by a permutation zone of a cell TXi(ic).
M i TX i ic
: Distance between the pixel, subscriber, or mobile Mi and a cell TXi(ic).
TX i ic
CNR Preamble : Preamble C/N from the cell TXi(ic) as calculated in "Preamble C/N Calculation" on page 715.
CINR Preamble : Preamble C/(I+N) from the cell TXi(ic) as calculated in "Preamble C/(I+N) Calculation" on page 716.
Mobility M i : Speed of the pixel, subscriber, or mobile Mi.
TX i ic
Calculations
Mi is assigned the permutation zone with the highest priority among the permutation zones whose selection criteria Mi
satisfies. Mi satisfies the selection criteria of a permutation zone if:
The distance between Mi and TXi(ic) is less than or equal to the maximum distance covered by the permutation zone:
d
M i TX i ic
TX i ic
d Max PZ
The preamble C/N or C/(I+N) at Mi is better than or equal to the quality threshold defined for the permutation zone:
TX i ic
TX i ic
CNR Preamble QT PZ
TX i ic
TX i ic
or CINR Preamble QT PZ
The mobility of Mi is less than or equal to the maximum mobile speed supported by the permutation zone:
TX i ic
Mobility M i Speed Max PZ

Therefore, the permutation zones assigned to a pixel, subscriber, or mobile Mi in the downlink and uplink are:
718

AT332_TRR_E0
PZ DL
Mi
PZ UL
TX ic
i
= Highest Priority PZ DL
TX ic
TX ic
i
i
CNR
QT
Preamble
PZ
TX i ic
TX i ic
M i TX i ic
d Max PZ AND
AND Mobility M i Speed Max PZ
OR
d
TX i ic
TX i ic
CINR
QT
TX ic
i
= Highest Priority PZ UL
TX i ic
TX i ic
CNR
QT
Preamble
PZ
TX i ic
TX i ic
M i TX i ic
d
AND
AND Mobility M Speed
OR
d
Max PZ
i
Max PZ
TX ic
TX ic
i
CINR i
QT
Preamble
Preamble
PZ
PZ
If more than 1 permutation zone satisfies the distance, speed, and quality threshold criteria, and all have the same priority,
the permutation zone assigned to the pixel, subscriber, or mobile will be the first in the list of permutation zones (frame
configuration) among these zones.
Output
Mi
Mi
PZ DL and PZ UL : Downlink and uplink permutation zones assigned to the pixel, subscriber, or mobile Mi.
10.4.6 Traffic and Pilot Signal Level and Quality Calculations

Traffic and pilot subcarriers can be transmitted with different transmission powers than the preamble power of a cell, and do
not suffer the same interference and noise as the preamble. The following sections describe how traffic and pilot signal levels,
noise and interference, C/N, and C/(I+N) ratios are calculated on the downlink and uplink.

"Traffic and Pilot Noise Calculation (DL)" on page 721.
"Traffic and Pilot Interference Calculation (DL)" on page 722.
"Traffic and Pilot C/N Calculation (DL)" on page 729.
"Traffic Noise Calculation (UL)" on page 734.
"Traffic Interference Calculation (UL)" on page 735.
"Traffic C/N Calculation (UL)" on page 738.
10.4.6.1 Traffic and Pilot Signal Level Calculation (DL)

Input
TX i ic
P Preamble : Preamble transmission power of the cell TXi(ic).
P Traffic : Traffic power reduction of the cell TXi(ic).
P Pilot : Pilot power reduction of the cell TXi(ic).
TX i ic
TX i ic
TX i
Without smart antennas: G
With smart antennas: G

G
TX i
TX i
TX i
is the transmitter antenna gain, i.e., G
TX i
TX i
= G Ant .
is the smart antenna gain in the direction of the pixel, subscriber, or mobile Mi, i.e.,
= G SA . Where is the direction in which Mi is located. For more information on the calculation of
G SA , refer to section "Beamforming Smart Antenna Models" on page 43.
719

Array
G SA
G SA
G SA : Smart antenna diversity gain (for cross-polarised smart antennas) defined per clutter class.
: Smart antenna array gain offset defined per clutter class.
Combining
Div
TX
TX
= L Total DL ).
TX i
TX
is selected.
Mi
Mi

Mi
Mi
M
Mi
Calculations
The received traffic and pilot signal levels (dBm) from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi as
follows:
TX i ic
TX i ic
Mi
TX i ic
Mi
C Traffic = EIRP Traffic L Path M Shadowing Model L Indoor + G

TX i ic
C Pilot
= EIRP Pilot L Path M Shadowing Model L Indoor + G
L
L
Mi
Mi
Mi
Mi
Mi
Mi
L Ant L Body and

L Ant L Body
TX ic
i
TX ic
i
TX
TX ic
i
TX
EIRP Traffic = P Traffic + G

TX ic
i
EIRP Pilot
= P Pilot + G
TX ic
i
TX ic
i
With P Traffic and P Pilot

TX i ic
TX i ic
+ G SA
+ G SA
Array
+ G SA
Array
+ G SA
+ G SA L
Div
Combining
+ G SA L
Div
TX
TX
and
being the traffic and pilot transmission powers of the cell TXi(ic) calculated as follows:
TX i ic
TX i ic
P Traffic = P Preamble P Traffic and P Pilot
720
Combining
TX i ic
TX i ic
= P Preamble P Pilot

AT332_TRR_E0
If you wish to exclude the the energy corresponding to the cyclic prefix part of the total
symbol duration from the useful signal level, you must add the following lines in the
Atoll.ini file:
[WiMAX]
TX i ic
When this option is active, the cyclic prefix energy is excluded from C Preamble . In other
TX ic
i
TX ic
is added to C i
Preamble .
Output
TX i ic
C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i ic
10.4.6.2 Traffic and Pilot Noise Calculation (DL)

For determining the traffic and pilot C/N and C/(I+N), Atoll calculates the downlink noise over the channel bandwidth used by
the cell. The used bandwidth depends on the number of used subcarriers. The numbers of subcarriers used by different
permutation zones can be different.
The downlink noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on
the temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the
used bandwidth.
Input
M
i
PZ DL
N SCa Used : Number of subcarriers used by the downlink permutation zone of a cell TXi(ic) assigned to Mi.
N SCa Total : Total number of subcarriers defined for the frame configuration of a cell TXi(ic).
nf
TX i ic
TX i ic
M
Calculations
n 0 = 10 Log K T 1000 = 174 dBm/Hz
The thermal noise for a cell is calculated as:
Mi
TX i ic
n 0 DL
PZ DL
TXi ic
N SCa Used
------------------------= n 0 + 10 Log F Sampling TX ic
i
N SCa Total
TX i ic
n DL
TX i ic
= n 0 DL + nf
Mi
Effect of Segmentation:
If you select downlink segmentation support for the frame configuration used by the cell, it means that the first
downlink PUSC permutation zone is segmented. All other zones are pooled together to form a non-segmented zone.
The downlink segmenting factor, f Segment DL , is calculated from the number of secondary subchannel groups
assigned to the permutation zone in the Permutation Zones table.
721

PSG + 2 SSGf Segment DL = 3

-------------------------------------------15
Where, PSG is the number of primary subchannel groups and SSG is the number of used secondary subchannel groups.
The multiplicative coefficients of 3 and 2 are derived from the ratio of the numbers of
subchannels that belong to the primary and to the secondary subchannel gourps. For
example, for the FFT size of 1024 (or 2048), each primary subchannel group contains 6
(or 12) subchannels, and each secondary subchannel group contains 4 (or 8)
subchannels, which gives the ratio of 3:2. And, the denominator of 15 = 3 x 3 + 2 x 3.
f Segment DL represents the fraction of the channel bandwidth used by a downlink segment. The power transmitted
1
over a segment has ---------------------------- times the spectral density of the power transmitted over the entire channel
f Segment DL
1
bandwidth. When calculating the downlink C/N and C/(I+N) ratios, the increase in power by ---------------------------- due to this
f Segment DL
power concentration is equivalent to a reduction in the noise level by f Segment DL . Hence, if downlink segmentation
is used, the thermal noise power at the pixel, subscriber, or mobile Mi covered by the downlink segmented
permutation zone is reduced by the factor f Segment DL . Which means that the thermal noise for the a segment of the
channel used by a cell is calculated as:
Mi
TX i ic
n 0 DL
PZ DL
N SCa Used
TX i ic
f Segment DL
= n 0 + 10 Log FSampling ------------------------TX i ic
N SCa Total
Output
TX i ic
n DL
: Downlink noise for the cell TXi(ic).
10.4.6.3 Traffic and Pilot Interference Calculation (DL)

the traffic and pilot signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the
channels used by the cells TXi(ic) and TXj(jc), on the traffic loads of the interfering cells TXj(jc), and whether the cells use
downlink segmentation or not. Moreover, the interference can come from cells using simple as well as smart antennas.
The calculation can be divided into the two parts.
10.4.6.3.1
"Traffic and Pilot Interference Signal Levels Calculation (DL)" on page 722.
"Effective Traffic and Pilot Interference Calculation (DL)" on page 726.
Traffic and Pilot Interference Signal Levels Calculation (DL)

The traffic and pilot signal levels received from interfering cells TXj(jc) at a pixel, subscriber, or mobile Mi, covered by a cell
TXi(ic), are calculated in a different manner than the traffic and pilot signal levels from the studied cell TXi(ic). This section
explains how these interfering signals are calculated.
Input
722
TX j jc
P Preamble : Preamble transmission power of the cell TXj(jc).
P Pilot : Pilot power reduction of the interfering cell TXj(jc).
P Traffic : Traffic power reduction of the interfering cell TXj(jc).
P Idle Pilot : Idle pilot power reduction of the interfering cell TXj(jc).
TX j jc
TX j jc
TX j jc
TX
: Total transmitter losses for the transmitter TXj ( L
TX
= L Total DL ).
TX j

AT332_TRR_E0
TX
L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXj.

is selected.
Mi
Mi
Mi
M
L Ant is determined in the direction of TXj(jc) from the antenna patterns of the antenna used by Mi while the antenna
is pointed towards TXi(ic).
Mi
TL DL
TX jc
j
TX jc
j
AU DL
: Downlink AAS usage ratio of the interfering cell TXj(jc).
Downlink AAS usage ratios are calculated using Monte Carlo simulations as explained in "Simulation Process" on
page 703.
TX j jc
N SCa Used : Number of used subcarriers defined for the first downlink permutation zone in the frame configuration
assigned to the interfering cell TXj(jc).
TX j jc
N SCa Data : Number of data subcarriers defined for the first downlink permutation zone in the frame configuration
assigned to the interfering cell TXj(jc).
Calculations
WiMAX cells can transmit different powers on pilot (NUsed NData) and data (NData) subcarriers for the part of the frame with
traffic, and a different pilot power for the part of the frame that does not have traffic bursts. Data subcarriers are off during
the empty part of the frame. Therefore, the interference received from a cell depends on the traffic load and the different
powers of the cell, i.e., pilot, traffic, and idle pilot powers.
Monte Carlo simulations and coverage prediction calculations present different scenarios for interference calculations in the
case of smart antennas.

In the case of Monte Carlo simulations, the interferer is either using the transmitter antenna or the smart antenna at
any given moment. So, for each interfered pixel, subscriber, or mobile, Atoll already knows the type of the
interference source. Therefore, the interference received from any cell TXj(jc) can be given by:
TX jc
TX j jc
Without smart antennas: I Total
TX jc
j
I j
I Idle
Non AAS
----------------------------------------------
10
10
+ 10
= 10 Log 10
TX jc
TX j jc
With smart antennas: I Total
I j
AAS -
-----------------10
= 10 Log 10
723

In the case of coverage prediction calculations, the interferer could either be transmitting using the transmitter
antenna, or using the smart antenna, or it could be empty, or not transmitting.Therefore, the interference received
from any cell TXj(jc) can be given by:
TX jc
TX jc
TX jc
j
j
I j
I
I
Non AAS
Idle
AAS
----------------------------------------------------------------
10
10
10
+ 10
+ 10
= 10 Log 10
TX jc
j
I Total
Where, the three components of the interference are:

TX j jc
I Non AAS : Interference from the loaded part of the frame transmitted using the main antenna,
I AAS
I Idle
TX jc
j
TX j jc
: Interference from the loaded part of the frame transmitted using the smart antenna,
: Interference from the empty, or idle, part of the frame.
The above components of the interference are calculated as follows:

The interference from the loaded part of the frame transmitted using the main antenna is calculated as follows:
The received interfering traffic and pilot signal levels (dBm) from any cell TXj(jc) are calculated for a pixel, subscriber,
or mobile Mi as follows:
In Monte Carlo simulations:
TX j jc
TX j jc
Mi
TX j jc
Mi
I Traffic = EIRP Traffic L Path M Shadowing Model L Indoor + G

TX j jc
= EIRP Pilot L Path M Shadowing Model L Indoor + G
I Pilot
L
L
Mi
Mi
Mi
Mi
Mi
Mi
L Ant L Body
L Ant L Body
In coverage prediction:
TX j jc
TX j jc
Mi
TX j jc
Mi
I Traffic = EIRP Traffic L Path M Shadowing Model + M Shadowing C I L Indoor + G

TX j jc
= EIRP Pilot L Path M Shadowing Model + M Shadowing C I L Indoor + G
I Pilot
L
L
Mi
Mi
Mi
Mi
Mi
Mi
L Ant L Body
L Ant L Body
TX j jc
TX j jc
EIRP Traffic = P Traffic + G

TX j jc
TX j jc
With P Traffic and P Pilot

TX jc
j
TX jc
j
TX j
TX j
TX j jc
and EIRP Pilot
TX j
TX j
TX j
being the traffic and pilot transmission powers of the cell TXj(jc) calculated as follows:
TX jc
j
TX jc
j
P Traffic = P Preamble P Traffic and P Pilot

And G
TX j jc
= P Pilot + G
TX jc
j
TX jc
j
= P Preamble P Pilot
TX j
= G Ant , i.e., the transmitter antenna gain for the antenna used by the transmitter TXj.
The interference from the loaded part of the frame transmitted using the main antenna is given as:
TX jc
TX j jc
I Non AAS
TX jc
j
I j
I Pilot
TX j jc
TX j jc
Traffic
------------------ TX jc
------------------
TX
jc
N
N
j
j
10
10
SCa
Data
SCa
Data
- + 10
----------------------- 1 -------------------------
1 AU DL 10
= 10 Log TL DL
TX j jc
TX j jc
N SCa Used
N SCa Used
If you wish to include the effect of the number of antennas in case of MIMO, you must
[WiMAX]
When the multi-antenna interference option is active, the interference is incremented by
TX jc
TX jc
j
j
+ 10 Log N Ant TX . Where N Ant TX is the number of MIMO transmission
(downlink) antennas defined for the cell TXj(jc).
724

AT332_TRR_E0
The interference from the loaded part of the frame transmitted using the smart antenna is calculated as follows:
The received interfering traffic signal level (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi
as follows:
In Monte Carlo simulations:
TX j jc
I AAS
TX j jc
= EIRPAAS
Mi
Mi
Mi
Mi
L Ant L Body
In coverage prediction:
TX j jc
I AAS
TX j jc
= EIRPAAS
L Path M Shadowing Model + M Shadowing C I L Indoor + G
Mi
Mi
Mi
Mi
L Ant L Body
TX j jc
EIRP AAS
TX j jc
= P Traffic + G
TX j
TX j
TX j jc
With P Traffic being the traffic transmission power of the cell TXj(jc) calculated as follows:
TX j jc
TX j jc
TX j jc
P Traffic = P Preamble P Traffic

And, G
TX j
= G SA is the smart antenna gain in the direction of the victim mobile Mi, calculated from the angular
distributions of the downlink traffic power density of the interfering cells. The angular distribution of the downlink
traffic power density is determined from the array correlation matrices calculated during Monte Carlo simulations.
is the direction in which the victim pixel, subscriber, or mobile Mi is located. For more information on the
calculation of G SA , see "Beamforming Smart Antenna Models" on page 43.
The gain of the interfering signal, G SA , transmitted in the direction of each pixel is given by:
H
G SA = g n S R Avg S
Where S is the steering vector in the direction (probe mobile/pixel), H denotes the Hilbert transform, R Avg is the
average array correlation matrix, and g n is the gain of the nth antenna element in the direction .
The interference from the empty, or idle, part of the frame transmitted using the transmitter antenna is calculated
as follows:
The received interfering pilot signal level (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi
as follows:
TX j jc
TX j jc
I Idle Pilot = EIRP Idle Pilot L Path L Indoor + G
Mi
Mi
Mi
Mi
L Ant L Body
TX j jc
TX j jc
EIRP Idle Pilot = P Idle Pilot + G
TX j
TX j
TX j jc
With P Idle Pilot being the idle pilot transmission power of the cell TXj(jc) calculated as follows:
TX j jc
TX j jc
TX j jc
P Idle Pilot = P Preamble P Idle Pilot

And, G
TX j
TX j
= G Ant , i.e., the transmitter antenna gain for the antenna used by the transmitter TXj.
The interference from the empty, or idle, part of the frame transmitted using the transmitter antenna is given as:
TX jc
TX j jc
I Idle
I j
TX j jc
Idle Pilot
----------------------------
TX j jc
N
10
SCa Data
1 -----------------------= 10 Log 1 TL DL 10

TX j jc
N SCa Used
725

If you wish to include the effect of the number of antennas in case of MIMO, you must
[WiMAX]
When the multi-antenna interference option is active, the interference is incremented by
TX jc
TX jc
j
j
+ 10 Log N Ant TX . Where N Ant TX is the number of MIMO transmission
(downlink) antennas defined for the cell TXj(jc).
Output
10.4.6.3.2
TX j jc
I Total : Interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc).
Effective Traffic and Pilot Interference Calculation (DL)

The effective downlink traffic and pilot interference received at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) from
interfering cells TXj(jc) depends on the co- and adjacent channel overlap that exists between the channel used by the studied
cell and the interfering cells, and the downlink segmentation parameters of the studied and interfering cells. The first
downlink PUSC zone can be segmented at the studied and the interfering cells. The probability of subcarrier collision depends
on the lengths of the segmented zones and on the subchannel groups used at both sides.
Input
TX j jc
I Total : Interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc) as calculated in
"Traffic and Pilot Interference Signal Levels Calculation (DL)" on page 722.
TX i ic TX j jc
SU DL
rO
TX i ic
TX j jc
and SU DL
: Downlink segmentation usage ratios defined for cells TXi(ic) and TXj(jc).
Calculations
The total traffic and pilot interference (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows:
TX j jc
I DL
TX j jc
TX i ic TX j jc
= I Total + f O
TX i ic TX j jc
+ f Seg DL
Inter Tech
+ I DL
Calculations for the interference reduction factors due to channel overlapping and downlink segmentation are explained
below:
as follows:
TX i ic TX j jc
fO
TX i ic TX j jc
= 10 Log r O
Interference reduction due to downlink segmentation:

If you select downlink segmentation support for the frame configuration that you are using, it means that the first
zone in the downlink, i.e., the DL PUSC zone, is segmented. All other zones are pooled together to form a group of
non-segmented zones. There are two effects of segmentation:
1. Power concentration, which means that the spectral density of the power transmitted over one segment is higher
than the spectral density of the same power transmitted over the entire channel bandwidth. The effect of power
concentration is visible when calculating the downlink C/(I+N). The power transmitted over a segmented zone has
1
--------------------------- times the spectral density of the power transmitted over the entire channel bandwidth. When
f Segment DL
1
calculating the C/(I+N) ratio, the increase in power by ---------------------------- is equivalent to decreasing the noise and
f Segment DL
726

AT332_TRR_E0
interference by f Segment DL . Hence, if downlink segmentation is used, the interference received at the pixel,
subscriber, or mobile Mi covered by the segmented zone is reduced by a factor of f Segment DL .
2. Collision probability between the subcarriers used by the subchannels belonging to the segment of the studied cell
and the subcarriers used by other sectors, segmented or not. The following paragraphs explain how the collision
probability is calculated.
The downlink segmentation usage (SU) ratio is the percentage of the total downlink traffic load present in the
segmented downlink PUSC zone. For example, if the downlink traffic load is 80 %, and the downlink segmentation
usage ratio is 50 %, then this means that the downlink traffic load of the segmented zone is 40 % (i.e., 50 % of 80 %),
and the downlink traffic load of the non-segmented zones is 40 %.
In coverage predictions, Atoll uses the downlink segmentation usage ratios stored in the cell properties for
determining the interference. In simulations, Atoll resets the downlink segmentation usage ratios for all the cells to
0, and then calculates the downlink segmentation usage ratios according to the traffic loads of the mobiles allocated
to the segmented zone and in the non-segmented zones.
Figure 10.3: Downlink Segmentation

Atoll determines the switching point between the segmented and the non-segmented zones using the downlink
segmentation usage ratio. The switching points between the segmented and non-segmented zones of the victim and
interfering cells, TXi(ic) and TXj(jc) respectively, are calculated as follows:
SP
SP
TX i ic
TX ic
i
SU DL
and
= ----------------------------------------------------------------------------------------------TX i ic
TX i ic
TX i ic
SU DL
+ f Segment DL 1 SU DL
TX j jc
SU DL
= ----------------------------------------------------------------------------------------------TX j jc
TX j jc
TX j jc
SU DL
+ f Segment DL 1 SU DL
TX jc
j
Where, SP is the switching point between the segmented and the non-segmented zones, SU is the downlink
segmentation usage ratios of the cells, and f Segment DL is downlink segmenting factor, which gives the bandwidth
used by a segment.
The downlink segmenting factor, f Segment DL , is calculated from the number of secondary subchannel groups
assigned to the first downlink PUSC permutation zone in the Permutation Zones table.
PSG + 2 SSGf Segment DL = 3
-------------------------------------------15
Where, PSG is the number of primary subchannel groups and SSG is the number of secondary subchannel groups.
The multiplicative coefficients of 3 and 2 are derived from the ratio of the numbers of
subchannels that belong to the primary and to the secondary subchannel gourps. For
example, for the FFT size of 1024 (or 2048), each primary subchannel group contains 6
(or 12) subchannels, and each secondary subchannel group contains 4 (or 8)
subchannels, which gives the ratio of 3:2. And, the denominator of 15 = 3 x 3 + 2 x 3.
727

If the downlink segmentation usage ratio is set to 0, it means that the segmented zone does not exist. Setting SU to 0
gives SP = 0, and setting SU to 1 gives SP = 1 (or 100%), which shows how the switching point varies with the downlink
segmentation usage ratio.
Derivation of the switching point formula: The downlink segmentation usage ratio is used
to partition the total downlink traffic load into segmented and non-segmented zones.
Therefore, the switching point formula is derived from the equation:
SU DL TL DL
1 SU DL TL DL
-------------------------------------------------------------------- = ----------------------------------------------SP fSegment DL W Channel
1 SP W Channel
With cells using downlink segmentation, there can be four different interference scenarios.
Between the segmented zone of the victim and the segmented zone of the interferer.
Between the segmented zone of the victim and the non-segmented zone of the interferer.
Between the non-segmented zone of the victim and the segmented zone of the interferer.
Between the non-segmented zone of the victim and the non-segmented zone of the interferer.
Figure 10.4: Downlink Segmentation Interference Scenarios

Therefore, Atoll calculates the probabilities of collision for each scenario and weights the total interference according
to the total collision probability. The probability of collision p Coll for each scenario is given by the following formula:
3 PSG Com + 2 SSG Com
p Coll = ---------------------------------------------------------------------TX i ic
TX i ic
3 PSG
+ 2 SSG
Where, PSGCom is the number of primary subchannel groups common in TXi(ic) and TXj(jc), SSGCom is the number of
secondary subchannel groups common in TXi(ic) and TXj(jc), PSG
in the cell TXi(ic), and SSG
TX i ic
TX i ic
is the number of primary subchannel groups
is the number of secondary subchannel groups in the cell TXi(ic).
The segment numbers and the cell permutation base numbers (Cell PermBase) are determined from the cells
preamble index. The mapping between the preamble index, the segment number, and Cell PermBase is available in
the IEEE specifications. This mapping is performed in Atoll as follows:
Preamble Index ( PI )
Range: 0 to 113
Cell PermBase ( PB )
Range: 0 to 31
Segment Number ( N Seg )
Range: 0, 1, 2
PI 96
96 PI 114
PI Modulo 32
PI 96
PI
Floor ------
32
PI 96 Modulo 3
There can be 2 cases for calculating the total probability of collision.
728

AT332_TRR_E0
Case 1: If the pixel, subscriber, or mobile Mi is covered by the segmented zone of TXi(ic), the total collision
probability for the pixel, subscriber, or mobile Mi is calculated as follows:
TX ic TX jc
i
j
p Collision DL
TX jc
TX ic
j
i
SS
p Coll
If SP
SP
TX jc
TX ic
TX jc
= SS
j
i
j
SN
+ p Coll SP
SP
TX j jc
TX i ic
p Coll SP
SP
If SP
------------------------------------------------------------------------------------------------------------TX ic
i
SP
Case 2: If the pixel, subscriber, or mobile Mi is covered by the non-segmented zone of TXi(ic), the total collision
probability for the pixel, subscriber, or mobile Mi is calculated as follows:
TX i ic TX j jc
p Collision DL
TX j jc
TX i ic
NN
p Coll
If SP
SP
TX j jc
TX jc
TX ic
+ p NS SP j SP i
= p NN
TX jc
TX ic
Coll 1 SP
Coll
---------------------------------------------------------------------------------------------------------------------------- If SP j SP i
TX ic
1 SP i
The interference reduction factor due to downlink segmentation for the pixel, subscriber, or mobile Mi is calculated
as follows:
TX i ic TX j jc
f Seg DL
TX ic TX jc
i
j
= 10 Log p Collision DL

Inter Tech
I DL
TX k
P DL Rec
--------------------------------------
F
TX
ic
TX
i
k
TX k ICP DL
TX k
F TX i ic TX k
ICPDL

TX k
Output
TX j jc
I DL
: Effective downlink traffic and pilot interference received at the pixel, subscriber, or mobile Mi from any
interfering cell TXj(jc).
Inter Tech
I DL
10.4.6.4 Traffic and Pilot C/N Calculation (DL)

Input
TX i ic
C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Traffic
and Pilot Signal Level Calculation (DL)" on page 719.
TX ic
i
C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Traffic
TX ic
i
n DL
: Downlink noise for the cell TXi(ic) as calculated in "Traffic and Pilot Noise Calculation (DL)" on page 721.
T AMS : AMS threshold defined for the cell TXi(ic).
TX i ic
729

M
T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used by Mis terminal.
or mobile Mi.
or mobile Mi.
TX i ic
N Ant TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic).
N Ant RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel, subscriber, or
mobile Mi.
Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel, subscriber, or mobile
Mi as calculated in "Permutation Zone Selection" on page 718.
i
BLER BDL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the

Calculations
The traffic and pilot C/N for a cell TXi(ic) are calculated as follows for any pixel, subscriber, or mobile Mi:
TX i ic
TX i ic
TX i ic
TX i ic
TX i ic
CNR Traffic = C Traffic n DL

TX i ic
CNR Pilot
= C Pilot n DL
The bearers available for selection in the pixel, subscriber, or mobile Mis WiMAX equipment are the ones:
Whose selection thresholds are less than the traffic or pilot C/N at Mi: T B CNR Traffic or T B CNR Pilot
TX ic
i
TX ic
i
Mi
If the cell supports MIMO, the STTD/MRC or SU-MIMO diversity gain, G Div DL , corresponding to the bearer is
applied to its selection threshold. The gain is read from the properties of the WiMAX equipment assigned to the
TX i ic
Mi
Mi
pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX , the subchannel allocation mode of PZ DL , Mobility M i ,
M
i
BLER B DL .

DL
The additional diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is also
applied. Therefore, the bearers available for selection are all the bearers defined in the WiMAX equipment for
M
DL
TX ic
i
Mi
DL
TX i ic
T B G Div DL G Div CNR Traffic

Mi
T B G Div DL G Div CNR Pilot
Bearer Index
730
Peak MAC Throughput

AT332_TRR_E0
From among the bearers available for selection, the selected bearer is the one with the highest downlink peak
MAC channel throughput as calculated in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput,
and Per-User Throughput Calculation" on page 747.
Effective MAC Throughput

MIMO STTD/MRC and SU-MIMO Diversity Gain:

Once the bearer is known, the traffic and pilot C/N calculated above become:
TX i ic
TX i ic
Mi
DL
TX i ic
Mi
DL
CNR Traffic = CNR Traffic + G Div DL + G Div

TX i ic
CNR Pilot
= CNR Pilot + G Div DL + G Div
Mi
Where G Div DL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer.
Output
TX i ic
CNR Traffic : Traffic C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
CNR Pilot : Pilot C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i ic
10.4.6.5 Traffic and Pilot C/(I+N) and Bearer Calculation (DL)

the studied cell (as explained in "Traffic and Pilot Signal Level Calculation (DL)" on page 719) at the pixel, subscriber, or mobile
under study. Next, Atoll calculates the interference received at the same studied pixel, subscriber, or mobile from all the
interfering cells (as explained in "Traffic and Pilot Interference Calculation (DL)" on page 722). Interference from each cell is
weighted according to the co- and adjacent channel overlap between the studied and the interfering cells, the traffic loads of
the interfering cells, and the probabilities of subcarrier collision if downlink segmentation is used. Finally, Atoll takes the ratio
of the signal level and the sum of the total interference from other cells and the downlink noise (as calculated in "Traffic and
Pilot Noise Calculation (DL)" on page 721).
Input
TX i ic
C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Traffic
TX i ic
C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Traffic
TX i ic
n DL
TX jc
j
I DL
: Downlink noise for the cell TXi(ic) as calculated in "Traffic and Pilot Noise Calculation (DL)" on page 721.
: Effective downlink traffic and pilot interference from any cell TXj(jc) calculated for a pixel, subscriber, or
mobile Mi covered by a cell TXi(ic) as explained in "Traffic and Pilot Interference Calculation (DL)" on page 722.
NR DL
T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used by Mis terminal.
Inter Tech
TX i ic
Mi
M
or mobile Mi.
Mi
or mobile Mi.
731

TX ic
i
mobile Mi.
Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel, subscriber, or mobile
Mi
i
BLER BDL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the

Inter Tech
I DL
: Downlink inter-technology interference as calculated in "Traffic and Pilot Interference Calculation (DL)"
on page 722.
Calculations
The traffic and pilot C/(I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX i ic
CINR Traffic
TX i ic
CINR Pilot
TXj jc
TX i ic
n
IDL
DL
Inter Tech
Inter Tech
-
-----------------and
+ -------------------+ NR DL
= C Traffic 10 Log
10 + I DL
10
10
10
All TXj jc
TX i ic
TX i ic
= C Pilot
TXj jc
TX i ic
n DL
IDL
Inter Tech
Inter Tech
-
-----------------10 + I DL
+ -------------------+ NR DL
10 Log
10
10
10
All TXj jc
The Traffic Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX ic
TX i ic
I + N DL
TX j jc
n DL
I DL
---------------------
Inter
Tech
10
----------------- + NR Inter Tech
= 10 Log
+ 10
10 + I DL
DL
10
All TX j jc
The bearers available for selection in the pixel, subscriber, or mobile Mis WiMAX equipment are the ones:
Whose selection thresholds are less than the traffic or pilot C/(I+N) at Mi: T B CINR Traffic or T B CINR Pilot
Mi
TX i ic
TX i ic
Mi
If the cell supports MIMO, the STTD/MRC or SU-MIMO diversity gain, G Div DL , corresponding to the bearer is
applied to its selection threshold. The gain is read from the properties of the WiMAX equipment assigned to the
TX ic
i
pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX , the subchannel allocation mode of PZ DL , Mobility M i ,
M
i
BLER B DL .

DL
Mi
Mi
DL
TX i ic
T B G Div DL G Div CINR Traffic
732

AT332_TRR_E0
M
TX ic
i
DL
T B G Div DL G Div CINR Pilot
Bearer Index
Peak MAC Throughput

From among the bearers available for selection, the selected bearer is the one with the highest downlink peak

MIMO STTD/MRC and SU-MIMO Diversity Gain:

Once the bearer is known, the traffic and pilot C/(I+N) calculated above become:
TX i ic
TX i ic
Mi
TX ic
i
DL
CINR Traffic = CINR Traffic + G Div DL + G Div

TX ic
i
CINR Pilot
DL
= CINR Pilot + G Div DL + G Div
Mi
Output
TX i ic
CINR Traffic : Traffic C/(I+N) from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
CINR Pilot : Pilot C/(I+N) from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
I + N DL
TX i ic
TX i ic
: Traffic Total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by a cell
TXi(ic).
10.4.6.6 Traffic Signal Level Calculation (UL)

Input
Mi
P Max : Maximum transmission power of the terminal used by the pixel, subscriber, or mobile Mi without power
control.
P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi after power control as
Mi
calculated in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 740.

TX i
TX
Without smart antennas: G
With smart antennas: G
TX i
TX i
is the transmitter antenna gain, i.e., G
TX i
TX i
= G Ant .
is the uplink smart antenna beamforming gain, i.e., G
TX i
TX
i
= G SA = 10 Log E SA .
For more information on the calculation of G SA , refer to section "Beamforming Smart Antenna Models" on
page 43.
TX i

TX
TX i
= L Total UL ).
733

TX
is selected.
Mi
Mi
Mi
Mi
Mi
Calculations
The received traffic signal level (dBm) from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated as follows:
Mi
Mi
C UL = EIRP UL L Path M Shadowing Model L Indoor + G
TX i
TX i
Mi
Mi
L Ant L Body
Mi
EIRP UL = P
With P
Mi
Mi
+G
Mi
Mi
Mi
= P Max without power control at the start of the calculations, and is the P
Mi
Mi
= P Eff after power control.
Output
Mi
C UL : Received uplink signal level from the pixel, subscriber, or mobile Mi at a cell TXi(ic).
10.4.6.7 Traffic Noise Calculation (UL)

For determining the uplink C/N and C/(I+N), Atoll calculates the uplink noise over the channel bandwidth used by the cell. The
used bandwidth depends on the number of used subcarriers. The numbers of subcarriers used by different permutation zones
can be different.
The uplink noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on the
temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the used
bandwidth.
Input
734
PZ
Mi
UL
N SCa Used : Number of subcarriers used by the uplink permutation zone of a cell TXi(ic) assigned to Mi.
nf
TX i ic
TX i ic
TX i ic

AT332_TRR_E0
Calculations
n 0 = 10 Log K T 1000 = 174 dBm/Hz
Mi
TX i ic
n 0 UL
PZ UL
N SCa Used
TXi ic
= n 0 + 10 Log F Sampling ------------------------
TX i ic
N
SCa Total
TX i ic
n UL
TX i ic
= n 0 UL + nf
TX i ic
Output
TX i ic
n UL
: Uplink noise for the cell TXi(ic).
10.4.6.8 Traffic Interference Calculation (UL)

The uplink traffic interference is only calculated during Monte Carlo simulations. In coverage predictions, the uplink noise rise
values already available in simulation results or in the Cells table are used.
The interference received by a cell TXi(ic) from an interfering mobile covered by a cell TXj(jc) can be defined as the uplink signal
level received from interfering mobiles Mj depending on the overlap that exists between the channels used by the cells TXi(ic)
and TXj(jc), on the traffic loads of the interfering mobile Mj.
10.4.6.8.1
Calculation of the uplink interference from each individual interfering mobile as explained in "Traffic Interference
Signal Levels Calculation (UL)" on page 735.
Calculation of the uplink noise rise which represents the total uplink interference from all the interfering mobiles as
Traffic Interference Signal Levels Calculation (UL)

Input
Mj
C UL : Uplink signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) as calculated in


In Monte Carlo simulations, interfering signal levels already include M Shadowing Model , as explained in "Traffic Signal
Level Calculation (UL)" on page 733.
information, see "Shadow Fading Model" on page 90). As the interfering signal levels already include
Mj
Mj
C UL = C UL + M Shadowing C I
is selected.
TX i ic TX j jc
rO
Mj

Traffic loads are calculated during Monte Carlo simulations as explained in "Scheduling and Radio Resource
735

Calculations
M
TX ic TX jc
i
j
I UL = C UL + f O
TX ic TX jc
i
j
+ f TL UL + f Seg UL
Calculations for the interference reduction factors due to channel overlapping, uplink traffic load, and uplink segmentation
are explained below:
as follows:
TX i ic TX j jc
fO
TX i ic TX j jc
= 10 Log r O
Interference reduction due to interfering mobiles traffic load:

The interference reduction factor due to the interfering mobiles uplink traffic load is calculated as follows:
M
j
j
Interference reduction due to uplink segmentation:

If you select uplink segmentation support for the frame configuration that you are using, it means that the first zone
in the uplink, i.e., the UL PUSC zone, is segmented. All other zones are pooled together to form a group of nonsegmented zones. The interference reduction factor due to uplink segmentation is calculated as follows:
TX i ic TX j jc
f Seg UL
TX ic TX jc
i
j
= 10 Log p Collision UL
TX i ic TX j jc
Where p Collision UL
is the collision probability between the subcarriers of the uplink segments being used by the
interfered and interfering cells. It is determined during Monte Carlo simulations as follows:
TX i ic TX j jc
p Collision UL
SC Com
= -----------------TX i ic
SC
Where, SCCom is the number of subchannels common in TXi(ic) and TXj(jc), SC

the cell TXi(ic).
TX i ic
is the number of subchannels in
The segment numbers and the cell permutation base numbers (Cell PermBase) are determined from the cells
preamble index. The mapping between the preamble index, the segment number, and Cell PermBase is available in
the IEEE specifications. This mapping is performed in Atoll as follows:
Preamble Index ( PI )
Range: 0 to 113
Cell PermBase ( PB )
Range: 0 to 31
Segment Number ( N Seg )
Range: 0, 1, 2
PI 96
96 PI 114
PI Modulo 32
PI 96
PI
Floor ------
32
PI 96 Modulo 3
In Monte Carlo simulations, Atoll calculates two separate noise rise values; for the mobiles served by the segmented
zone of the interfered cell Atoll calculates the uplink segmented noise rise, and for the mobiles served by the nonsegmented zones of the interfered cell Atoll calculates the uplink noise rise.
In coverage predictions, point analysis, and calculations on subscriber lists, according to the zone, segmented or nonsegmented, that covers the pixel, receiver, or subscriber, Atoll uses either the uplink segmented noise rise or the
uplink noise rise to calculate the C/(I+N). For more information on the calculation of the uplink noise rise, see "Noise
Rise Calculation (UL)" on page 737.
Output
736
Mj
I UL : Uplink interference signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc).

AT332_TRR_E0
10.4.6.8.2

The uplink noise rise is defined as the ratio of the total uplink interference received by any cell TXi(ic) from interfering mobiles
Mj present in the coverage areas of other cells TXj(jc) to the uplink noise of the cell TXi(ic). In other words, it is the ratio (I+N)/N.
Input
Mj
I UL : Uplink interference signal levels received at a cell TXi(ic) from interfering mobiles Mj covered by other cells TXj(jc)
as calculated in "Traffic Interference Signal Levels Calculation (UL)" on page 735.
TX i ic
n UL
NR UL
: Uplink noise for the cell TXi(ic) as calculated in "Traffic Noise Calculation (UL)" on page 734.
Inter Tech
Calculations
The uplink noise rise and total noise (I+N) for the cell TXi(ic) are calculated as follows:
Without smart antennas:

For any mobile Mi covered by a non-segmented zone in the interfered cell TXi(ic), Atoll calculates the UL noise rise as
follows:
TX i ic
NR UL
TX i ic
IMj
n UL
UL
non-seg M i
---------------------
TX i ic
Inter Tech
10
-------------------------------------------= 10 Log
+ NRUL
n UL
10
10
+ 10
All Mj
All
TX
jc
For any pixel, subscriber, or mobile Mi covered by the non-segmented zone in the interfered cell TXi(ic), Atoll
calculates the uplink total noise (I+N) as follows:
TX i ic
I + N UL
TX i ic
= NR UL
TX i ic
+ n UL
For any mobile Mi covered by the segmented zone in the interfered cell TXi(ic), Atoll calculates the segmented zone
UL noise rise as follows:
TX i ic
NR UL Seg
TX i ic
IMj
n
UL
UL
seg M i
---------------------
TX ic
10
-------------------------------- + NR Inter Tech n i
= 10 Log
10
10
+ 10
UL
UL
All M j
All TX j jc
For any pixel, subscriber, or mobile Mi covered by the segmented zone in the interfered cell TXi(ic), Atoll calculates
the uplink total noise (I+N) as follows:
TX i ic
I + N UL
TX i ic
TX i ic
= NR UL Seg + n UL

The angular distribution of the uplink noise rise is calculated during Monte Carlo simulations and can be stored in the
Cells table in order to be used in coverage predictions. The angular distribution of the uplink noise rise is given by:
2
I UL + n I
NR UL = --------------------------------2
n I
TX i ic
I + N UL
= I UL + n I
Output
TX i ic
NR UL
: Non-segmented uplink noise rise for the cell TXi(ic).
NR UL Seg : Segmented uplink noise rise for the cell TXi(ic).
TX i ic
737

TX ic
i
: Angular distribution of the uplink noise rise for the cell TXi(ic).
NRUL
I + N UL
TX ic
i
TX ic
i
or I + N UL
: Total Noise for a cell TXi(ic) calculated for any pixel, subscriber, or mobile Mi.
10.4.6.9 Traffic C/N Calculation (UL)

Input
Mi
C UL : Received uplink signal level from the pixel, subscriber, or mobile Mi at its serving cell TXi(ic) as calculated in
TX ic
i
n UL
: Uplink noise for the cell TXi(ic) as calculated in "Traffic Noise Calculation (UL)" on page 734.
TX i ic
T B Lowest : Bearer selection threshold of the lowest bearer in the WiMAX equipment assigned to the cell TXi(ic).
TX i ic
Mi
PZ UL
N SC
: Number of subchannels per channel defined for the uplink permutation zone assigned to the pixel, subscriber,
or mobile Mi as calculated in "Permutation Zone Selection" on page 718.

Mi
PZ UL = 8
N SC Seg : Number of subchannels per segment for the first uplink PUSC permutation zone.
M PC : Power control margin defined in the global network settings.
T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used bythe cell TXi(ic).
Mi
M
Mi
Mi
or mobile Mi.
Mi
or mobile Mi.
Mi
N Ant TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel, subscriber, or
mobile Mi.
TX ic
i
N Ant RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic).
Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel, subscriber, or mobile
Mi
i
BLER BUL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell

TXi(ic).
Calculations
The uplink C/N from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated as follows:
Mi
Mi
TX i ic
CNR UL = C UL n UL
The bearers available for selection in the cell TXi(ic)s WiMAX equipment are the ones:
738

AT332_TRR_E0
Whose selection thresholds are less than the uplink C/N at Mi: T B CNR UL
TX ic
i
If the cell supports MIMO, the STTD/MRC, SU-MIMO diversity or MU-MIMO diversity gain, G Div UL ,
Mi
TX i ic
Mi
WiMAX equipment assigned to the cell TXi(ic) for N Ant RX , N Ant TX , the subchannel allocation mode of PZ UL ,
M
i
Mobility M i , BLER B UL .

UL
Mi
TX i ic
Mi
UL
T B G Div UL G Div CNR UL

Bearer Index
Peak MAC Throughput

From among the bearers available for selection, the selected bearer is the one with the highest uplink peak MAC
Per-User Throughput Calculation" on page 747.

MIMO STTD/MRC, SU-MIMO Diversity, and MU-MIMO Diversity Gain:

Once the bearer is known, the uplink C/N calculated above becomes:
Mi
Mi
TX i ic
UL
CNR UL = CNR UL + G Div UL + G Div

TX i ic
Where G Div UL is the STTD/MRC, SU-MIMO diversity, or MU-MIMO diversity gain corresponding to the selected
bearer.
Uplink Subchannelisation:
The uplink subchannelisation depends on the uplink bandwidth allocation target defined for the scheduler used by the
cell TXi(ic). The uplink C/N calculated above is given for the total number of subchannels associated with the
PZ
permutation zone, i.e., N SC
Mi
UL
. Subchannelisation is performed for all the pixels, subscribers, or mobiles in the uplink,
and may reduce the number of used subchannels in order to satisfy the selected target.
Full Bandwidth
transmission, there is no gain in the uplink C/N.
Maintain Connection
The bandwidth used for transmission by a mobile is reduced only if the uplink C/N is not enough to even access
the lowest bearer. For example, as a mobile moves from good to bad radio conditions, the number of subchannels
used by it for transmission in uplink are reduced one by one in order to improve the uplink C/N. The calculation of
the gain introduced by the subchannelisation is explained below.
Best Bearer
The bandwidth used for transmission by a mobile is reduced in order to improve the uplink C/N enough to access
the best bearer. For example, if using 5 subchannels, a mobile is able to access the best bearer, and using 6 it would
only get access to the second best, it will be assigned 5 subchannels as the used uplink bandwidth. Although using
4 subchannels, its uplink C/N will be better than when using 5, the uplink bandwidth is not reduced to 4 because
739

it does not provide any gain in terms of the bearer, i.e., the mobile already has the best bearer using 5
subchannels. The calculation of the gain introduced by the bandwidth reduction is explained below.
i.e., bearer with the highest index, with the highest peak MAC throughput, or with the highest effective MAC
throughput.
The uplink subchannelisation may result in the use of a number of subchannels which is less than the total number of
subchannels associated with the permutation zone. The gain related to this bandwidth reduction is applied to the
uplink C/N:
M
i
CNR UL
Final
PZUL
M
N SC
i
= CNR UL+ 10 Log ----------------
N Mi
All SC
SC
UL
Mi
PZ UL
Mi
Min
Where N SC UL Service N SC UL N SC
for any pixel, subscriber, or mobile Mi covered by a non-segmented
permutation zone in the interfered cell TXi(ic), and
Min
N SC UL Service
Mi
M
i
PZ UL = 8
N SC UL N SC Seg for any pixel, subscriber,
or mobile Mi covered by the segmented uplink PUSC zone in the interfered cell TXi(ic).
Uplink Power Control:
Once the subchannelisation is performed, Atoll continues to work with the C/N given by the subchannelisation, i.e.,
Mi
Mi
CNR UL = CNR UL .
Final
The pixel, subscriber, or mobile Mi reduces its transmission power so that the uplink C/N from it at its cell is just
enough to get the selected bearer.
If with P
Mi
Mi
Mi
= P Max AND CNR UL T
TX i ic
Mi
+ M PC , where T
B UL
TX i ic
Mi
B UL
is the bearer selection threshold, from the WiMAX
equipment assigned to the cell TXi(ic), for the bearer selected for the pixel, subscriber, or mobile Mi.
Mi
M i TX ic
Mi
Mi
i
P Eff = Max PMax CNR UL T M + M PC P Min
B i
UL
Mi
Mi
CNR UL is calculated again using P Eff .

Output
Mi
CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
10.4.6.10 Traffic C/(I+N) and Bearer Calculation (UL)

or mobile as explained in "Traffic Signal Level Calculation (UL)" on page 733. Next, Atoll calculates the uplink carrier to noise
ratio as explained in "Traffic C/N Calculation (UL)" on page 738. Finally, determines the uplink C/(I+N) by dividing the
previously calculated uplink C/N by the uplink noise rise value of the cell as calculated in "Noise Rise Calculation (UL)" on
page 737.
740

AT332_TRR_E0
Input
M
CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic) as calculated in "Traffic C/N
NR UL
TX ic
i
: Non-segmented uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on
page 737.
TX i ic
NR UL Seg : Segmented uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 737.
NR UL
TX i ic
: Angular distribution of the uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation
(UL)" on page 737.

TX i ic
T B Lowest : Bearer selection threshold of the lowest bearer in the WiMAX equipment assigned to the cell TXi(ic).
TX i ic
Mi
PZ UL
N SC

Mi
PZ UL = 8
N SC Seg : Number of subchannels per segment for the first uplink PUSC permutation zone.
T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used bythe cell TXi(ic).
Mi
M
Mi
Mi
or mobile Mi.
Mi
or mobile Mi.
Mi
mobile Mi.
TX ic
i
Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel, subscriber, or mobile
Mi
i
BLER B UL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell

TXi(ic).
Calculations
The uplink C/(I+N) for any pixel, subscriber, or mobile Mi at a cell TXi(ic) is calculated as follows:
Without smart antennas:

For any pixel, subscriber, or mobile Mi covered by the non-segmented zone in the interfered cell TXi(ic):
Mi
Mi
TX i ic
CINR UL = CNR UL NR UL
For any pixel, subscriber, or mobile Mi covered by the segmented zone in the interfered cell TXi(ic):
Mi
Mi
TX i ic
CINR UL = CNR UL NR UL Seg
741

Monte Carlo simulations: The uplink C/(I+N) is calculated as described in the section "Beamforming Smart
Antenna Models" on page 43. Victim and interfering mobiles are generated by a time-slot scenario as explained
in "Simulation Process" on page 703.
Coverage predictions: CINR UL = CNR UL NR UL
TX ic
i
The bearers available for selection in the cell TXi(ic)s WiMAX equipment are the ones:
Whose selection thresholds are less than the uplink C/(I+N) at Mi: T B CINR UL and T B CINR UL
Mi
Mi
Mi
Mi
TX ic
i
M
TX ic
i
WiMAX equipment assigned to the cell TXi(ic) for N Ant RX , N Ant TX , the subchannel allocation mode of PZ UL ,
M
i
Mobility M i , BLER B UL .

UL
Mi
TX i ic
UL
Mi
Mi
TX i ic
UL
Mi
T B G Div UL G Div CINR UL and

T B G Div UL G Div CINR UL
Bearer Index
Peak MAC Throughput

From among the bearers available for selection, the selected bearer is the one with the highest uplink peak MAC


Once the bearer is known, the uplink C/(I+N) calculated above becomes:
M
TX ic
i
UL

Mi
TX i ic
Mi
UL
CINR UL = CINR UL + G Div UL + G Div and

Mi
Mi
TX i ic
UL
CINR UL = CINR UL + G Div UL + G Div

TX i ic
bearer.
Uplink Subchannelisation:
742

AT332_TRR_E0
The uplink subchannelisation depends on the uplink bandwidth allocation target defined for the scheduler used by the
cell TXi(ic). The uplink C/(I+N) calculated above is given for the total number of subchannels associated with the
M
PZ
permutation zone, i.e., N SC
i
UL
. Subchannelisation is performed for all the pixels, subscribers, or mobiles in the uplink,
and may reduce the number of used subchannels in order to satisfy the selected target.
Full Bandwidth
transmission, there is no gain in the uplink C/(I+N).
Maintain Connection
The bandwidth used for transmission by a mobile is reduced only if the uplink C/(I+N) is not enough to even access
the lowest bearer. For example, as a mobile moves from good to bad radio conditions, the number of subchannels
used by it for transmission in uplink are reduced one by one in order to improve the uplink C/(I+N). The calculation
of the gain introduced by the subchannelisation is explained below.
Best Bearer
The bandwidth used for transmission by a mobile is reduced in order to improve the uplink C/(I+N) enough to
access the best bearer. For example, if using 5 subchannels, a mobile is able to access the best bearer, and using
6 it would only get access to the second best, it will be assigned 5 subchannels as the used uplink bandwidth.
Although using 4 subchannels, its uplink C/(I+N) will be better than when using 5, the uplink bandwidth is not
reduced to 4 because it does not provide any gain in terms of the bearer, i.e., the mobile already has the best
bearer using 5 subchannels. The calculation of the gain introduced by the bandwidth reduction is explained below.
i.e., bearer with the highest index, with the highest peak MAC throughput, or with the highest effective MAC
throughput.
The uplink subchannelisation may result in the use of a number of subchannels which is less than the total number of
subchannels associated with the permutation zone. The gain related to this bandwidth reduction is applied to the
uplink C/(I+N):
Mi
Mi
CINR UL
Final
Where
PZUL
Mi
N SC
= CINR UL+ 10 Log ----------------
NMi
All SC
SC UL
Min
N SC UL Service
M
i
PZ UL
Mi
N SC UL N SC
for any pixel, subscriber, or mobile Mi covered by a non-segmented
permutation zone in the interfered cell TXi(ic), and
Min
N SC UL Service
i
N SC UL
PZ
Mi
UL
=8
N SC Seg for any pixel, subscriber,
or mobile Mi covered by the segmented uplink PUSC zone in the interfered cell TXi(ic).
Once the subchannelisation is performed, Atoll continues to work with the C/(I+N) given by the subchannelisation,
Mi
Mi
i.e., CINR UL = CINR UL .

Final
The pixel, subscriber, or mobile Mi reduces its transmission power so that the uplink C/(I+N) from it at its cell is just
If with P
Mi
Mi
Mi
= P Max AND CINR UL T
TX i ic
Mi
+ M PC , where T
B UL
TX i ic
Mi
B UL
is the bearer selection threshold, from the WiMAX
Mi
M i TX i ic
Mi
Mi
P Eff = Max P Max CINR UL T M + M PC P Min
B i
UL
Mi
Mi
CINR UL is calculated again using P Eff .
743

Output
M
CINR UL or CINR UL : Uplink C/(I+N) from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
N SC UL : Number of subchannels used by the pixel, subscriber, or mobile Mi in the uplink after subchannelisation.
Mi
M

page 744.
Calculation of throughputs as explained in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and

The total amount of resources in a cell is the number of modulation symbols that can be used for data transfer in each frame.
The total cell resources can be calculated separately for the downlink and the uplink subframes. The following sections
describe how the cell capacities are calculated for TDD and FDD networks.
10.4.7.1.1
Calculation of Sampling Frequency

Input
TX i ic
f Sampling : Sampling factor defined for the frequency band of the cell TXi(ic).
W Channel : Channel bandwidth of the cell TXi(ic).
TX i ic
Calculations
Atoll determines the sampling frequency as follows:
TX i ic
TX i ic
W Channel 10
- 8000
F Sampling = Floor f Sampling ----------------------------------8000
Output
10.4.7.1.2
TX i ic
F Sampling : Sampling frequency for the cell TXi(ic).
Calculation of Symbol Duration

Input
TX i ic
r CP
TX ic
i
TX i ic
: Cyclic prefix ratio defined for the frame configuration of TXi(ic) or, otherwise, in the global network settings.
Calculations
From the sampling frequency, Atoll determines the inter-subcarrier spacing.
F
TX i ic
TX i ic
F Sampling 10
= ------------------------------------TX i ic
N SCa Total
Atoll calculates the useful symbol duration.
744

AT332_TRR_E0
TX ic
i
1
D Sym Useful = ------------------TX ic
i
F
And, the duration of the cyclic prefix.

TX ic
i
D CP
TX ic
i
r CP
= -------------F
Adding the Cyclic prefix ratio to the useful symbol duration, Atoll determines the total symbol duration.
TX i ic
TX i ic
TX i ic
D Symbol = D Sym Useful + D CP

Output
10.4.7.1.3
TX i ic
D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic).
Calculation of Total Cell Resources - TDD Networks

Input
D TTG : TTG duration.
D RTG : RTG duration.
D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic) as calculated in "Calculation of Symbol
Duration" on page 744.
r DL Frame : DL ratio.
N SD DL : Number of symbol durations that correspond to the downlink subframe.
N SD UL : Number of symbol durations that correspond to the uplink subframe.
O Fixed : Downlink fixed overhead.
O Variable : Downlink variable overhead.
O Fixed : Uplink fixed overhead.
O Variable : Uplink variable overhead.
TDD
TDD
TX ic
i
TDD
TDD
TDD
DL
DL
UL
UL
Mi
PZ DL
N SCa Data : Number of data subcarriers of the downlink permutation zone of a cell TXi(ic) assigned to Mi.
Mi
PZ UL
N SCa Data : Number of data subcarriers of the uplink permutation zone of a cell TXi(ic) assigned to Mi.
Calculations
The downlink and the uplink subframes of a TDD frame are separated in time by the TTG and the RTG time guards.
First of all, Atoll calculates the useful frame duration by removing the TTG and RTG from the frame duration:
Used
TDD
TDD
D Frame = D Frame D TTG D RTG
Then, Atoll calculates the frame duration in terms of number of symbol durations:
D Used
TX i ic
Frame
N SD Used Frame = Floor ---------------- TXi ic
D Symbol
Next, Atoll calculates the downlink and uplink cell capacities as follows:
Downlink Subframe:
Atoll calculates the number of symbol durations in the downlink subframe excluding the fixed overhead defined in the
global network settings:
745

TX ic

TX ic
i
i
TDD
DL
N SD DL Subframe = RoundUp N SD Used Frame r DL Frame O Fixed if DL:UL ratio is defined in percentage.
TDD
TX ic
TXi ic
N SD DL
i
DL
Or N SD DL Subframe = RoundUp N SD Used Frame ----------------------------------------- O Fixed if DL:UL ratio is defined in
TDD
TDD
N SD DL + N SD UL
fraction.
The RoundUp function rounds a float value up to the nearest integer value.
The total number of symbols in the downlink subframe after removing the variable overhead is:
TX ic
i
R DL
TX ic
i
N Sym DL Subframe
Mi
DL
PZ
O Variable
TXi ic
DL
= Floor N SD DL Subframe N SCa Data 1 ---------------------

100
Uplink Subframe:
Atoll calculates the number of symbol durations in the uplink subframe excluding the fixed overhead defined in the
global network settings:
TX i ic
TX i ic
N SD UL Subframe = RoundDown N SD Used Frame 1 r DL Frame O Fixed
percentage.
TDD
UL
if DL:UL ratio is defined in
TDD
TX ic
TX i ic
N SD UL
i
UL
Or N SD UL Subframe = RoundDown N SD Used Frame ----------------------------------------- O Fixed if DL:UL ratio is defined in
TDD
TDD
N SD DL + N SD UL
fraction.
The RoundDown function rounds a float value down to the nearest integer value.
The total number of symbols in the uplink subframe after removing the variable overhead is:
TX i ic
R UL
TX i ic
= N Sym UL Subframe
Mi
UL
PZ UL
O Variable
TX i ic
-
= Floor N SD UL Subframe N SCa Data 1 -------------------
100
Output
10.4.7.1.4
TX i ic
R DL
R UL
TX i ic
TX i ic
= N Sym DL Subframe : Amount of downlink resources in the cell TXi(ic).

TX i ic
= N Sym UL Subframe : Amount of uplink resources in the cell TXi(ic).
Calculation of Total Cell Resources - FDD Networks

The total cell resources calculation is the same for downlink and uplink subframes in FDD networks. Therefore, the symbol X
is used to represent DL or UL in the expressions below.
Input
D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic) as calculated in "Calculation of Symbol
TX ic
i
Duration" on page 744.

X
O Fixed : Downlink or uplink fixed overhead.
O Variable : Downlink or uplink variable overhead.
Mi
PZ X
N SCa Data : Number of data subcarriers of the downlink or uplink permutation zone of a cell TXi(ic) assigned to Mi.
Calculations
There are no transmit and receive time guards in FDD systems. Therefore, the downlink and the uplink subframe durations
are the same as the frame duration.
746

AT332_TRR_E0
X
D Subframe = D Frame
The subframe durations in terms of the number of symbol durations excluding the fixed overheads are:
DX
TX ic
i
Subframe
- O XFixed
N SD X Subframe = Floor ---------------------TX ic
i
D Symbol
The total numbers of symbols in the downlink or uplink subframes after removing the variable overheads are:
TX ic
i
RX
TX ic
i
N Sym X Subframe
Mi
X
PZ
O Variable
TXi ic
X
= Floor N SD X Subframe N SCa Data 1 ---------------------

100
Output
TX i ic
TX i ic
= N Sym X Subframe : Amount of downlink or uplink resources in the cell TXi(ic).
RX
10.4.7.2 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput,

and Per-User Throughput Calculation
capacities are similar to channel throughputs but upper-bound by the maximum downlink and uplink traffic loads. Allocated
bandwidth throughputs are calculated for the number of used subchannels in uplink allocated to the pixel, subscriber, or
mobile Mi. Per-user throughputs are calculated by dividing the downlink cell capacities or uplink allocated bandwidth
throughputs by the average number of downlink or uplink users defined for the cell, respectively.
Input
TX i ic
R DL
TX i ic
TX i ic
page 744.
TX i ic
page 744.
M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the uplink in
R UL
i
B DL
i
UL

f Segment DL : Downlink segmenting factor for the first downlink PUSC zone as calculated in "Effective Traffic and Pilot
Interference Calculation (DL)" on page 726.
TX i ic
CNR Preamble : Preamble C/N the cell TXi(ic) as calculated in "Preamble C/N Calculation" on page 715.
T MU MIMO : MU-MIMO threshold defined for the cell TXi(ic).
G MU MIMO : MU-MIMO gain defined for the cell TXi(ic).
i
i
BLER B DL : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the WiMAX equipment
TX i ic
TX i ic
TX i ic
TX ic
747

i
i
BLER BUL : Uplink block error rate read from the BLER vs. CINR UL graph available in the WiMAX equipment assigned

to the cell TXi(ic).
M
Mi.
Mi
PZ
N SC
Mi
UL
Mi
N SC UL : Number of uplink subchannels after subchannelisation with which the pixel, subscriber, or mobile Mi can get
the highest available bearer, as calculated in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 740.
TX i ic
TX ic
i
Calculations
Downlink:
TX i ic
Peak MAC Channel Throughput:
i
CTP P DL
R DL
Mi
DL
= --------------------------------D Frame
TX i ic
For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG DL read
from the scheduler properties for the Mobility M i and the number of users connected to the cell in downlink.
TX i ic
R DL
Mi
Mi
TX ic
i
DL
CTP P DL = -------------------------------- G MUG DL
D Frame
TX i ic
Mi
Max
G MUG DL = 1 if CINR Traffic CINR MUG

number of users.
Downlink Segmentation:
Mi
If the permutation zone assigned to the pixel, subscriber, or mobile Mi is the first downlink PUSC zone ( PZ DL = 0 )
and it is segmented, the channel throughput is calculated as:
Mi
Mi
CTP P DL = CTP P DL f Segment DL

MIMO SU-MIMO Gain:
If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports SU-MIMO or AMS, SU-MIMO gain
Max
G SU MIMO is applied to the bearer efficiency. The gain is read from the properties of the WiMAX equipment assigned
to the pixel, subscriber, or mobile Mi for:
TX i ic
N Ant RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel, subscriber,
Mi
or mobile Mi.
748

AT332_TRR_E0
Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel, subscriber, or
mobile Mi as calculated in "Permutation Zone Selection" on page 718.
B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink as explained in "Traffic and Pilot C/
i
BLER B DL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the

TX ic
i
terminal used by the pixel, subscriber, or mobile Mi. BLER is determined for CINR Traffic .
In case of SU-MIMO:
B DL
In case of AMS:
Mi
Mi
B DL
TX i ic
Max
B DL
Max
Mi
Mi
TX i ic
1 + f SU MIMO G SU MIMO 1 if CNR Preamble T AMS
TX i ic
TX i ic
B DL
i
i
i
Effective MAC Channel Throughput: CTP E DL = CTP P DL 1 BLER B DL
Mi
f TP Scaling
- TP Offset
= CTP E DL -----------------------100
Mi
Mi
Application Channel Throughput: CTP A DL
Peak MAC Cell Capacity: Cap P DL = CTP P DL TL DL Max
i
i
i
Effective MAC Cell Capacity: Cap E DL = Cap P DL 1 BLER B DL
Mi
TX i ic
Mi
Mi
Application Cell Capacity: Cap A DL
Mi
f TP Scaling
- TP Offset
= Cap E DL -----------------------100
Mi
Mi
Peak MAC Throughput per User: PUTP P DL
Mi
Cap P DL
= ----------------------TX i ic
N Users DL
M
Mi
Cap E DL
Effective MAC Throughput per User: PUTP E DL = ----------------------TX ic
i
N Users DL
Mi
Application Throughput per User: PUTP A DL
Mi
Mi
f TP Scaling
- TP Offset
= PUTP E DL -----------------------100
Mi
Uplink:
TX ic
i
Mi
R UL
Mi
B UL
Peak MAC Channel Throughput: CTP P UL = --------------------------------D Frame

TX i ic
For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG UL read
from the scheduler properties for the Mobility M i and the number of users connected to the cell in uplink.
TX i ic
R UL
Mi
Mi
TX ic
i
UL
CTP P UL = -------------------------------- G MUG UL
D Frame
749

TX ic
i
Max
G MUG UL = 1 if CINR UL CINR MUG

number of users.
MIMO SU-MIMO Gain:
If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports SU-MIMO or AMS, SU-MIMO gain
Max
G SU MIMO is applied to the bearer efficiency. The gain is read from the properties of the WiMAX equipment assigned
to the cell TXi(ic) for:
N Ant TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel, subscriber,
or mobile Mi.
TX i ic
Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel, subscriber, or
Mi
mobile Mi as calculated in "Permutation Zone Selection" on page 718.
B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink as explained in "Traffic C/(I+N) and Bearer
i
BLER B UL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell

Mi
TXi(ic). BLER is determined for CINR UL .

In case of SU-MIMO:
B UL
In case of AMS:
Mi
B UL
Max
Mi
Mi
1 + fSU MIMO G SU MIMO 1
B UL
TX i ic
Max
Mi
TX i ic
1 + f SU MIMO G SU MIMO 1 if CNR Preamble T AMS
TX i ic
TX i ic
B UL
MIMO MU-MIMO Gain (for uplink throughput coverage predictions only):
If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports MU-MIMO and
TX i ic
TX i ic
TX i ic
TX i ic
CNR Preamble T MU MIMO and N Ant RX 2 , the MU-MIMO gain G MU MIMO is applied to the channel throughput.
The MU-MIMO gain is read from the properties of the cell TXi(ic).
Mi
Mi
TX i ic
CTP P UL = CTP P UL G MU MIMO
Mi
Mi
f TP Scaling
- TP Offset
= CTP E UL -----------------------100
Mi
Mi
Application Channel Throughput: CTP A UL
Peak MAC Cell Capacity: Cap P UL = CTP P UL TL UL Max
i
i
i
Effective MAC Cell Capacity: Cap E UL = Cap P UL 1 BLER B UL
750
i
i
i
Effective MAC Channel Throughput: CTP E UL = CTP P UL 1 BLER B UL
Mi
Mi
TX i ic
Mi
Mi
Mi
f TP Scaling
- TPOffset
= Cap E UL -----------------------100
Mi

AT332_TRR_E0
Peak MAC Allocated Bandwidth Throughput:
i
ABTP P UL
i
CTP P UL
N SC UL
----------------M
PZ
N SC
i
UL
i
i
i
Effective MAC Allocated Bandwidth Throughput: ABTP E UL = ABTP P UL 1 BLER B UL

Mi
Application Allocated Bandwidth Throughput: ABTPA UL
Peak MAC Throughput per User:
i
PUTP P UL
Cap M i
M
P UL
- ABTPP i UL
= Min ----------------------TX i ic
N Users UL
Mi
Effective MAC Throughput per User: PUTP E UL
Mi
Mi
Mi
f TP Scaling
- TPOffset
= ABTP E UL -----------------------100
Mi
Application Throughput per User: PUTP A UL
Cap Mi
M
E UL
- ABTP E i UL
= Min ----------------------TX i ic
N Users UL
Mi
Mi
f TP Scaling
- TP Offset
= PUTP E UL -----------------------100
Mi
Output
Mi
CTP P DL : Downlink peak MAC channel throughput at the pixel, subscriber, or mobile Mi.
CTP E DL : Downlink effective MAC channel throughput at the pixel, subscriber, or mobile Mi.
Cap P DL : Downlink peak MAC cell capacity at the pixel, subscriber, or mobile Mi.
Cap E DL : Downlink effective MAC cell capacity at the pixel, subscriber, or mobile Mi.
PUTP P DL : Downlink peak MAC throughput per user at the pixel, subscriber, or mobile Mi.
PUTP E DL : Downlink effective MAC throughput per user at the pixel, subscriber, or mobile Mi.
CTP P UL : Uplink peak MAC channel throughput at the pixel, subscriber, or mobile Mi.
CTP E UL : Uplink effective MAC channel throughput at the pixel, subscriber, or mobile Mi.
Cap P UL : Uplink peak MAC cell capacity at the pixel, subscriber, or mobile Mi.
Cap E UL : Uplink effective MAC cell capacity at the pixel, subscriber, or mobile Mi.
ABTP P UL : Uplink peak MAC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
ABTP E UL : Uplink effective MAC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
ABTP A UL : Uplink application allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
PUTP P UL : Uplink peak MAC throughput per user at the pixel, subscriber, or mobile Mi.
PUTP E UL : Uplink effective MAC throughput per user at the pixel, subscriber, or mobile Mi.
Mi
M
Mi
Mi
Mi
Mi
M
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
Mi
751


Atoll WiMAX BWA module includes a number of scheduling methods which can be used for scheduling and radio resource
allocation during Monte Carlo simulations. These resource allocation algorithms are explained in "Scheduling and Radio
Resource Allocation" on page 752 and the calculation of user throughputs is explained in "User Throughput Calculation" on
page 761.

Input
TX ic
i
QoS
TX i ic
TX i ic
Mi
Mi
: QoS class of the service (UGS, ErtPS, rtPS, nrtPS, or Best Effort) accessed by a mobile Mi.
: Priority of the service accessed by a mobile Mi.
i
i
BLER BDL : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the WiMAX equipment
Mi
Mi
M
Mi
TX ic
i
i
BLER BUL : Uplink block error rate read from the BLER vs. CINR UL graph available in the WiMAX equipment assigned

M
CTP P DL : Downlink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
Mi
Mi
page 744.
Mi
CTP E DL : Downlink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 744.
Mi
CTP P UL : Uplink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 744.
Mi
CTP E UL : Uplink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 744.
ABTP P UL : Uplink peak MAC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput
QoS
f Bias : Bias factor defined for the Biased (QoS Class) scheduling method.
Calculations
Mobile Selection:
752

AT332_TRR_E0
TX ic
i
TX ic
i
TX ic
TX ic
TX ic
i
i
i
TX i ic
Sel
Calculation of Actual Minimum and Maximum Throughput Demands:

Depending on the selected target throughput of the scheduler assigned to the cell TXi(ic), the actual minimum and maximum
throughput demands can be considered as the peak MAC, effective MAC, or application throughput. Therefore:
Target Throughput = Peak MAC Throughput

M
Sel
i
Sel
i
Downlink: TPD Min DL , TPD Max DL

Sel
Sel
Mi
Mi
Mi
Uplink: TPD Min UL , Min TPD Max UL ABTP P UL
Target Throughput = Effective MAC Throughput
Downlink:
Sel
i
TPD Min DL
M
Sel
Mi
Sel
Mi
Sel
M
TPD Min DL
TPD Max DL
i
= --------------------------------------------- , TPD Max DL = --------------------------------------------Sel
Sel
Mi
Mi
1 BLER B DL
1 BLER BDL
Sel
Uplink:
Sel
Mi
TPD Min UL
Sel
Mi
TPD Min UL
= ---------------------------------------------,
Sel
Mi
1 BLER BUL
Sel
Mi
TPD Max UL
Mi
Mi
Min TPD Max UL ABTP P UL
= ------------------------------------------------------------------------Sel
Mi
1 BLER B UL
Target Throughput = Application Throughput

M
Sel
Mi
Sel
i
Sel
Sel
i
Mi
TPD Min DL + TP Offset
TPD Max DL + TP Offset
- , TPD Max DL = ----------------------------------------------------------------------------Downlink: TPD Min DL = ----------------------------------------------------------------------------Sel
Sel
Mi Mi
Mi Mi
Uplink:
Sel
i
TPD Min UL
M
Sel
Mi
Mi
TPD Min UL + TP Offset

= -----------------------------------------------------------------------------,
Sel
Mi Mi
1 BLER BUL f TP Scaling
Sel
Sel
Mi
TPD Max UL
Mi
Mi
Mi
Min TPD Max UL ABTP P UL + TP Offset
= -------------------------------------------------------------------------------------------------Sel
Mi Mi
1 BLER BUL f TP Scaling
The Min() function selects the lower of the two values. This calculation is performed in order to limit the maximum uplink
throughput demand to the maximum throughput that a user can get in uplink using the allocated bandwidth (number of used
subchannels) calculated for it in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 740.
Sel
1. For the QoS classes UGS, ErtPS, rtPS, and nrtPS, Atoll sorts the M i
p
Sel
Mi
TX i ic
N Users in order of decreasing service priority,
753

Sel
Mi
QoS
Sel
i
UGS
ErtPS
rtPS
Sel
Mi
Sel
Mi
nrtPS
N1
TX i ic
=n
Sel
Mi
=n
> 0 ...
=0
=n
Sel
Mi
Sel
Mi
> 0 ...
=0
Sel
i
Sel
i
... n > p
> 0 ...
=0
Sel
Mi
Sel
Mi
... n > p
=n
Sel
Mi
Sel
Mi
... n > p
Sel
i
Sel
i
... n > p
> 0 ...
=0
TX i ic
Where N N Users , if there are some Best Effort users, or N = N Users if there are no Best Effort users selected.
Sel
Sel
= 1 up to M i
= N , Atoll allocates the downlink and uplink resources required to satisfy each
users minimum throughput demands in downlink and uplink as follows:

Sel
Mi
R Min DL
Sel
Mi
Sel
Sel
Mi
Mi
TPD Min DL
TPD Min UL
= -------------------------- and R Min UL = -------------------------Sel
Mi
Sel
Mi
CTP P DL
CTP P UL
When/If in downlink
Sel
Mi
TX i ic
Sel
i
When/If in uplink
Sel
Mi
TX i ic
Sel
Mi

one direction, it is rejected, and the resources, that were allocated to it in the one direction in which it was able to get
a throughput, are allocated to other mobiles.
5. Mobiles which are active UL and whose minimum throughput demand in UL is higher than the uplink allocated
Sel
Mi
Sel
Mi
bandwidth throughput ( TPD Min UL ABTP P UL ) are rejected due to Resource Saturation.
754

AT332_TRR_E0
6. If
Sel
i
TX ic
i
Sel
i
Sel
i
TX ic
i
Sel
i
throughput demands.
If at this stage, a sites downlink or uplink effective MAC aggregate throughput exceeds its maximum downlink or uplink
backhaul throughput, respectively, mobiles are rejected one by one due to Backhaul Saturation, starting from the mobile with
the lowest priority service, among all the cells of the site in order to reach a downlink or uplink effective MAC aggregate site
throughput the sites maximum downlink or uplink backhaul throughput.
TX i ic
TX i ic
Sel
Mi
R Min DL
Sel
Mi
TX i ic
TX i ic
Sel
Mi
R Min UL
Sel
Mi
For each mobile, the throughput demands remaining once the minimum throughput demands have been satisfied are the
difference between the maximum and the minimum throughput demands:
Downlink:
Sel
i
TPD Rem DL
M
Sel
i
TPD Max DL
M
Sel
Mi
Sel
Mi
Sel
i
TPD Min DL
M
Sel
Mi
Uplink: TPD Rem UL = TPD Max UL TPD Min UL

For the remaining throughput demands of the mobiles belonging to the QoS classes ErtPS, rtPS, nrtPS, and Best Effort, the
following resource allocation methods are available:
Proportional Fair:
The goal of this scheduling method is to distribute resources among users fairly in such a way that, on the average,
each user gets the highest possible throughput that it can get under the radio conditions at its location.
Sel
Let the total number of users belonging to the QoS classes ErtPS, rtPS, nrtPS, and Best Effort, be N M i
TX i ic
TX i ic
a. Each users channel throughput is increased by the multi-user diversity gain G MUG DL or G MUG UL read from the
Sel
scheduler properties for the Mobility M i assigned to mobile M i
and the number of connected users, DL or
UL, in the cell TXi(ic) in the iteration k-1.

Sel
Mi
Sel
Mi
CTP P DL = CTP P DL
Without MUG
Sel
Mi
TX i ic
Sel
Mi
TX i ic
Sel
Mi
G MUG DL and CTP P UL = CTP P UL

Max
TX i ic
TX i ic
Without MUG
Sel
Mi
G MUG DL = 1 if CINR Traffic CINR MUG and G MUG UL = 1 if CINR UL
G MUG UL
Max
CINR MUG .
If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph,
it is interpolated from the gain values available for the numbers of users just less than and just greater than the
actual number of users.
b. Atoll divides the remaining resources in the cell into equal parts for each user:
TX i ic
TX i ic
R Rem DL
R Rem UL
--------------------- and -------------------N
N
c. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
755

Sel
i
RD Rem DL
M
Sel
i
Sel
Sel
i
M
TPD Rem DL
TPD Rem UL
i
= --------------------------- and RD Rem UL = --------------------------M
Sel
i
CTP P DL
Sel
i
CTP P UL
d. The resources allocated to each user by the Proportional Fair scheduling method for satisfying its maximum
Sel
Mi
R Max DL
TX i ic
Sel
Sel
Sel
TX i ic
Mi
Mi
Mi
R Rem DL
R Rem UL
N
N
e. Atoll stops the resource allocation in downlink or uplink,
When/If in downlink
Sel
Mi
TX i ic
Sel
i
When/If in uplink
Sel
Mi
TX i ic
Sel
Mi

f. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of
remaining users.
g. Atoll recalculates the remaining resources as follows:
TX ic
i
TX ic
i
Sel
i
R Min DL
Sel
Mi
TX i ic
TX i ic
Sel
i
R Max DL and
Sel
Mi
Sel
Mi
R Min UL
Sel
M
i
Sel
Mi
R Max UL
Sel
i
h. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied
TX i ic
TX i ic
The goal of this scheduling method is to allocate resources to users weighted according to their remaining throughput
demands. Therefore, the user throughputs for users with high throughput demands will be higher than those with low
throughput demands. In other words, this scheduler distributes channel throughput between users proportionally to
their demands.
a. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel
Mi
RD Rem DL
Sel
Mi
Sel
Sel
Mi
Mi
TPD Rem DL
TPD Rem UL
= --------------------------- and RD Rem UL = --------------------------Sel
Mi
Sel
Mi
CTP P DL
CTP P UL
b. Atoll calculates the amount effective remaining resources for the cell of each user to distribute among the users
as follows:
Sel
Sel
TX i ic
Mi
TX i ic
Mi
TXi ic
TX i ic
R Eff Rem DL = Min R Rem DL

RD Rem DL and R Eff Rem UL = Min R Rem UL
RD Rem UL
Sel
Sel
Mi
Mi
756

AT332_TRR_E0
c. The resources allocated to each user by the Proportional Demand scheduling method for satisfying its maximum
Sel
i
R Max DL
M
TX ic
i
R Eff Rem DL
Sel
i
Sel
Sel
i
M
TX ic
RD Rem DL
RD Rem UL
i
i
- and R Max
--------------------------------- UL = R Eff Rem UL ---------------------------------Sel
Sel
M
RDRem DL
RDRem UL
Sel
Mi
Sel
Mi
Biased (QoS Class):

The goal of this scheduling method is to distribute resources among users of each QoS class fairly in such a way that,
on the average, each user gets the highest possible throughput that it can get under the radio conditions at its
location. The resources available for allocation to users of each QoS class depend on a bias factor. The QoS Class Bias
Factor controls the amount of resources available for each QoS class.
Calculation of the Remaining Resources per QoS Class:
QoS
The bias factor f Bias represents the bias in terms of resources allocated to 1 user of a QoS class with rank r to the
resources allocated to 1 user of a QoS class with rank r1:
QoS
Sel
i
Sel
i
Sel
i
R Max ErtPS
f Bias
R Max rtPS
R Max nrtPS
- = ------------------------- = 1 + --------= -------------------------= -------------------------Sel
Sel
Sel
100
M
M
M
i
i
i
R Max rtPS
R Max nrtPS
R Max BE
The ranks of QoS classes are:
QoS Class
QoS Class Rank r QoS
ErtPS
rtPS
nrtPS
Best Effort
The resources available for the users of each QoS class from among the remaining resources is calculated as follows:
r
TX i ic
R QoS DL
1 QoS
1 QoS
N QoS ---
N QoS ---
TX i ic
TX i ic
TX i ic

= R Rem DL ------------------------------------------------------- and R QoS UL = R Rem UL ------------------------------------------------------r
r
1 QoS
1 QoS
N QoS ---
N QoS ---

All QoS
All QoS
Resource Allocation:
Once the remaining resources available for the users of each QoS class have been determined, the allocation of
resources within each QoS class is performed as for the proportional fair scheduler.
Sel
Let the number of users belonging to a QoS class N QoS M i .

a. Atoll divides the remaining resources of the QoS class into equal parts for each user:
TX i ic
TX i ic
R QoS DL
R QoS UL
-------------------- and ------------------N QoS
N QoS
demands:
Sel
Mi
RD Rem DL
Sel
Mi
Sel
Sel
Mi
Mi
TPD Rem DL
TPD Rem UL
= --------------------------- and RD Rem UL = --------------------------Sel
Mi
CTP P DL
Sel
Mi
CTP P UL
c. The resources allocated to each user by the Biased scheduling method for satisfying its maximum throughput
demands are:
757

Sel
i
R Max DL
M

TX ic
i
Sel
Sel
Sel
TX ic
i
M
R QoS DL
R QoS UL
Mi
Mi
i
= Min RD Rem DL -------------------- and R Max UL = Min RD Rem UL --------------------
N
N QoS
QoS
the remaining resources of the QoS class, whichever is smaller.
d. Atoll stops the resource allocation for a QoS class in downlink or uplink,
When/If in downlink
Sel
i
TX ic
i
R Max DL = R QoS DL , i.e., the resources available in downlink for the QoS class have
Sel
Mi
been used up for satisfying the maximum throughput demands of the mobiles.
Sel
Mi
When/If in uplink
TX i ic
R Max UL = R QoS UL , i.e., the resources available in uplink for the QoS class have been
Sel
Mi
used up for satisfying the maximum throughput demands of the mobiles.

remaining users.
TX i ic
TX i ic
Sel
Mi
R QoS DL = TL DL Max
R Min DL
Sel
Mi
TX ic
i
TX ic
i
Sel
Mi
M
R QoS UL = TL UL Max
Sel
Mi
R Max DL and
Sel
i
R Min UL
Sel
Mi
Sel
i
R Max UL
Sel
Mi
g. Atoll repeats the all the above steps for the users of the QoS class whose maximum throughput demands have not
TX i ic
TX i ic
been satisfied until either R QoS DL = 0 and R QoS UL = 0 , or all the maximum throughput demands are
satisfied.
Max Aggregate Throughput:

The goal of this scheduling method is to achieve the maximum aggregate throughput for the cells. This is done by
allocating as much resources as needed to mobiles with high C/(I+N) conditions. As mobiles with high C/(I+N) can get
higher bearers, and therefore require less amount of resources, more mobiles can therefore be allocated resources
in the same frame, and the end-throughput for each cell will be the highest compared to other types of schedulers.
Sel
a. Atoll sorts the M i
TX i ic
N Users in order of decreasing downlink or uplink traffic C/(I+N), depending on whether the
allocation is being performed for the downlink or for the uplink.

b. Starting with the mobile with the highest rank, Atoll allocates the downlink and uplink resources required to
satisfy each users remaining throughput demands in downlink and uplink as follows:
Sel
Mi
R Max DL
Sel
Mi
Sel
Mi
Sel
Mi
TPD Rem DL
TPD Rem UL
= --------------------------- and R Max UL = --------------------------Sel
Mi
Sel
Mi
CTP P DL
CTP P UL
c. Atoll stops the resource allocation in downlink or uplink,
When/If in downlink
Sel
Mi
TX i ic
Sel
Mi
When/If in uplink
Sel
i
TX ic
i
Sel
Mi
Round Robin:
The goal of this scheduling method is to allocate equal resources to users fairly.
758

AT332_TRR_E0
Sel
Let the total number of users belonging to the QoS classes ErtPS, rtPS, nrtPS, and Best Effort, be N M i
a. Atoll divides the remaining resources in the cell into equal parts for each user:
TX ic
i
TX ic
i
R Rem DL
R Rem UL
--------------------- and -------------------N
N
demands:
Sel
Mi
RD Rem DL
Sel
Mi
Sel
Mi
Sel
Mi
TPD Rem DL
TPD Rem UL
= --------------------------- and RD Rem UL = --------------------------Sel
Mi
Sel
Mi
CTP P DL
CTP P UL
c. The resources allocated to each user by the Round Robin scheduling method for satisfying its maximum
Sel
TX i ic
Sel
Sel
Sel
TX i ic
Mi
Mi
R Rem DL
R Rem UL
Mi
Mi
- and R Max
R Max DL = Min RD Rem DL ------------------- UL = Min RD Rem UL ---------------------
N
N
d. Atoll stops the resource allocation in downlink or uplink,
When/If in downlink
Sel
Mi
TX i ic
Sel
Mi
When/If in uplink
Sel
Mi
TX i ic
Sel
Mi

remaining users.
TX i ic
TX i ic
TX i ic
TX i ic
Sel
Mi
R Min DL
Sel
i
Sel
Mi
R Min UL
Sel
Mi
Sel
Mi
R Max DL and
Sel
i
Sel
Mi
R Max UL
Sel
Mi
g. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied
TX i ic
TX i ic
Spatial Multiplexing with Uplink Multi-User MIMO:
Using MU-MIMO cells are able to schedule two users on two different antennas using the same frequency-time resources. A
mobile connected to antenna 1 makes resources available on antenna 2 for another mobile. These resources made available
on antenna 2 can be assigned to another mobile without increasing the overall cell load.
When the second mobile is assigned to antenna 2, it consumes the resources made available by the first mobile and, if it
requires more resources than the first mobile made available, the second mobile also makes some resources available on
antenna 1.
Each new mobile is either connected to antenna 1 or antenna 2. The part of the mobiles resources which are not coupled
with resources allocated to another mobile on the other antenna is called the real resource consumption. The part of the
mobiles resources which are coupled with the resources allocated to another mobile on the other antenna is called the virtual
resource consumption.
759

MU-MIMO can be used if the permutation zone assigned to the pixel, subscriber, or mobile Mi supports MU-MIMO,
TX ic
i
TX ic
i
TX ic
i
CNR Preamble T MU MIMO , and N Ant RX 2 .

Let i be the index of connected MU-MIMO mobiles: i = 1 to N
MU MIMO
Each mobile M i
MU MIMO
M
i=0
RR UL
MU MIMO
Mi
has a corresponding traffic load TL UL
MU MIMO
M
i=0
V UL
= 100 % and available virtual resources
. The scheduling starts with available real resources

= 0 % . i = 0 means no MU-MIMO mobile has yet
been scheduled.
The virtual resource consumption of a mobile
MU MIMO
Mi
MU MIMO
Mi
is given by: VC UL
MU MIMO
Mi
MU MIMO
The real resource consumption of a mobile M i
is given by: RC UL
MU MIMO
The virtual resources made available by the mobile M i

MU MIMO
Mi
V UL
MU MIMO
Mi 1
= V UL
MU MIMO
Mi
VC UL
MU MIMO
Mi
RC UL
MU MIMO
Mi
= Min TL UL
MU MIMO
Mi
= TL UL
MU MIMO
Mi 1
V UL
MU MIMO
Mi
VC UL
are given by:
MU MIMO
Mi
+ RC UL
TX i ic
= TL UL Max .
The following table gives an example:

Mobile
MU MIMO
Mi
TL UL
(%)
MU MIMO
Mi
VC UL
(%)
MU MIMO
Mi
RC UL
MU MIMO
Mi
V UL
(%)
M1
10
10
10
M2
M3
20
15
15
M4
40
15
25
25
(%)

Sel
Sel
M
Mi
i
R Max DL CTP E DL
Sel
M i Site
= Max 1 ------------------------------------------------------------------------------------------------------- and
Sel
Sel
M
Mi
Site
i
TP
CTP
BH DL
Min DL
E DL
Sel
M i Site
Site
BHOF DL
Sel
Sel
Mi
Mi
R Max UL CTP E UL
Sel
M i Site
= Max 1 -------------------------------------------------------------------------------------------------------
Sel
Sel
Mi
Mi
Site
TP
CTP
BH UL
Min UL
E UL
Sel
M i Site
Site
BHOF UL

Sel
760
(which can also be

AT332_TRR_E0
Sel
Downlink:
Sel
i
TL DL
M
Sel
i
R DL
M
Sel
i
Sel
i
M
Mi
i
Max DL CTP P DL

BHOF DL
= -----------------------------------------------------------------------------------------------------------------Sel
Sel
i
CTP P DL
Sel
Sel
Mi
Uplink: TL UL
Sel
Mi
= R UL
Sel
i
Sel
i
Mi
Mi
Max UL CTP P UL

BHOF UL
= -----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P UL
Output
Sel
Mi
TL DL
TL UL
Sel
Mi
Sel
Mi
= R DL
Sel
Sel
Mi
Sel

Sel

Input
Sel
Mi
R DL
Sel
Sel
Mi
Sel
Sel
i
Sel
CTP P DL : Downlink peak MAC channel throughput at the mobile M i
page 744.
Sel
i
Sel
CTP P UL : Uplink peak MAC channel throughput at the mobile M i
page 744.
Sel
TX i ic
Mi
BLER B DL : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the WiMAX equipment
Sel
assigned to the terminal used by the mobile M i
Sel
Mi
Mi
BLER B UL : Uplink block error rate read from the BLER vs. CINR UL graph available in the WiMAX equipment
assigned to the cell TXi(ic).

Sel
Mi
Sel
Sel
Mi
Sel
Calculations
Downlink:
Sel
Mi
Sel
Mi
Sel
Mi
CTP P DL
Peak MAC User Throughput: UTP P DL = R DL
Mi
Mi
Mi
Effective MAC User Throughput: UTP E DL = UTP P DL 1 BLER B DL
Sel
Sel
Sel
761

Application User Throughput:
Sel
i
UTP A DL
M
Sel
i
UTP E DL
M
Sel
i
Sel
M
f TP Scaling
i
------------------------- TP Offset
100
Uplink:
Sel
Mi
Sel
Mi
Sel
Mi
CTP P UL
Peak MAC User Throughput: UTP P UL = R UL
M
M
Mi
i
i
Effective MAC User Throughput: UTP E UL = UTP P UL 1 BLER B UL
Sel
Sel
Mi
Sel
Sel
Mi
Sel
Sel
Mi
Sel
Mi
f TP Scaling
- TP Offset
= UTP E UL -----------------------100
Output
M
Sel
i
Sel
UTP P DL : Downlink peak MAC user throughput at the pixel, subscriber, or mobile M i
UTP E DL : Downlink effective MAC user throughput at the pixel, subscriber, or mobile M i .
UTP P UL : Uplink peak MAC user throughput at the pixel, subscriber, or mobile M i
UTP E UL : Uplink effective MAC user throughput at the pixel, subscriber, or mobile M i .
UTP A UL : Uplink application user throughput at the pixel, subscriber, or mobile M i
Sel
Mi
Sel
Sel
Mi
Sel
Sel
i
Sel
Sel
Mi
Sel
Sel
Mi
Sel


"Automatic Preamble Index Planning Using the AFP" on page 771.
"Automatic Zone PermBase Planning Using the AFP" on page 775.

They are active,

They belong to the folder on which the allocation has been executed. This can be the Transmitters folder or a group
of transmitters (subfolder).

zone.
762

AT332_TRR_E0
We assume a reference cell TXi(ic) and a candidate neighbour cell TXj(jc). When automatic planning starts, Atoll checks the
discarded.

[Neighbours]

decreasing rank.
neighbour.
763

i.
[Neighbours]
3. If the Use Coverage Conditions check box is selected, the coverage areas of TXi(ic) and TXj(jc) must have an overlap
( S TX ic S TX jc ).Otherwise, only the distance criterion is taken into account.
i
The overlapping zone ( S TX ic S TX jc ) is defined as follows:

i
i
The received preamble signal level is greater than or equal to the preamble signal level threshold. The received
TX ic
i
TX ic
i
preamble signal level ( C Preamble ) and the preamble signal level threshold are calculated from CNR Preamble
TX i ic
TX i ic
and T Preamble , respectively, by adding the value of the noise ( n Preamble ) to them.
TX i ic
TX i ic
S TX ic is the surface area covered by TXi(ic) within C Preamble + HO Start and C Preamble + HO End , or
TX ic
i
TX ic
i
CINR Preamble + HO Start and CINR Preamble + HO End . HOStart is the margin with respect to the best preamble
signal level or C/(I+N) at which the handover starts, and HOEnd is the margin with respect to the best
preamble signal level or C/(I+N) at which the handover ends.

S TX jc is the coverage area where the candidate cell TXj(jc) is the best server.
j
TX i ic
If a global value of the preamble C/N threshold ( T Preamble ) is set in the coverage
conditions dialog box, for each cell, Atoll uses the higher of the two values, i.e., global
value and the value defined for that cell.
gives the largest possible preamble C/N coverage areas for the cells.
Atoll
S TX ic S TX jc
i
j
- 100 ),
i
and compares this value with the % Min Covered Area.
764

AT332_TRR_E0

S TX ic S TX jc
i
j
- 100 % Min Coverage Area .
TXj(jc) is considered a neighbour of TXi(ic) if -------------------------------------S TX ic
i
Neighbourhood cause
When
Importance value
Existing neighbour

Existing importance
Exceptional pair
100 %
Co-site cell

selected
Adjacent cell

selected

relationship
transmitter.
d Di = 1 ---------d max
of antennas.
[Neighbours]
765


Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
30%
Min(A)
30%
Max(A)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

Neighbourhood cause
Importance Function

Co-site
Adjacent
No
No
10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
No
Yes
30%+30%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Yes
Yes
60%+40%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Where:
overlapping.
causes.
some coverage overlapping.If the Min and Max value ranges of the importance
function factors overlap, the neighbours may be ranked differently. There can be a
mix of the neighbourhood causes.
In the results, Atoll lists only the cells for which it finds new neighbours. Cells whose channels have the same start frequency,
the same channel width, and the same total number of subcarriers are listed as intra-carrier neighbours. Otherwise,
neighbour cells are listed as inter-carrier neighbours.

They are active,

766

AT332_TRR_E0
zone.
We assume a reference cell A and a candidate neighbour B. When automatic planning starts, Atoll checks following
conditions:

[Neighbours]
CDMA Carriers: This option is available when an WiMAX network is being co-planned with a UMTS, CDMA, or TDSCDMA network. This option enables you to select the CDMA carrier(s) that you want Atoll to consider as potential
neighbours of WiMAX cells. You may choose one or more carriers. Atoll will allocate only the cells using the
selected carriers as neighbours.

neighbour.

1st case: SA is the area where the cell A is the best serving cell, with a 0dB margin.
This means that the preamble signal received from A is greater than the minimum required (calculated from
the preamble C/N threshold), and is the highest one. .
767

2nd case: The margin is other than 0dB. SA is the area where:
The preamble signal level received from A exceeds the minimum required (calculated from the preamble C/N
threshold) and is within a margin from the highest signal level.
to 0dB.
2nd case: The margin is other than 0dB. SB is the area where:
level.
SA SB
SA
SA SB
- 100 % Min Covered Area .
Min Covered Area. B is considered a neighbour of A if ----------------SA

Neighbour cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
100 %

distance conditions
d1 ---------d max
distance.

Neighbour cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
IF

IF
transmitter.
d Di = 1 ---------d max

768

AT332_TRR_E0
Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

cause
IF

No
10%+50%{10%(Di)+90%(O)}
Yes
60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.

Co-channel interference is the main reason for overall network quality degradation in WiMAX. In order to improve network
performance, the WiMAX AFP tries to minimise co- and adjacent channel interference as much as possible while respecting
any constraints input to it. The main constraints are the resources available for allocation, i.e., the number of frequencies with
The following describes the AFPs automatic planning method for frequencies in WiMAX networks, which takes into account
They are active,

zone.

769


You can modify these weights in your WiMAX document. The absolute values of the constraint weights are calculated from
the relative weights (%) defined in the Constraint Weights dialog box as follows:
% Neighbour
% IM
% Dis tan ce

$
TX i ic TX j jc
TX i ic TX j jc
= rO
TX i ic TX j jc
Where r O
TX i ic TX j jc
Neighbour Neighbour
TX i ic TX j jc
TX ic TX j jc
+ i
IM
IM
page 706.
TX i ic TX j jc
TX i ic TX j jc
Neighbour
is
TX i ic TX j jc
IM
TX i ic TX j jc
TX i ic TX j jc
IM
= r CCO
TX i ic TX j jc
TX i ic TX j jc
IM CC
and IM CC
TX i ic TX j jc
IM CC
TX i ic TX j jc
+ r ACO
TX i ic TX j jc
IM AC
TX i ic TX j jc
TX i ic TX j jc
and r ACO
TX i ic TX j jc
Dis tan ce
TX i ic TX j jc
them. Dis tan ce
QRF
TX i ic TX j jc
= 1$
TX i ic TX j jc

cell has relations:
770

AT332_TRR_E0
TX ic
i
$ Total = 1
QRF
TX ic TX jc
i
j
TX jc
j
above, i.e.,
$ Total =
TX i ic
$ Total
TX i ic

Calculates the cost (as described above) of the initial frequency plan,
solution.
10.5.4 Automatic Preamble Index Planning Using the AFP

IEEE 802.16e defines 114 preamble indexes. Each preamble index, from 0 to 113, contains the following information:
Segment number (0, 1, or 2),

DL PermBase (0 to 31) for the obligatory first DL PUSC zone, and
A pseudo-noise sequence transmitted using the subcarriers corresponding to the preamble carrier set.
The downlink subframe can be divided into a 3-segment structure, and includes a preamble which begins the transmission
(the first symbol of the downlink transmission). The preamble subcarriers are divided into 3 carrier sets. There are three
possible groups consisting of a carrier set each which may be used by any segment. These are defined by allocation of different
subcarriers to each one of them. The subcarriers are modulated using a BPSK modulation with a specific Pseudo-Noise (PN)
sequence.
Preamble carrier sets are defined using equation below:
PreambleCarrierSet n = n + k 3
Where PreambleCarrierSetn gives the subcarriers used by the preamble, n is the number of the preamble carrier set indexed
0, 1, or 2, k is a running index from 0 to 567 for FFT 2048, from 0 to 283 for FFT 1024, from 0 to 142 for FFT 512, and from 0
to 35 for FFT 128.
In a WiMAX network, each base station transmits a different PN sequence, out of the 114 available, on the preamble carrier
set. A mobile trying to connect to the network scans all the preamble subcarriers, listens to all the preambles (i.e., PN
sequences) from all the base stations it can receive, and compares the PN sequences it is receiving with the 114 stored in its
memory in order to detect the preamble index from the PN sequence.
It selects the base station as its server whose preamble it receives with either the highest signal level or the highest C/(I+N).
Once the best server is known, its PN sequence is used to identify its transmission. The PN sequence of the best server gives
the preamble index, which in turn gives the segment number, and the IDCell (DL PermBase of the first DL PUSC zone, referred
to as Cell PermBase in Atoll). Therefore, the mobile knows which subcarriers to listen to for the FCH, DCD, UCD, DL-MAP, and
UL-MAP.
As can be understood from the above description, if all the cells in the network transmit the same preamble index, the
network will have 100% interference on downlink preambles, and it will be impossible for a mobile to identify different cells.
Cell search and selection will be impossible. Therefore, it is important to intelligently plan preamble indexes to cells so as to
reduce preamble interference, and allow easy recognition of cells by mobiles.
The following describes the AFPs automatic planning method for preamble indexes in a WiMAX network, which takes into
account interference matrices, neighbour relations (first-order neighbours, first-order neighbours of a common WiMAX cell,
and optionally second-order neighbours), distance between transmitters, and the frequency plan of the network.
They are active,

Their preamble index status or segment is not set to locked,
771

zone.

The AFP is based on a cost-based function which takes into account the following constraints, in the order of priority:
1. Same preamble index,
Assigned weight PI = 0,75
2. Same segment number,
Assigned weight Seg = 0,02
3. Same cell permbase,
Assigned weight PB = 0,23
The sum of the weights assigned to the above constraints is 1.
% PI
PI = -----------------------------------------------------% PI + % Seg + % PB
% Seg
Seg = -----------------------------------------------------% PI + % Seg + % PB
% PB
PB = -----------------------------------------------------% PI + % Seg + % PB
The above constraints are studied between each TBA cell and its related cells. Atoll calculates the cost between each
individual TBA and related cell, and then the overall cost for the TBA cell.

You can choose to not take into account the preamble index collision between neighbours of a common cell by adding
772

AT332_TRR_E0
% Neighbour
Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% Inter Neighbour
Inter Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% 2nd Neighbour
2nd Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% IM
IM = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce
% Dis tan ce
Dis tan ce = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter Neighbour + % 2nd Neighbour + % IM + % Dis tan ce

TX i ic TX j jc
VL 1
TX i ic TX j jc
VL 2
PI
PB
= PI p Coll + PB p Penalty
Seg
= Seg p Coll
If TX i ic and TX i jc are co-transmitter cells, and the option Allocate Same Segment to Co-transmitter Cells has been
TX i ic
selected, and N Seg
TX i jc
N Seg
TX i ic TX j jc
, then VL 1
TX i ic TX j jc
+ VL 2
= 1.
Where PI , Seg , and PB are the weights assigned to the preamble index, segment number, and cell permbase constraints.
PI
PI
p Coll is the preamble index collision probability given by p Coll = 1
if PI
if PI
TX i ic
TX i ic
= PI
PI
TX j jc
TX j jc
Seg
p Coll is the segment number collision probability. If TX i ic and TX j jc are co-transmitter cells, and the option Allocate
0
Seg
Seg
Same Segment to Co-transmitter Cells has been selected, p Coll is given by p Coll =
1
Seg
p Coll
PB
1
=
p Penalty
TX i ic
if NSeg
TX i ic
if N Seg
TX i ic
if N Seg
TX i ic
if N Seg
TX j jc
= N Seg
TX j jc
. Otherwise,
N Seg
TX j jc
= N Seg
TX j jc
N Seg
TX ic
TX jc
TX i ic
TX j jc
1 if PB i PB j
AND Site
= Site
PB
is the cell permbase penalty given by p Penalty =
TX i ic
TX j jc
TX i ic
TX j jc if the
PB
AND Site
Site
0,001 if PB
0 Otherwise
PB
cell permbase planning strategy is set to "Same per site", and by p Penalty = 0 if the cell permbase planning strategy is set to
"Free". The cell permbase penalty models the cell permbase constraint.
TX i ic TX j jc
Neighbours
TX i ic TX j jc
773

TX ic TX jc
i
j

TX ic TX jc
i
j
Where Neighbour
planning. If two cells are neighbours of a common cell and have the same preamble index assigned, the importance of the
preamble index collision is the average of their neighbour importance values with the common neighbour cell. If more than
one pair of neighbours of the TBA cell has the same preamble index assigned, then the importance is the highest value among
all the averages:
TX i ic TX j1 j1c
TX i ic TX j2 j2c
Neighbour
+ Neighbour
---------------------------------------------------------------------------------
2
All Neighbour Pairs
Inter Neighbour =
Max
with PI Collisions
Where TX j1 j1c and TX j2 j2c are two neighbours of the TBA cell TX i ic that have the same preamble index assigned.
planning. If the TBA cell has the same preamble index assigned as one of its second-order neighbours, the importance of the
preamble index collision is the multiple of the importance values of the first order neighbour relations between the TBA cell
2nd Neighbour =
TX ic TX jc
j
i
Neighbour
All Neighbour Pairs
Max
TX j jc TX k kc
Neighbour
with PI Collisions

TX i ic TX j jc
Interference
TX i ic TX j jc
IM
TX i ic TX j jc
IM
TX i ic TX j jc
= IM IM
TX i ic TX j jc
= r CCO
TX ic TX jc
i
j
TX ic TX jc
i
j
TX i ic TX j jc
f Overlap
and IM
IM CC
TX i ic TX j jc
TX i ic TX j jc
IM CC
TX ic TX jc
i
j
= IM CC
TX ic TX jc
i
j
and IM CC
TX i ic TX j jc
+ r ACO
TX i ic TX j jc
IM AC
otherwise.
TX i ic TX j jc
TX i ic TX j jc
, r CCO
TX i ic TX j jc
, and r ACO
are the total,
page 706.
TX i ic TX j jc
Dis tan ce
them.
TX i ic TX j jc
Dis tan ce
TX i ic TX j jc
f Overlap
is
calculated
TX i ic TX j jc
= rO
as
explained
in
"Distance
Importance
TX i ic TX j jc
Calculation"
on
page 778.
= 1 otherwise.
QRF
TX i ic TX j jc
TX i ic TX j jc
TX i ic TX j jc
TX ic TX j jc
TX i ic TX j jc
TX i ic TX j jc
TX i ic TX j jc
i
= 1 VL 1
+ VL 2
+ VL 1
Neighbours f Overlap
Interference

The total cost of the current preamble index plan for any TBA cell is given as follows, considering all the cells with which the
TX i ic
$ Total = 1
QRF
TX jc
j
774
TX i ic TX j jc

AT332_TRR_E0
And, the total cost of the current preamble index plan for the entire network is simply the sum of the total TBA cell costs
$ Total =
TX ic
i
$ Total
TX i ic

Calculates the cost (as described above) of the initial preamble index plan,
Tries different preamble index plans in order to reduce the cost,
Stops when it is unable to improve the cost of the network, and proposes the last known best preamble index plan as
the solution.
10.5.5 Automatic Zone PermBase Planning Using the AFP

PermBases are numbers which are used as seeds in the permutation of subcarriers (mapping between physical and logical
subcarrier numbers) and their allocation to subchannels. Subchannels in a channel contain different physical subcarriers when
different permbases are used as seeds.
Downlink PUSC permutation zones use 2 permbases:
1. The first DL PUSC permutation zone uses the cell permbase (mapped to the preamble index of the cell). It is called
IDCell in the IEEE specifications. It is a number from 0 to 31.
2. The second DL PUSC permutation zone uses the zone permbase, also a number from 0 to 31.
Other downlink permutation zones only use zone permbases.
Uplink permutation zones also use only zone permbases. However, the uplink zone permbase is a number from 0 to 69.
The following describes the AFPs automatic planning method for zone permbases in a WiMAX network, which takes into
account interference matrices, neighbour relations (first-order neighbours, first-order neighbours of a common WiMAX cell,
and optionally second-order neighbours), distance between transmitters, and the frequency plan of the network.
They are active,

Their zone permbase status is not set to locked,
In the following description, ZPB is used for the downlink zone permbases ( ZPBDL )
and uplink zone permbases ( ZPBUL ) without distinction.
If no focus zone exists in the ATL document, Atoll takes into account all the cells
whose path loss matrix calculation radius intersects the rectangle containing the
computation zone.

The AFP is based on a cost-based function which takes into account the following constraint:
Same zone permbase,

Assigned weight ZPB = 1
The above constraint is studied between each TBA cell and its related cells. Atoll calculates the cost between each individual
TBA and related cell, and then the overall cost for the TBA cell.

775


You can choose to not take into account the zone permbase collision between neighbours of a common cell by adding

Atoll calculates the constraint violation level between the TBA cell TXi(ic) and its related cell TXj(jc) as follows:
VL
TX ic TX jc
i
j
ZPB
= ZPB p Coll
Where ZPB is the weight assigned to the zone permbase constraint.
ZPB
p Coll
is the zone permbase collision probability given by
ZPB
p Coll
= 1
if ZPB
if ZPB
TX ic
i
TX ic
i
= ZPB
ZPB
TX jc
j
TX jc
j
TX ic TX jc
i
j
Neighbours
TX ic TX jc
i
j
TX ic TX jc
i
j
Where Neighbour

TX ic TX jc
i
j
planning. If two cells are neighbours of a common cell and have the same zone permbase assigned, the importance of the
zone permbase collision is the average of their neighbour importance values with the common neighbour cell. If more than
one pair of neighbours of the TBA cell has the same zone permbase assigned, then the importance is the highest value among
all the averages:
TX i ic TX j1 j1c
Inter Neighbour
TX i ic TX j2 j2c
Neighbour
+ Neighbour
=
Max
---------------------------------------------------------------------------------
2
All Neighbour Pairs

with ZPB Collisions
Where TX j1 j1c and TX j2 j2c are two neighbours of the TBA cell TX i ic that have the same zone permbase assigned.
planning. If the TBA cell has the same zone permbase assigned as one of its second-order neighbours, the importance of the
zone permbase collision is the multiple of the importance values of the first order neighbour relations between the TBA cell
2nd Neighbour =
with ZPB Collisions
776
TX ic TX jc
j
i
Neighbour
All Neighbour Pairs
Max
TX j jc TX k kc
Neighbour

AT332_TRR_E0

TX ic TX jc
i
j
Interference
TX i ic TX j jc
IM
TX i ic TX j jc
IM
TX ic TX jc
i
j
= IM IM
TX ic TX jc
i
j
TX ic TX jc
i
j
f Overlap
TX i ic TX j jc
= r CCO
TX ic TX jc
i
j
and IM
TX i ic TX j jc
TX i ic TX j jc
IM CC
TX ic TX jc
i
j
= IM CC
TX i ic TX j jc
IM CC
and IM CC
TX i ic TX j jc
+ r ACO
TX i ic TX j jc
IM AC
otherwise.
TX i ic TX j jc
TX i ic TX j jc
, r CCO
TX i ic TX j jc
, and r ACO
are the total,
page 706.
TX i ic TX j jc
Dis tan ce
TX i ic TX j jc
Dis tan ce
them.
TX i ic TX j jc
f Overlap
is
TX i ic TX j jc
= rO
calculated
as
explained
in
"Distance
Importance
TX i ic TX j jc
Calculation"
on
page 778.
= 1 otherwise.
From the constraint violation level and the total importance of the relation between the TBA and its related cell, Atoll
QRF
TX i ic TX j jc
= 1 VL
TX i ic TX j jc
TX i ic TX j jc
Interference
TX i ic TX j jc
TX i ic TX j jc
+ Neighbours f Overlap

The total cost of the current zone permbase plan for any TBA cell is given as follows, considering all the cells with which the
TX i ic
$ Total = 1
QRF
TX i ic TX j jc
TX j jc
And, the total cost of the current zone permbase plan for the entire network is simply the sum of the total TBA cell costs
$ Total =
TX i ic
$ Total
TX i ic

Calculates the cost (as described above) of the initial zone permbase plan,
Tries different zone permbase plans in order to reduce the cost,
Stops when it is unable to improve the cost of the network, and proposes the last known best zone permbase plan as
the solution.
10.5.6 Appendices
777

S TX ic
i
TX ic TX jc
i
j
IM CC
TX jc
TX ic
j
i
n
C Preamble + M Quality
Preamble
----------------------------------------------------------------------------------------
TX ic
TX ic
i
10
10
T i
C
10 Log 10
+ 10
Preamble
Preamble
= -------------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX ic
i

S TX ic
i
TX i ic TX j jc
IM AC
TX j jc
TX i ic
TX i ic
n Preamble
C Preamble + M Quality + f ACS FB
-----------------------------------------------------------------------------------------------------------------------
TX i ic
TX ic
10
10
T i
C
10 Log 10
+ 10
Preamble
Preamble
= -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX ic
i
TX i ic
TX i ic
Here S TX ic is the best server coverage area of the cell TXi(ic), that comprises all the pixels where CNR Preamble T Preamble
i
as calculated in "Service Area Calculation" on page 717. S TX ic

i
TX i ic
Condition
is the best server coverage area of the cell TXi(ic)
TX j jc
where the given condition is true. C Preamble and C Preamble are the received preamble signal levels from the cells TXi(ic) and
TX ic
i
TXj(jc) respectively, n Preamble the preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on
TX i ic
page 712, M Quality is the quality margin used for the interference matrices calculation, and f ACS FB is the adjacent channel
suppression factor defined for the frequency band of the cell TXi(ic).

TX i ic TX j jc
TX i ic TX j jc
Dis tan ce
D Reuse
2
= Log --------------------------------
TX i ic TX j jc
D

--------------------------------------------------------2
Log D Reuse
if D
TX i ic TX j jc
Otherwise
as follows:
D
D
TX i ic TX j jc
TX i ic TX j jc
them. d
= d
TX i ic TX j jc
TX i ic TX j jc
1 + x cos cos 2
TX i ic TX j jc
TX i ic TX j jc

778
AT332_TRR_E0


779

780
Chapter 11
Wi-Fi Networks

782

Chapter 11: Wi-Fi Networks
AT332_TRR_E0
11 Wi-Fi Networks
This chapter describes all the calculations performed in Atoll Wi-Fi documents. The first part of this chapter lists all the input
parameters in the Wi-Fi documents, their significance, location in the Atoll GUI, and their usage. It also contains the lists of
the formulas used for the calculations.
signal levels, noise, and interference for downlink and uplink, and the radio resource management algorithms used in Monte
Carlo simulations.
cells.
All the calculation algorithms in this section are described for two types of receivers.

11.1 Definitions
Name
Value
Unit
Description
1.38 x 10-23
J/K
Boltzmanns constant
290
Ambient temperature
n0
dBm/Hz

noise
r CP
None
Cyclic Prefix Ratio (guard interval)

Choice List: 1/4 (long), 1/8 (short)
M PC
Global parameter
dB
CNR Min
Global parametera
dB

N SCa Total
None
Total number of subcarriers per

channel (FFT size)
N SCa Used
None
Number of used subcarriers per

channel
N SCa Data
Frame configuration zone parameter
None
Number of subcarriers per channel

used for data transfer
N SCa DC
None
Number of DC subcarriers per

channel
N SCa Pilot
Calculation result ( N SCa Pilot = N SCa Used N SCa Data )
None
Number of pilot subcarriers per

channel
783

Name
Value
Unit
Description
N SCa Guard
Calculation result
( N SCa Guard = N SCa Total N SCa Used N SCa DC )
None
Number of guard subcarriers per

channel
W Channel
MHz
Channel bandwidth
First
None

frequency band
N Channel
Last
None

frequency band
F Start FB DL
MHz
DL Start frequency of the frequency

band
F Start FB UL
MHz
UL Start frequency of the frequency

band
f ACS FB
dB
Adjacent Channel Suppression Factor
ICS FB
MHz
CN FB
None
Channel number step
Inter Tech
Network parameter
dB
reduction factor
Bearer parameter
None
Bearer index
Mod B
Bearer parameter
None
CR B
Bearer parameter
None
Bearer parameter
bits/
symbol
Bearer Efficiency
TB
Bearer parameter
dB
TP BH DL
Site
Site parameter
kbps
Maximum backhaul site downlink

throughput
Site
Site parameter
kbps
Maximum backhaul site uplink

throughput
dB
N Ant TX
None

in transmission
N Ant RX
None

in reception
TX
Antenna parameter
dB
TX
dB
Transmitter loss
N Channel
Cell parameter
None
P DL
Cell parameter
dBm
Power
TL DL
Cell parameter
TL UL
Cell parameter
Uplink traffic load
TL DL Max
Cell parameter
TL UL Max
Cell parameter
NR UL
Cell parameter
dB
Uplink noise rise
N Users Max
Cell parameter
None
N Users DL
Cell parameter
None

cell in downlink
N Channel
f IRF
TP BH UL
nf
G
L
784
TX

AT332_TRR_E0
Name
Value
Unit
Description
N Users UL
Cell parameter
None

cell in uplink
T AMS
Cell parameter
dB
T MU MIMO
Cell parameter
dB
T Min
Cell parameter
dB
Minimum C/N threshold
D Reuse
Cell parameter
Channel reuse distance
G MU MIMO
Cell parameter
None
Uplink MU-MIMO gain
Inter Tech
Cell parameter
dB
Inter Tech
Cell parameter
dB
G SU MIMO
Max
Cell Wi-Fi equipment parameter
None
G Div UL
Cell Wi-Fi equipment parameter
dB
Service parameter
None
Service priority
B DL Highest
Service parameter
None

the downlink
B UL Highest
Service parameter
None

the uplink
B DL Lowest
Service parameter
None

the downlink
B UL Lowest
Service parameter
None

the uplink
f Act
UL
Service parameter
f Act
DL
Service parameter
TPD Min UL
Service parameter
kbps

uplink
TPD Min DL
Service parameter
kbps

downlink
TPD Max UL
Service parameter
kbps

uplink
TPD Max DL
Service parameter
kbps

downlink
UL
Service parameter
kbps

uplink
TP Average
DL
Service parameter
kbps

downlink
TP Offset
Service parameter
kbps
Throughput offset
f TP Scaling
Service parameter
Scaling factor
L Body
Service parameter
dB
Body loss
P Min
Terminal parameter
dBm
P Max
Terminal parameter
dBm
nf
Terminal parameter
dB
Terminal parameter
dB
Terminal parameter
dB
Terminal loss
NRDL
NRUL
TP Average
785

Name
Value
Unit
Description
N Ant TX
Terminal parameter
None

in transmission
N Ant RX
Terminal parameter
None

in reception
G SU MIMO
Max
Terminal Wi-Fi equipment parameter
None
G Div DL
Terminal Wi-Fi equipment parameter
dB

diversity gain
UL
Clutter parameter
dB
G Div
DL
Clutter parameter
dB
f SU MIMO
Clutter parameter
None
SU-MIMO gain factor
L Indoor
Clutter parameter
dB
Indoor loss
L Path
dB
Path loss
G Div
F
ICPDL
Network parameter
None

M Shadowing Model

standard deviation
dB
M Shadowing C I

dB
a.


Name
TX i ic
F Start
Value
TX i ic
TX ic
TX i ic
TX i ic TX j jc
W CCO
TX jc
TX i ic TX j jc
r ACO
786
TX jc
TX ic
MHz

MHz

MHz
None
MHz

None

overlap ratio
TX ic TX jc
i
j
W CCO
----------------------------------TX ic
i
W Channel
r CCO
TX i ic TX j jc
TX ic
j
i
j
i
TX i ic TX j jc
W ACO
TX i ic
F Start + W Channel
F End
Description
TX ic
i
i
F Start FB + W Channel + ICS FB
TX i ic
TXi ic N First TXi ic

Channel
Channel
N
-------------------------------------------------------
TX i ic
CN FB
Unit
TX jc
TX ic
TX jc
TX ic
TX ic
j
i
j
i
i
TX ic TX jc
i
j
W ACO
L
---------------------------------TX i ic
W Channel

AT332_TRR_E0
Name
TX ic TX jc
i
j
W ACO
Value
TX jc
j
Unit
Description
MHz

None

overlap ratio
None
None
Total overlap ratio
Unit
Description
dBm
Received signal level
dBm
EIRP of a cell
Value
Unit
Description
N TXi ic
SCa Used
n 0 + 10 Log ------------------------ TXi ic
N SCa Total
dBm
dBm
Unit
Description
dBm
Interference generated by an
interfering cell
dB

dB

downlink traffic load
TX ic
i
Min F End F End
TX ic
TX jc
TX ic
i
j
i
TX ic TX jc
i
j
W ACO
H
---------------------------------TX i ic
W Channel
TX ic TX jc
i
j
r ACO
H
TX i ic TX j jc
TX ic TX jc
i
j
r ACO
r ACO
TX i ic TX j jc
+ r ACO
H
TX ic
f ACS FB
----------------------------
j
i
j
10
r i
+ r ACO
10
CCO
TX ic TX jc
i
j
rO
TX i ic
TX j jc
TX i ic
f ACS FB
TX ic
TX ic TX jc TX ic TX jc --------------------------- W i
j
i
j
10
Channel
r i
--------------------+ r ACO
10
TX j jc
CCO
W Channel
TX i ic
TX j jc

Name
TX i ic
C DL
EIRP
Value
TX i ic

L
Mi
L Ant L Body
TX i ic
TX i ic
P DL
+G
TX i
TX i

Name
TX i ic
n 0 DL
TX i ic
TX i ic
n 0 DL + nf
n DL
Mi

Name
TX j jc
I DL
TX i ic TX j jc
fO
TX j jc
f TL DL
Value
TX j jc
C DL
TX i ic TX j jc
+ fO
TX j jc
Inter Tech
+ f TL DL + I DL
TX i ic TX j jc
10 Log r O
TX j jc
10 Log TLDL
787

Name
Value
Unit
Description
Inter Tech
I DL
TX
k
P DL Rec
--------------------------------------
F TX ic TX
i
k
TX ICP DL
k
interference
Unit
Description
dB
Downlink C/N for a cell
Unit
Description

Name
Value
TX i ic
C DL
TX i ic
CNR DL
TX i ic
n DL
TX i ic
With MIMO: CNR DL
Mi
DL
+ G Div DL + G Div

Name
Value
TX i ic
TX ic
i
CINR DL
C DL
TXj jc
I DL
----------------- 10 Log
10 10
All TXj jc
TX i ic
n DL
Inter
Tech
+I
+ -------------------10
DL
10
TX i ic
With MIMO: CINR DL
Mi
TX i ic
Downlink C/(I+N) for a cell
DL
+ G Div DL + G Div
TX jc
I + N DL
Inter
Tech
+ NR
DL
dB
TX ic
I j
n DL
DL
------------------
---------------------
10 Log
DL
DL
All TXj jc
dBm
Unit
Description
dBm
dBm
Value
Unit
Description
N TXi ic
SCa Used
n 0 + 10 Log ------------------------ TXi ic
N SCa Total
dBm
dBm

Name
Value
Mi
Mi
C UL
TX
L Ant L Body
P
EIRP UL
With P
TX i
Mi
+G
Mi
Mi
= P Eff after
power control

Name
TX i ic
n 0 UL
TX i ic
n UL
788
TX i ic
n 0 UL + nf
TX i ic

AT332_TRR_E0
11.2.9 Interference Calculation (UL)

Name
M
Value
M
TX ic TX jc
i
j
C UL + f O
I UL
+ f TL UL
TX i ic TX j jc
TX ic TX jc
i
j
10 Log r O
fO
j
10 Log TL UL
Mj
f TL UL
TX i ic
NR UL
TX i ic
I + N UL
Unit
Description
dBm
Uplink interference received at a cell
dB

dB

load
TX i ic
IMj
n
UL
UL
non-seg M i
---------------------
TX i ic
Inter Tech
10
-------------------------------------------10 Log
+ NR UL
n UL dB
10
10
+ 10
All M j
All TX j jc
TX i ic
TX i ic
Uplink noise at a cell
dBm
Unit
Description
dB
Unit
Description
dB
Unit
Description
W Channel 10
-----------------------------------TX i ic
N SCa Total
kHz
Inter-subcarrier distance
1
------------------TX i ic
F
sec
Useful symbol duration
sec
sec
Symbol duration
Symbols
Total cell resources
+ n UL
NRUL

Name
Value
TX i ic
Mi
Mi
CNR UL
C UL n UL
M
TX ic
i
UL
With MIMO: CNR UL + G Div UL + G Div

Name
Value
TX i ic
Mi
Mi
CINR UL
CNR UL NR UL
Mi
TX i ic
UL
11.2.12 Calculation of Total Cell Resources

Name
TX ic
i
TX i ic
D Sym Useful
Value
TX i ic
TX i ic
TX i ic
r CP
-------------F
D CP
TX ic
i
D Symbol
TX i ic
R DL
TX ic
i
TX ic
i
D Sym Useful + D CP
TX ic
1
- N SCai Data
Floor ----------------TX i ic
D
Symbol
789

11.2.13 Channel Throughput, Cell Capacity, and Per-user Throughput

Calculation
Name
Value
TX i ic
R DL
M
CTP P DL
With MIMO (AMS):

B
Mi
DL
Mi
i
i
Mi
Mi
f TP Scaling
- TP Offset
CTP E DL -----------------------100
TX i ic
Mi
CTP P DL TL DL Max
M
Mi
i
i
Cap P DL 1 BLER BDL
Mi
Mi
f TP Scaling
- TP Offset
Cap E DL -----------------------100
Cap E DL
Cap A DL

throughput
kbps

throughput
kbps
kbps
kbps
kbps

user
kbps

per user
kbps

user
kbps

throughput
kbps

throughput
kbps

throughput
kbps
kbps
Mi
Mi
Mi
kbps
TX i ic
Mi
Cap P DL

throughput
T AMS
if CNR DL
CTP A DL
kbps
DL
TX i ic
CTP E DL
Description
M
i
B DL
Max
Unit
Mi
Mi
Mi
Cap P DL
----------------------TX i ic
N Users DL
Mi
PUTP P DL
PUTP E DL
Mi
PUTP A DL
Mi
i
PUTP E DL
M
f TP Scaling
i
- TPOffset
-----------------------100
TX i ic
R UL
With MIMO (AMS):
Mi
B UL
Mi
CTP P UL
Cap E DL
----------------------TX ic
i
N Users DL
Mi
Mi
B UL
Max
Mi
B UL
TX i ic
if CNR DL
TX i ic
T AMS

TX ic
i
R UL
Mi
B UL
TX ic
--------------------------------- G MUi MIMO

D Frame
M
CTP E UL
Mi
CTP A UL
Mi
Mi
i
CTP E UL
M
f TP Scaling
i
- TP Offset
-----------------------100
Mi
TX i ic
Cap P UL
CTP P UL TL UL Max
i
i
Cap E UL
790
i
i

AT332_TRR_E0
Name
M
Value
M
Cap A UL
i
Cap E UL
PUTP P UL
Description
kbps
kbps

user
kbps

user
kbps

user
M
f TP Scaling
i
------------------------- TP Offset
100
i
Cap P UL
----------------------TX i ic
N Users UL
Mi
Unit
Mi
Cap E UL
----------------------TX i ic
N Users UL
Mi
PUTP E UL
Mi
PUTP A UL
Mi
Mi
f TP Scaling
100
Mi

Name
Value
Unit
Description
Sel
Mi
R Min DL
TPD Min DL
---------------------------
None

demand in downlink
None

demand in uplink
None

throughput demands
R Min UL
None

demands
Sel
Mi
Sel
Mi
kbps

mobile in downlink
Sel
Mi
Sel
Mi
kbps

mobile in uplink
None

mobile in downlink
None

mobile in uplink
None

demand in downlink
None

demand in uplink
Sel
i
Sel
i
CTP P DL
Sel
Mi
R Min UL
TX i ic
R Rem DL
TX i ic
R Rem UL
Sel
Mi
TPD Rem DL
Sel
Mi
Sel
Mi
TPD Min UL
--------------------------Sel
Mi
CTP P UL
TX i ic
TL DL Max
Sel
Mi
R Min DL
Sel
Mi
TX ic
i
TL DL Max
Sel
i
Sel
Mi
TPD Max DL TPD Min DL
TPD Rem UL
TPD Max UL TPD Min UL
Sel
Mi
RD Rem DL
TPD Rem DL
----------------------------
Sel
Mi
Sel
Mi
CTP P DL
M
Sel
i
Sel
Mi
RD Rem UL
TPD Rem UL
----------------------------
Sel
i
R Max DL
Mi
R Rem DL
-
Min RD Rem DL -------------------N
Sel
Mi
R Max UL
Sel
i
CTP P UL
M
Sel
Sel
TX i ic
TX i ic
R Rem UL
Mi
-
Min RD Rem UL -------------------N
791

Name
Value
Unit
Description
Sel
Sel
M
Mi
i
R Max DL CTP E DL
Sel
M
Site
i
-
Max 1 -----------------------------------------------------------------------------------------------------Sel
Sel
Mi
Mi
Site
TP
CTP
Min DL
BH DL
E DL
Sel
M i Site
None

downlink
Sel
Sel
Mi
Mi
CTP
Max UL
UL
Sel
M i Site
Max 1 -------------------------------------------------------------------------------------------------------
Sel
Sel
Mi
Mi
Site
R Min UL CTP E UL
TP BH UL
Sel
M i Site
None
None

in downlink
None

in uplink
Unit
Description
kbps
Downlink peak MAC user throughput
kbps
Downlink effective MAC user

throughput
kbps

throughput
kbps
Uplink peak MAC user throughput
kbps
Uplink effective MAC user

throughput
kbps
Site
BHOF DL
Site
BHOF UL
Sel
Sel
Mi
TL DL
Sel
Mi
= R DL
Sel
Mi
Sel
Mi
Mi
Mi
Max DL CTP P DL

BHOF DL
-----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P DL
Sel
Sel
Mi
TL UL
Sel
Mi
= R UL
Sel
Mi
Sel
Mi
Mi
Mi
R Max UL CTP P UL

BHOF UL
-----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P UL

Name
Sel
Mi
UTP P DL
Sel
i
UTPE DL
M
Sel
Mi
UTP A DL
Sel
Mi
UTP P UL
Sel
Mi
UTP E UL
Sel
Mi
UTP A UL
792
Value
Sel
Mi
R DL
Sel
Mi
CTP P DL
Sel
Sel
Mi
i
Sel
Mi
Sel
Mi
Sel
Mi
f TP Scaling
- TP Offset
UTP E DL -----------------------100
Sel
Mi
R UL
Sel
Mi
CTP P UL
Sel
Sel
Mi
Mi
UTPP UL 1 BLER B UL
Sel
Mi
Sel
Mi
Sel
Mi
f TP Scaling
- TP Offset
UTP E UL -----------------------100

AT332_TRR_E0

TX ic
i
Downlink signal level C DL
Path loss L Path
,G

for which calculated path loss matrices are available.
depends on the signal level of the best server. The bar graph displays cells whose received signal levels are higher than their
C/N thresholds and are within a 30 dB margin from the highest signal level.
You can use a value other than 30 dB for the margin from the highest signal level, for example a smaller value for improving
the calculation speed. For more information on defining a different value for this margin, see the Administrator Manual.

Atoll displays the best server signal level and interference from other cells.
Interference level bar graphs show the interference levels in decreasing order. The maximum number of bars in the graph
depends on the highest interference level. The bar graph displays cells whose C/N are higher than the minimum interferer C/
N threshold and whose interference levels are within a 30 dB margin from the highest interference level.
You can use a value other than 30 dB for the margin from the highest interference level, for example a smaller value for
Manual.

11.3.2.1 Signal Level Coverage Predictions
The following coverage predictions are based on the received signal levels:
Overlapping Zones
For these calculations, Atoll calculates the received signal level, then determines the selected display parameter on each pixel
inside the cells calculation area. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver.
L
Mi
,G
Mi
Mi
Mi
, L Ant , and L Body are not considered in the calculations performed for the signal level based coverage predictions.
For more information on signal level calculations, see "Signal Level Calculation (DL)" on page 807

793


All Servers
TX i ic
MinimumThreshold C DL
TX ic
TX ic
i
or L i
Total or L Path MaximumThreshold

TX i ic
TX ic
TX ic
i
or L i
AND
TX i ic
C DL
TX j jc
Best C DL M
ji
If M = 0 dB, Atoll considers pixels where the received signal level from TXi(ic) is the highest.
If M = 2 dB, Atoll considers pixels where the received signal level from TXi(ic) is either the highest or within a 2 dB
margin from the highest.
If M = -2 dB, Atoll considers pixels where the received signal level from TXi(ic) is 2 dB higher than the received
signal levels from the cells which are 2nd best servers.

TX i ic
TX ic
TX ic
i
or L i
AND
TX i ic
C DL
TX jc
nd
j
2 Best C DL
ji
values.
If M = 0 dB, Atoll considers pixels where the received signal level from TXi(ic) is the second highest.
If M = 2 dB, Atoll considers pixels where the received signal level from TXi(ic) is either the second highest or within
a 2 dB margin from the second highest.
signal levels from the cells which are 3rd best servers.

such as:
794

level.
Path Loss (dB)
Total Losses (dB)
highest signal level) and evaluates the path loss from this cell.
highest signal level) and evaluates the total losses from this cell.

AT332_TRR_E0
of these cells).

The following coverage predictions are based on the received signal levels and noise, and take into account the receiver
characteristics ( L
Mi
,G
Mi
Mi
Mi

For these calculations, Atoll calculates the received signal level or C/N level at each pixel. Each pixel within the calculation area
of TXi(ic) is considered a non-interfering receiver. The properties of the non-interfering probe receiver are set by selecting a
terminal, a mobility type, and a service.




page 819.
options:
Signal Level (DL) (dBm)

C/N Level (DL) (dB)
options:

C/N Level (UL) (dB)


Mi
,G
Mi
Mi
Mi
795


"Noise Rise Calculation (UL)" on page 815
For more information on throughput calculations, see:
"Channel Throughput, Cell Capacity, and Per-user Throughput Calculation" on page 821.


page 819.
options:
C/(I+N) Level (DL) (dB)

Total Noise (I+N) (DL) (dBm)
options:
Bearer (DL)
Service
options:

display options:
Quality indicators available in the document (Quality Indicators table): Atoll calculates the downlink C/(I+N) levels
quality indicator from the quality graphs defined in the Wi-Fi equipment of the selected terminal.
options:
796


AT332_TRR_E0
options:
Bearer (UL)
Service
options:

display options:
Quality indicators available in the document (Quality Indicators table): Atoll calculates the uplink C/(I+N) levels
quality indicator from the quality graphs defined in the Wi-Fi equipment of the best serving cell.

station.

"Throughput Calculation" on page 820.


797


[Simulation]
RandomTotalUsers=0
11.3.4.1.1

points.
Voice Service (v)
N Call D Call
UL
DL
UL
DL

UL
UL
DL
798

AT332_TRR_E0
DL
DL
UL

UL + DL
UL
DL
UL
UL

DL
DL

UL + DL
UL + DL
Therefore, a user can be either active on both links, inactive on both links, active on UL only, or active on DL only.
Data Service (d)
UL
DL
TP Average
DL
and the uplink
and the uplink V
UL
TP Average
UL
during a session.
for the service d.
UL
DL
N Session V 8
N Session V 8
DL
= ------------------------------------------ and f = -----------------------------------------UL
DL
TP Average 3600
TP Average 3600
UL
DL

UL

DL
UL
DL
1 f
DL
UL
1 f
UL + DL
UL
DL

UL
UL

DL
DL

UL + DL
UL + DL
UL
DL
UL + DL

799

11.3.4.1.2

Sector traffic maps per sector are also referred to as live traffic maps. Live traffic data from the OMC is spread over the best
server coverage areas of the transmitters included in the traffic map. Either throughput demands per service or the number
of active users per service are assigned to the coverage areas of each transmitter.

N
UL
UL
DL
TP Cell
TP Cell
DL
= ---------------------- and N = ---------------------UL
DL
TP Average
TP Average
UL
DL
UL
DL

UL

DL
UL
and N
DL
UL
UL + DL
DL
accurately determine the number of active users in the uplink ( n Active ), in the downlink ( n Active ), and both ( n Active ).
UL
DL

UL
DL

UL
UL
DL

DL
DL
UL

UL + DL
UL
DL
We have: N
UL
UL + DL
UL
DL
UL + DL
DL
UL
UL + DL
DL
UL + DL
N p Active
N p Active
UL + DL
--------------------------------------Number of users active in the uplink and downlink both: n Active = Min -------------------------------------- or
UL
UL + DL
DL
+ DL
Active
UL + DL
UL
DL
f Act N
DL
UL
f Act
UL

DL
UL

UL
DL
UL + DL
n Active
DL
UL + DL
n Active
UL + DL
800

AT332_TRR_E0
nv
- p Inactive
them.

Each Monte Carlo simulation in Atoll Wi-Fi is a snap-shot of the network with resource allocation carried out over a duration
of 1 second. The steps of this algorithm are listed below.
TX i ic
Cell transmission power ( P DL
) is set to the value defined by the user.
Cell loads ( TL DL
Mi
TX i ic
TX i ic
, TL UL
TX i ic
, and NR UL
) are set to their current values in the Cells table.
3. Determines the best servers for all the mobiles generated for the simulation as explained in "Best Server
Figure 11.1: Wi-Fi Simulation Algorithm

4. Determines the mobiles which are within the service areas of their best serving cells as explained in "Service Area
801

5. Determines the downlink and uplink C/(I+N) and bearers for each of these mobiles as explained in "C/(I+N) and Bearer
Calculation (DL)" on page 812 and "C/(I+N) and Bearer Calculation (UL)" on page 818 respectively.
6. Determines the channel throughputs at the mobile as explained in "Channel Throughput, Cell Capacity, and Per-user
Throughput Calculation" on page 821.
mobile according to the throughput demands of each mobile using the selected scheduler as explained in "Scheduling
and Radio Resource Allocation" on page 825.
as follows:
Calculation of Traffic Loads: Atoll calculates the traffic loads for all the cells TXi(ic).
TX i ic
TL DL
TX i ic
Mi
RDL and TLUL
Mi
RUL
Mi
Mi
TX i ic
MU MIMO
Mi
For uplink MU-MIMO, TL UL
RC UL
MU MIMO
Mi
Calculation of Uplink Noise Rise: For each victim cell TXi(ic), the uplink noise rise is calculated and updated by
considering each interfering mobile Mj as explained in "Noise Rise Calculation (UL)" on page 815.
Calculation of Uplink MU-MIMO Gain:
Atoll calculates the uplink MU-MIMO gain for all the cells as follows:
MU MIMO
Mi
TX i ic
G MU MIMO =
R UL
MU MIMO
Mi
------------------------------------------------------------MU MIMO
Mi
RC UL
MU MIMO
Mi
MU MIMO
Mi
Where
R UL
is the sum of the percentages of the uplink cell resources allocated to MU-MIMO
MU MIMO
Mi
MU MIMO
Mi
mobiles and
RC UL
is the sum of the real resource consumption of MU-MIMO mobiles.
MU MIMO
Mi
TX i ic
TL DL
TX i ic
TL UL
TX i ic
Req
TL DL
TL UL
TX i ic
TX ic
i
Max NR UL
All TX ic
i
TX i ic
If TL DL
TX i ic
TX ic
i
Max TL UL
All TX ic
i
TX i ic
NR UL
TX ic
i
Max TL DL
All TX ic
, TL UL
k 1
k 1
TX i ic
NR UL
k 1
TX i ic
Req
, and NR UL
Req

TX i ic
TL DL
802
TX i ic
TL DL
TX i ic
Req
AND TL UL
TX i ic
TL UL
TX i ic
Req
AND NR UL
TX i ic
NR UL
Req

AT332_TRR_E0
when creating the simulation, if:
TX ic
i
TL DL
TX ic
i
TL DL
TX ic
i
Req
OR TL UL
TX ic
i
TL UL
TX ic
i
Req
OR NR UL
TX ic
i
NR UL
Req
Simulation Results

Uplink traffic load
Uplink noise rise
Uplink MU-MIMO capacity gain

due to:
No Coverage: If the mobile does not have any best serving cell (step 3.) or if the mobile is not within the service area
of its best server (step 4.).
No Service: If the mobile is not able to access a bearer in the direction of its activity (step 5.), i.e., UL, DL, or DL+UL.
Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 7.)
uplink, the minimum uplink throughput demand is higher than the uplink allocated bandwidth throughput (step 7.)
Backhaul Saturation: If allocating resources to a mobile makes the effective MAC aggregate site throughputs exceed
the maximum backhaul throughputs defined for the site. This condition is only verified if the simulation was created



A Wi-Fi network can consist of cells that use different channel bandwidths. Therefore, the start and end frequencies of all the
channels may not exactly coincide. Channel bandwidths of cells can overlap each other with different ratios.
803


TX i ic
TX i ic
TX i ic
TX i ic
N Channel + 1 .


Input
TX i ic
TX j jc
F Start FB and F Start FB : Start frequency of the frequency band assigned to the cells TXi(ic) and TXj(jc).
F Start FB can represent the uplink or the downlink start frequencies ( F Start FB UL or F Start FB DL ).
First TX i ic
N Channel
TX ic
i
N Channel
and
First TX j jc
and N Channel
TX jc
j
N Channel
: Channel numbers assigned to cells TXi(ic) and TXj(jc).
Atoll considers that the same channel number is assigned to a cell in the downlink and uplink, i.e., the channel number
you assign to a cell is considered for uplink and downlink both.
TX i ic
TX j jc
ICS FB
TX i ic
TX i ic
CN FB
TX j jc
and ICS FB
: Inter-channel spacing of the frequency bands assigned to cells TXi(ic) and TXj(jc).
TX j jc
and CN FB
: Channel number step of the frequency bands assigned to cells TXi(ic) and TXj(jc).
Calculations
804

AT332_TRR_E0
For cell TXi(ic):

TX ic
i
TX ic
i
TX ic
i
TX i ic
F End
TX i ic
N TX i ic N First TXi ic
Channel
Channel
-
------------------------------------------------------TX ic
CN FB
TX ic
i
F Start = F Start FB + W Channel + ICS FB

TX i ic
For cell TXj(jc):

TX j jc
F Start
TX j jc
F End
TX j jc
TX jc
TX j jc
Channel
Channel
-------------------------------------------------------
TX j jc

CN FB
TX jc
j
j
= F Start FB + W Channel + ICS FB
TX j jc
= F Start FB + W Channel
Output
TX i ic
TX j jc
F End
TX i ic
TX j jc
and F End

Input
TX i ic
TX j jc
F Start
TX i ic
TX j jc
F End
and F End
TX ic
i
Calculations
TX i ic
TX j jc
F Start F End
TX i ic
AND F End
TX j jc
F Start

TX i ic TX j jc
W CCO
TX jc
TX ic
TX jc
TX ic
j
i
j
i
= Min FEnd F End Max F Start F Start

TX i ic TX j jc
r CCO
TX i ic TX j jc
W CCO
= ---------------------------------TX i ic
W Channel
Output
TX i ic TX j jc
r CCO
805


Input
TX ic
i
TX jc
j
F Start
TX i ic
TX j jc
F End
and F End
TX i ic
Calculations
TX i ic
TX i ic
TX j jc
TX i ic
TX j jc
AND F Start F Start

TX i ic
F End
TX j jc
F End
TX i ic
AND F End
TX i ic
TX j jc
+ W Channel F Start

TX i ic TX j jc
W ACO
TX jc
TX ic
TX jc
TX ic
TX ic
j
i
j
i
i

TX i ic TX j jc
W ACO
L
= ---------------------------------TX i ic
W Channel
TX ic TX jc
i
j
r ACO
L

TX i ic TX j jc
W ACO
TX jc
TX ic
j
i
= Min F End F End
TX ic
TX jc
TX ic
i
j
i

TX i ic TX j jc
W ACO
H
= ---------------------------------TX i ic
W Channel
TX i ic TX j jc
r ACO

TX i ic TX j jc
r ACO
TX i ic TX j jc
= r ACO
TX i ic TX j jc
+ r ACO
Output
TX ic TX jc
i
j
r ACO

Input
TX i ic TX j jc
r CCO
806

AT332_TRR_E0
TX ic TX jc
i
j
r ACO

TX ic
i
f ACS FB : Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic).
W Channel and W Channel : Bandwidths of the channels assigned to the cells TXi(ic) and TXj(jc).
TX ic
i
TX jc
j
Calculations
TX i ic TX j jc
rO
TX ic
f ACS FB
----------------------------
10
j
i
j
r i
+ r ACO
10
CCO
TX ic
i
TX jc
j
TX ic
f ACS FB
TX i ic
----------------------------
TX i ic
TX j jc
W Channel
10
i
j
i
j
r
--------------------if W Channel W Channel

+ r ACO
10
TX j jc
CCO
W Channel
TX i ic
W Channel
TX j jc
W Channel
TX j jc
TX j jc
TX i ic
W Channel
TX j jc
W Channel
TX j jc
TX j jc
Output
TX i ic TX j jc
rO
11.4.2 Signal Level and Quality Calculations

The following sections describe how signal levels, noise and interference, C/N, and C/(I+N) ratios are calculated on the
downlink and uplink.


Input
TX i ic
P DL
TX i
TX i
: Transmission power of the cell TXi(ic).
: Transmitter antenna gain for the antenna used by the transmitter TXi. ( G
TX i
TX i
TX i
= G Ant ).
= L Total DL ).
807


TX
TX
is selected.
Mi
Mi
Mi
Mi
Mi
Mi
Calculations
TX i ic
C DL
= EIRP
TX i ic
Mi
Mi
Mi
Mi
L Ant L Body
EIRP
TX i ic
TX i ic
= P DL
+G
TX i
TX i
If you wish to exclude the energy corresponding to the cyclic prefix (guard interval) in the
total symbol duration from the useful signal level, you must add the following lines in the
Atoll.ini file:
[WiMAX]
TX i ic
When this option is active, the cyclic prefix energy is excluded from C DL
TX ic
i
TX ic
is added to C i
DL
. In other
Output
TX i ic
C DL
: Received signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.

For determining the C/N and C/(I+N), Atoll calculates the downlink noise over the channel bandwidth used by the cell. The
used bandwidth depends on the number of used subcarriers. The downlink noise comprises thermal noise and the noise figure
Input
808

AT332_TRR_E0
TX ic
i
N SCa Used : Number of used subcarriers defined for the frame configuration of a cell TXi(ic).
nf
TX ic
i
M
Calculations
n 0 = 10 Log K T 1000 = 174 dBm/Hz
TX i ic
n 0 DL
N TXi ic
SCa Used
= n 0 + 10 Log ------------------------ TXi ic
N SCa Total
TX i ic
n DL
TX i ic
= n 0 DL + nf
Mi
Output
TX i ic
n DL

cells TXi(ic) and TXj(jc), and on the traffic loads of the interfering cells TXj(jc).
Input
TX j jc
C DL
: Received signal level from the cell TXi(ic) as explained in "Signal Level Calculation (DL)" on page 807.

In Monte Carlo simulations, the received signal levels from interferers already include M Shadowing Model , as
explained in "Signal Level Calculation (DL)" on page 807.
information, see "Shadow Fading Model" on page 90). As the received signal levels from interferers already include
M Shadowing Model , M Shadowing C I is added to the signal levels from interferers in order to achieve the ratio
TX j jc
C DL
TX j jc
= C DL
+ M Shadowing C I
is selected.
TX j jc
TL DL
TX i ic TX j jc
rO

Calculations
Interference (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows:
809

TX jc
j
I DL
TX jc
j
= C DL
TX ic TX jc
i
j
+ fO

TX jc
j
Inter Tech
+ f TL DL + I DL
If you wish to exclude the energy corresponding to the cyclic prefix (guard interval) in the
total symbol duration from the useful signal level, you must add the following lines in the
Atoll.ini file:
[WiMAX]
TX i ic
When this option is active, the cyclic prefix energy is excluded from C DL
TX ic
i
TX ic
is added to C i
DL
. In other
Independent of the option, interference levels are calculated for the total symbol
Calculations for the interference reduction factors due to channel overlapping and traffic load are explained below:
as follows:
TX i ic TX j jc
fO
TX i ic TX j jc
= 10 Log r O
Interference reduction due to interfering cells traffic load:

The interference reduction factor due to the interfering cells traffic load is calculated as follows:
TX j jc
TX j jc
f TL DL = 10 Log TL DL

Inter Tech
I DL
TX k
P DL Rec
---------------------------------------
=
F TX i ic TX k
TX k ICP DL
TX k
F TX i ic TX k
ICP DL

TX k
Output
TX j jc
I DL
I DL
: Downlink interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc).
Inter Tech

Input
TX i ic
C DL
: Received signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Signal Level
810
TX i ic
n DL
TX i ic
: Downlink noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 808.

AT332_TRR_E0
M
T B : Bearer selection thresholds of the bearers defined in the Wi-Fi equipment used by Mis terminal.
or mobile Mi.
or mobile Mi.
TX i ic
mobile Mi.
i
BLER B DL : Downlink block error rate read from the graphs available in the Wi-Fi equipment assigned to the terminal

used by the pixel, subscriber, or mobile Mi.

Calculations
The C/N for a cell TXi(ic) are calculated as follows for any pixel, subscriber, or mobile Mi:
TX i ic
CNR DL
TX i ic
= C DL
TX i ic
n DL
The bearers available for selection in the pixel, subscriber, or mobile Mis Wi-Fi equipment are the ones:
Whose selection thresholds are less than the C/N at Mi: T B CNR DL
TX i ic
Mi
Mi
If the cells frame configuration supports AMS, the STTD/MRC or SU-MIMO diversity gain, G Div DL , corresponding
to the bearer is applied to its selection threshold. The gain is read from the properties of the Wi-Fi equipment
TX ic
i
i
i
assigned to the pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX , Mobility M i , BLER BDL .

DL
applied. Therefore, the bearers available for selection are all the bearers defined in the Wi-Fi equipment for which
the following is true:
M
TX ic
i
DL
T B G Div DL G Div CNR DL
The bearer selected for data transfer is the one with the highest index.
MIMO STTD/MRC and SU-MIMO Diversity Gains:
Once the bearer is known, the C/N calculated above become:
TX i ic
TX i ic
= CNR DL
CNR DL
Mi
DL
+ G Div DL + G Div
Mi
Output
TX i ic
CNR DL
: C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
811


adjacent channel overlap between the studied and the interfering cells, and the traffic loads of the interfering cells. Finally,
Atoll takes the ratio of the signal level and the sum of the total interference from other cells and the downlink noise (as
calculated in "Noise Calculation (DL)" on page 808).
Input
TX i ic
C DL

TX i ic
n DL
I DL
TX j jc
: Interference from any cell TXj(jc) calculated for a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) as
explained in "Interference Calculation (DL)" on page 809.

Inter Tech
NRDL
T B : Bearer selection thresholds of the bearers defined in the Wi-Fi equipment used by Mis terminal.
TX i ic
Mi
Mi
or mobile Mi.
M
or mobile Mi.
TX i ic
mobile Mi.
i
BLER BDL : Downlink block error rate read from the graphs available in the Wi-Fi equipment assigned to the terminal

used by the pixel, subscriber, or mobile Mi.

Inter Tech
I DL
Calculations
The downlink C/(I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX i ic
CINR DL
TX i ic
= C DL
TXj jc
TX i ic
n DL
IDL
Inter Tech
Inter Tech
-
-----------------+ -------------------+ NR DL
10 + I DL
10 Log
10
10
10
All TXj jc
The Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX ic
TX i ic
I + N DL
TX j jc
n DL
I DL
---------------------
Inter
Tech
10
-----------------
+ NR Inter Tech
= 10 Log
+ 10
10 + I DL
DL
10
All TX j jc
812

AT332_TRR_E0
The bearers available for selection in the pixel, subscriber, or mobile Mis Wi-Fi equipment are the ones:
Whose selection thresholds are less than the downlink C/(I+N) at Mi: T B CINR DL
TX i ic
Mi
Mi
If the cells frame configuration supports AMS, the STTD/MRC or SU-MIMO diversity gain, G Div DL , corresponding
to the bearer is applied to its selection threshold. The gain is read from the properties of the Wi-Fi equipment
TX ic
i
i
i
assigned to the pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX , Mobility M i , BLER BDL .
DL
Mi
Mi
TX i ic
DL
T B G Div DL G Div CINR DL
Once the bearer is known, the C/(I+N) calculated above become:
TX i ic
CINR DL
TX i ic
= CINR DL
Mi
DL
+ G Div DL + G Div
Mi
Output
TX i ic
CINR DL
: Downlink C/(I+N) from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX ic
i
N DL
I +
: Total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by a cell TXi(ic).
Mi

Input
Mi
control.
Mi
TX i
TX i
: Transmitter antenna gain for the antenna used by the transmitter TXi ( G
TX i
TX i
TX i
= G Ant ).
= L Total UL ).
TX i
TX i
is selected.
813


i
Mi
Mi
Mi
Calculations
Mi
Mi
TX i
TX i
Mi
Mi
L Ant L Body
Mi
EIRP UL = P
With P
Mi
+G
Mi
Mi
Output
Mi

used bandwidth depends on the number of used subcarriers. The uplink noise comprises thermal noise and the noise figure
Input
N SCa Used : Number of used subcarriers defined for the frame configuration of a cell TXi(ic).
nf
TX i ic
TX i ic
TX i ic
Calculations
n 0 = 10 Log K T 1000 = 174 dBm/Hz
TX i ic
n 0 UL
N TXi ic
SCa Used
= n 0 + 10 Log ------------------------TX i ic
N SCa Total
TX ic
i
n UL
814
TX ic
i
= n 0 UL + nf
TX ic
i

AT332_TRR_E0
Output
TX ic
i
n UL
11.4.2.8 Interference Calculation (UL)

The uplink interference is only calculated during Monte Carlo simulations. In coverage predictions, the uplink noise rise values
already available in simulation results or in the Cells table are used.
The interference received by a cell TXi(ic) from an interfering mobile covered by a cell TXj(jc) can be defined as the uplink signal
level received from interfering mobiles Mj depending on the overlap that exists between the channels used by the cells TXi(ic)
and TXj(jc), on the traffic loads of the interfering mobile Mj.
11.4.2.8.1
Calculation of the uplink interference from each individual interfering mobile as explained in "Interference Signal
Levels Calculation (UL)" on page 815.
Calculation of the uplink noise rise which represents the total uplink interference from all the interfering mobiles as
Interference Signal Levels Calculation (UL)

Input
Mj
C UL : Uplink signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) as calculated in
TX i ic TX j jc
rO
Traffic loads are calculated during Monte Carlo simulations as explained in "Scheduling and Radio Resource
Calculations
Mj
Mj
TX i ic TX j jc
I UL = C UL + f O
Mj
+ f TL UL
Calculations for the interference reduction factors due to channel overlapping and uplink traffic load are explained below:
as follows:
TX i ic TX j jc
fO
TX i ic TX j jc
= 10 Log r O
Interference reduction due to interfering mobiles traffic load:

The interference reduction factor due to the interfering mobiles uplink traffic load is calculated as follows:
M
j
j
Output
11.4.2.8.2
Mj
I UL : Uplink interference signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc).

The uplink noise rise is defined as the ratio of the total uplink interference received by any cell TXi(ic) from interfering mobiles
Mj present in the coverage areas of other cells TXj(jc) to the uplink noise of the cell TXi(ic). In other words, it is the ratio (I+N)/N.
815

Input
M
I UL : Uplink interference signal levels received at a cell TXi(ic) from interfering mobiles Mj covered by other cells TXj(jc)
as calculated in "Interference Signal Levels Calculation (UL)" on page 815.

TX ic
i
n UL
Inter Tech
NRUL
: Uplink noise for the cell TXi(ic) as calculated in "Noise Calculation (UL)" on page 814.
Calculations
The uplink noise rise and total noise (I+N) for the cell TXi(ic) are calculated as follows:
TX ic
i
NR UL
TX ic
i
Mj
n UL
I UL
-------------------TX ic
10
Inter Tech
i
--------
= 10 Log
+ NR UL
n UL
10 + 10
10
All M j
All TX jc
For any pixel, subscriber, or mobile Mi in the interfered cell TXi(ic), Atoll calculates the uplink total noise (I+N) as follows:
TX i ic
I + N UL
TX i ic
= NR UL
TX i ic
+ n UL
Output
TX i ic
NRUL
I + N UL
TX i ic
: Total noise for a cell TXi(ic) calculated for any pixel, subscriber, or mobile Mi.

Input
Mi
n UL
TX i ic
TX i ic
T B : Bearer selection thresholds of the bearers defined in the Wi-Fi equipment used bythe cell TXi(ic).
Mi
Mi
Mi
Mi
or mobile Mi.
Mi
or mobile Mi.
Mi
mobile Mi.
816
TX i ic
i
BLER BUL : Uplink block error rate read from the graphs available in the Wi-Fi equipment assigned to the cell TXi(ic).


AT332_TRR_E0
Calculations
M
TX ic
i
CNR UL = C UL n UL
The bearers available for selection in the cell TXi(ic)s Wi-Fi equipment are the ones:
Mi
Mi
TX i ic
corresponding to the bearer is applied to its selection threshold. The gain is read from the properties of the Wi-Fi
M
TX ic
i
i
i
equipment assigned to the cell TXi(ic) for N Ant RX , N Ant TX , Mobility M i , BLER B UL .
UL
TX i ic
Mi
Mi
UL

M
TX ic
i
UL

TX i ic
bearer.
If with P
TX ic
i
M
i
B UL
+ M PC , where T
TX ic
i
M
i
B UL
is the bearer selection threshold, from the Wi-Fi
Mi
M i TX i ic
Mi
Mi
P Eff = Max P Max CNR UL T M + M PC P Min
B i
UL
Mi
Mi

Output
Mi
817


uplink C/N by the uplink noise rise value of the cell as calculated in "Noise Rise Calculation (UL)" on page 815.
Input
Mi
CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic) as calculated in "C/N Calculation (UL)"
on page 816.
TX i ic
NRUL
: Uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 815.
TX i ic
T B : Bearer selection thresholds of the bearers defined in the Wi-Fi equipment used bythe cell TXi(ic).
Mi
Mi
Mi
Mi
or mobile Mi.
Mi
or mobile Mi.
Mi
mobile Mi.
TX i ic
i
BLER BUL : Uplink block error rate read from the graphs available in the Wi-Fi equipment assigned to the cell TXi(ic).
Calculations
Mi
Mi
TX i ic
CINR UL = CNR UL NR UL
The bearers available for selection in the cell TXi(ic)s Wi-Fi equipment are the ones:
818
Whose selection thresholds are less than the uplink C/(I+N) at Mi: T B CINR UL
Mi
Mi

AT332_TRR_E0
TX ic
i
corresponding to the bearer is applied to its selection threshold. The gain is read from the properties of the Wi-Fi
M
TX ic
i
i
i

UL
M
TX ic
i
UL

Mi
Mi
TX i ic
UL

TX i ic
Where G Div UL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer.
Mi
Mi
TX i ic
UL

TX i ic
bearer.
If with P
Mi
Mi
Mi
TX i ic
Mi
+ M PC , where T
B UL
TX i ic
Mi
B UL
is the bearer selection threshold, from the Wi-Fi
M
M
TX ic
Mi
Mi
i
i
i
P Eff = Max P Max CINR UL T M + M PC P Min
i
UL
Mi
Mi

Output
Mi
CINR UL : Uplink C/(I+N) from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
Mi

In Wi-Fi, best server refers to a cell ("serving transmitter"-"reference cell" pair) from which a pixel, subscriber, or mobile Mi
gets the highest signal level. This calculation also determines whether the pixel, subscriber, or mobile Mi is within the coverage
area of any transmitter or not.
819

Input
TX ic
i
C DL
: Downlink signal level received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Signal
Level Calculation (DL)" on page 807 using the terminal and service parameters ( L
,G
, L Ant , and L Body ) of Mi.
Calculations
The best server of any pixel, subscriber, or mobile Mi, BS M , is the cell from which the received downlink signal level is the
i
highest among all the cells. The best server is determined as follows:
BS M = TX i ic
i
TX i ic
TX i ic
C DL
= Best
C
All TX i ic DL
Here ic is the cell of the transmitter TXi with the highest power. However, if more than one cell of the same transmitter covers
the pixel, subscriber, or mobile, the final reference cell ic might be different from the initial cell ic (the one with the highest
power). In coverage prediction calculations and in calculations on subsriber lists, the cell of the highest priority layer is
selected as the serving (reference) cell. In Monte Carlo simulations, a random cell is selected as the serving (reference) cell.
Output

i

In Wi-Fi, a pixel, subscriber, or mobile Mi can be covered by a cell (as calculated in "Best Server Determination" on page 819)
TXi(ic) if the downlink C/N from the cell at the pixel, subscriber, or mobile is greater than or equal to the minimum C/N
Input
TX i ic
CNR DL
: Downlink C/N from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "C/N Calculation (DL)"
on page 810.
TX i ic
T Min
: Min C/N threshold defined for the cell TXi(ic).
Calculations
TX ic
i
CNR DL
TX ic
i
T Min
Output

False: Otherwise.

page 820.
Calculation of throughputs as explained in "Channel Throughput, Cell Capacity, and Per-user Throughput Calculation"
on page 821.

The total amount of resources in a cell is the number of modulation symbols that can be used for data transfer per second.
820

AT332_TRR_E0
Input
TX ic
i
W Channel : Channel bandwidth of the cell TXi(ic).
N SCa Data : Number of data subcarriers defined for the frame configuration of a cell TXi(ic).
r CP
TX ic
i
TX i ic
TX ic
i
: Cyclic prefix ratio defined for the cells frame configuration of TXi(ic) or, otherwise, in the global network
settings.
Calculations
Atoll determines the inter-subcarrier spacing.
F
TX ic
i
TX i ic
W Channel 10
= ----------------------------------TX i ic
N SCa Total
Atoll calculates the useful symbol duration.

TX i ic
1
D Sym Useful = ------------------TX ic
i
F
And, the duration of the cyclic prefix (guard interval).

TX ic
i
D CP
TX i ic
r CP
= -------------F
Adding the cyclic prefix ratio to the useful symbol duration, Atoll determines the total symbol duration.
TX i ic
TX i ic
TX i ic
D Symbol = D Sym Useful + D CP
The total number of modulation symbols in the downlink and uplink are:
TX ic
i
R DL
TX ic
i
= R UL
TX ic
1
- N SCai Data
= Floor ----------------TX i ic
D
Symbol
Output
TX ic
i
TX ic
i
R DL
and R UL
: Amount of downlink and uplink resources in the cell TXi(ic).
11.4.5.2 Channel Throughput, Cell Capacity, and Per-user Throughput Calculation

capacities are similar to channel throughputs but upper-bound by the maximum downlink and uplink traffic loads. Per-user
throughputs are calculated by dividing the cell capacities by the average number of connects users, downlink or uplink,
defined for the cell.
Input
TX i ic
R DL
TX i ic
TX ic
i
page 820.
TX ic
i
R UL
page 820.
B
i
DL
821

M
B
i
UL
: Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the uplink in "C/
(I+N) and Bearer Calculation (UL)" on page 818.

TX ic
i
CNR DL
TX ic
i
T AMS
: Downlink C/N the cell TXi(ic) as calculated in "C/N Calculation (DL)" on page 810.
: AMS threshold defined for the cell TXi(ic).
TX i ic
TX ic
i
TX ic
i
i
BLER BDL : Downlink block error rate read from the BLER vs. CINR DL

M
graph available in the Wi-Fi equipment
i
i
BLER BUL : Uplink block error rate read from the BLER vs. CINR UL graph available in the Wi-Fi equipment assigned

Mi
Mi.
Mi
TX i ic
TX ic
i
Calculations
Downlink:
TX i ic
Mi
Peak MAC Channel Throughput: CTP P DL = R DL
Mi
B DL
MIMO SU-MIMO Gain:

Max
If the frame configuration supports AMS, SU-MIMO gain G SU MIMO is applied to the bearer efficiency. The gain is read
from the properties of the Wi-Fi equipment assigned to the pixel, subscriber, or mobile Mi for:
TX i ic
Mi
or mobile Mi.
B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink as explained in "C/(I+N) and Bearer
Mi
i
BLER B DL : Downlink block error rate read from the graphs available in the Wi-Fi equipment assigned to the
TX i ic
terminal used by the pixel, subscriber, or mobile Mi. BLER is determined for CINR DL
Mi
B DL
TX ic
i
Max
Mi
1 + f SU MIMO G SU MIMO 1 if CNR DL
TX ic
i
T AMS
B DL
822
i
i
i

AT332_TRR_E0
M
i
CTP A DL
i
CTP E DL
M
f TP Scaling
i
------------------------- TP Offset
100
Peak MAC Cell Capacity: Cap P DL = CTP P DL TL DL Max
i
i
i
TX ic
i
Mi
Mi
Mi
f TP Scaling
- TP Offset
= Cap E DL -----------------------100
Mi
Mi
Mi
Cap P DL
= ----------------------TX i ic
N Users DL
Mi
Mi
Cap E DL
Effective MAC Throughput per User: PUTP E DL = ----------------------TX i ic
N Users DL
i
PUTP A DL
Mi
i
PUTP E DL
M
f TP Scaling
i
- TP Offset
-----------------------100
Uplink:
TX i ic
Mi
Peak MAC Channel Throughput: CTP P UL = R UL
M
i
B UL
MIMO SU-MIMO Gain:

Max
If the frame configuration supports AMS, SU-MIMO gain G SU MIMO is applied to the bearer efficiency. The gain is read
from the properties of the Wi-Fi equipment assigned to the cell TXi(ic) for:
or mobile Mi.
TX i ic
B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink as explained in "C/(I+N) and Bearer
Mi
i
BLER B UL : Uplink block error rate read from the graphs available in the Wi-Fi equipment assigned to the cell

Mi

Mi
B UL
TX i ic
Max
Mi
TX i ic
T AMS
B UL
TX i ic
CNR DL
TX i ic
TX i ic
TX i ic
T MU MIMO and N Ant RX 2 , the MU-MIMO gain G MU MIMO is applied to the channel throughput. The
MU-MIMO gain is read from the properties of the cell TXi(ic).

Mi
Mi
TX i ic
i
i
i
823


M
i
CTP A UL
i
CTP E UL
M
f TP Scaling
i
------------------------- TP Offset
100
Peak MAC Cell Capacity: Cap P UL = CTP P UL TL UL Max
i
i
i
TX ic
i
Mi
Mi
Mi
f TP Scaling
- TPOffset
= Cap E UL -----------------------100
Mi
Mi
Peak MAC Throughput per User: PUTP P UL
Mi
Cap P UL
= ----------------------TX i ic
N Users UL
Mi
Mi
Cap E UL
Effective MAC Throughput per User: PUTP E UL = ----------------------TX i ic
N Users UL
i
PUTP A UL
i
PUTP E UL
Mi
M
f TP Scaling
i
- TPOffset
-----------------------100
Output
Mi
Mi
Mi
Mi
Mi
Mi
Mi
M
Mi
Mi
Mi
Mi
M
Mi
Mi
Mi
Mi
M

Wi-Fi scheduling and RRM algorithms are explained in "Scheduling and Radio Resource Allocation" on page 825 and the
calculation of user throughputs is explained in "User Throughput Calculation" on page 829.
824

AT332_TRR_E0

Input
TX ic
i
i
i
BLER B DL : Downlink block error rate read from the BLER vs. CINR DL
TX i ic
TX i ic
Mi
: Priority of the service accessed by a mobile Mi.

Mi
Mi
M
Mi
TX ic
graph available in the Wi-Fi equipment
i
i
BLER B UL : Uplink block error rate read from the BLER vs. CINR UL graph available in the Wi-Fi equipment assigned

M
CTP P DL : Downlink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
Mi
Mi
page 820.
Mi
CTP E DL : Downlink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 820.
CTP P UL : Uplink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 820.
CTP E UL : Uplink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 820.
Calculations
Mobile Selection:
TX i ic
TX i ic
TX ic
TX ic
TX ic
i
i
i
Sel
TX i ic

Sel
1. Atoll sorts the M i
Sel
TX i ic
N Users in order of decreasing service priority, p

Sel
= 1 up to M i
Sel
Mi
= N , Atoll allocates the downlink and uplink resources required to satisfy each
users minimum throughput demands in downlink and uplink as follows:
825

Sel
i
R Min DL
M
Sel
i
Sel
Sel
i
M
TPD Min DL
TPD Min UL
i
= -------------------------- and R Min UL = -------------------------M
Sel
i
CTP P DL
Sel
i
CTP P UL
When/If in downlink
Sel
Mi
TX i ic
Sel
Mi
When/If in uplink
Sel
Mi
TX i ic
Sel
Mi

one direction, it is rejected, and the resources, that were allocated to it in the one direction in which it was able to get
a throughput, are allocated to other mobiles.
5. If
Sel
Mi
TX i ic
Sel
Mi
Sel
Mi
TX i ic
Sel
Mi
throughput demands.
If at this stage, a sites downlink or uplink effective MAC aggregate throughput exceeds its maximum downlink or uplink
backhaul throughput, respectively, mobiles are rejected one by one due to Backhaul Saturation, starting from the mobile with
the lowest priority service, among all the cells of the site in order to reach a downlink or uplink effective MAC aggregate site
throughput the sites maximum downlink or uplink backhaul throughput.
TX i ic
TX i ic
Sel
Mi
R Min DL
Sel
Mi
TX i ic
TX i ic
Sel
Mi
R Min UL
Sel
Mi
For each mobile, the throughput demands remaining once the minimum throughput demands have been satisfied are the
difference between the maximum and the minimum throughput demands:
M
Sel
i
Sel
i
Sel
i
Downlink: TPD Rem DL = TPD Max DL TPD Min DL

Sel
Mi
Sel
Mi
Sel
Mi
Uplink: TPD Rem UL = TPD Max UL TPD Min UL

Sel
Let the total number of users with remaining throughput demands greater than 0 be N M i .
1. Atoll divides the remaining resources in the cell into equal parts for each user:
TX i ic
TX i ic
R Rem DL
R Rem UL
--------------------- and -------------------N
N
2. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands:
Sel
Mi
RD Rem DL
Sel
Mi
Sel
Mi
CTP P DL
826
Sel
Sel
Mi
Mi
TPD Rem DL
TPD Rem UL
= --------------------------- and RD Rem UL = --------------------------Sel
Mi
CTP P UL

AT332_TRR_E0
3. The resources allocated to each user for satisfying its maximum throughput demands are:
Sel
i
R Max DL
M
TX ic
i
Sel
Sel
Sel
TX ic
i
M
Mi
Mi
R Rem DL
R Rem UL
i
N
N
Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from the
remaining resources of the cell, whichever is smaller.
Sel
Mi
When/If in downlink
TX i ic
R Max DL = R Rem DL , i.e., the resources available in downlink have been used up for
Sel
Mi
When/If in uplink
Sel
Mi
TX i ic
R Max UL = R Rem UL , i.e., the resources available in uplink have been used up for satisfying
Sel
Mi
the maximum throughput demands of the mobiles.

5. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of
remaining users.
6. Atoll recalculates the remaining resources as follows:
TX ic
i
TX ic
i
Sel
i
Sel
Mi
TX i ic
TX i ic
R Min DL
Sel
i
R Max DL and
Sel
Mi
Sel
Mi
R Min UL
Sel
i
Sel
Mi
R Max UL
Sel
i
7. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied until
TX i ic
TX i ic
either R Rem DL = 0 and R Rem UL = 0 , or all the maximum throughput demands are satisfied.
Spatial Multiplexing with Uplink Multi-User MIMO:
Using MU-MIMO cells are able to schedule two users on two different antennas using the same frequency-time resources. A
mobile connected to antenna 1 makes resources available on antenna 2 for another mobile. These resources made available
on antenna 2 can be assigned to another mobile without increasing the overall cell load.
When the second mobile is assigned to antenna 2, it consumes the resources made available by the first mobile and, if it
requires more resources than the first mobile made available, the second mobile also makes some resources available on
antenna 1.
Each new mobile is either connected to antenna 1 or antenna 2. The part of the mobiles resources which are not coupled
with resources allocated to another mobile on the other antenna is called the real resource consumption. The part of the
mobiles resources which are coupled with the resources allocated to another mobile on the other antenna is called the virtual
resource consumption.
TX i ic
MU-MIMO can be used if the cell supports MU-MIMO, CNR DL
TX i ic
TX i ic
T MU MIMO , and N Ant RX 2 .
Let i be the index of connected MU-MIMO mobiles: i = 1 to N

MU MIMO
Each mobile M i
MU MIMO
Mi = 0
RR UL
MU MIMO
Mi
has a corresponding traffic load TL UL
MU MIMO
Mi = 0
= 100 % and available virtual resources V UL
. The scheduling starts with available real resources

= 0 % . i = 0 means no MU-MIMO mobile has yet
been scheduled.
MU MIMO
The virtual resource consumption of a mobile M i
MU MIMO
Mi
is given by: VC UL
MU MIMO
Mi
= Min TL UL
MU MIMO
Mi 1
V UL
827

MU MIMO
The real resource consumption of a mobile M i
MU MIMO
The virtual resources made available by the mobile M i

M
MU MIMO
i
V UL
MU MIMO
i1
= V UL
MU MIMO
i
VC UL
MU MIMO
Mi
RC UL
MU MIMO
i
is given by: RC UL
MU MIMO
i
= TL UL
MU MIMO
i
VC UL
are given by:
MU MIMO
i
+ RC UL
TX i ic
= TL UL Max .
The following table gives an example:

MU MIMO
i
TL UL
M
Mobile
(%)
MU MIMO
i
VC UL
M
(%)
MU MIMO
i
RC UL
M
MU MIMO
i
V UL
M
(%)
M1
10
10
10
M2
M3
20
15
15
M4
40
15
25
25
(%)

Sel
Sel
Mi
Mi
CTP
Max DL
DL
Sel
M i Site
= Max 1 ------------------------------------------------------------------------------------------------------- and
Sel
Sel
Mi
Mi
Site
R Min DL CTP E DL
TP BH DL
Sel
M i Site
Site
BHOF DL
Sel
Sel
Mi
Mi
CTP
Max UL
UL
Sel
M i Site
= Max 1 -------------------------------------------------------------------------------------------------------
Sel
Sel
Mi
Mi
Site
R Min UL CTP E UL
TP BH UL
Sel
M i Site
Site
BHOF UL

Sel
(which can also be

Sel
Sel
Mi
Downlink: TL DL
Sel
Mi
= R DL
Sel
Mi
Sel
Mi
M
Mi
i
Max DL CTP P DL

BHOF DL
= -----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P DL
Sel
Sel
Mi
Uplink: TL UL
Sel
Mi
= R UL
Sel
i
Sel
i
Mi
Mi
Max UL CTP P UL

BHOF UL
= -----------------------------------------------------------------------------------------------------------------Sel
Sel
Mi
CTP P UL
Output
828
Sel
Mi
TL DL
Sel
Mi
= R DL
Sel

AT332_TRR_E0
Sel
i
TL UL
Sel
i
Sel

Sel

Input
Sel
Mi
R DL
Sel
Sel
Mi
Sel
Sel
Mi
Sel
CTP P DL : Downlink peak MAC channel throughput at the mobile M i
page 820.
Sel
Mi
Sel
CTP P UL : Uplink peak MAC channel throughput at the mobile M i
page 820.
Sel
TX i ic
Mi
BLER B DL : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the Wi-Fi equipment
Sel
assigned to the terminal used by the mobile M i
Sel
Mi
Mi
BLER B UL : Uplink block error rate read from the BLER vs. CINR UL graph available in the Wi-Fi equipment assigned

Sel
Mi
Sel
Sel
i
Sel
Calculations
Downlink:
Sel
Mi
Sel
Mi
Sel
Mi
CTP P DL
Peak MAC User Throughput: UTP P DL = R DL
M
M
Mi
i
i
Effective MAC User Throughput: UTP E DL = UTP P DL 1 BLER B DL
Sel
Sel
Mi
Application User Throughput: UTP A DL
Sel
Sel
Sel
Mi
Sel
Mi
Sel
Mi
f TP Scaling
- TPOffset
= UTP E DL -----------------------100
Uplink:
M
Sel
i
Sel
i
Sel
i
CTP P UL
Peak MAC User Throughput: UTP P UL = R UL
Mi
Mi
Mi
Effective MAC User Throughput: UTP E UL = UTP P UL 1 BLER B UL
Sel
Sel
Mi
Sel
Sel
Mi
Sel
Sel
Mi
Sel
Mi
f TP Scaling
- TP Offset
= UTP E UL -----------------------100
829

Output
M
Sel
i
Sel
i
Sel
UTP P DL : Downlink peak MAC user throughput at the pixel, subscriber, or mobile M i
UTP E DL : Downlink effective MAC user throughput at the pixel, subscriber, or mobile M i .
UTP P UL : Uplink peak MAC user throughput at the pixel, subscriber, or mobile M i
UTP E UL : Uplink effective MAC user throughput at the pixel, subscriber, or mobile M i .
UTP A UL : Uplink application user throughput at the pixel, subscriber, or mobile M i
.
Sel
Sel
Mi
Sel
Sel
Mi
Sel
Sel
Mi
Sel
Sel
Mi
Sel



They are active,


zone.
discarded.
830

AT332_TRR_E0

[Neighbours]

decreasing rank.
neighbour.
i.
831

[Neighbours]
i

i
i
The received signal level is greater than or equal to the signal level threshold. The received signal level
TX i ic
( C DL
TX i ic
) and the signal level threshold are calculated from CNR DL

TX i ic
value of the noise ( n DL
, respectively, by adding the
) to them.
TX i ic
S TX ic is the surface area covered by TXi(ic) within C DL
TX i ic
and T Min
TX i ic
+ HO Start and C DL
+ HO End . HOStart is the
margin with respect to the best signal level at which the handover starts, and HO End is the margin with
respect to the best signal level at which the handover ends.

j
TX ic
i
If a global value of the C/N threshold ( T Min
) is set in the coverage conditions dialog
box, for each cell, Atoll uses the higher of the two values, i.e., global value and the
value defined for that cell.
gives the largest possible C/N coverage areas for the cells.
Atoll
S TX ic S TX jc
i
j
- 100 ),
i
832

AT332_TRR_E0

S TX ic S TX jc
i
j
- 100 % Min Coverage Area .
TXj(jc) is considered a neighbour of TXi(ic) if -------------------------------------S TX ic
i
Neighbourhood cause
When
Importance value
Existing neighbour

Existing importance
Exceptional pair
100 %
Co-site cell

selected
Adjacent cell

selected

relationship
transmitter.
d Di = 1 ---------d max
of antennas.
[Neighbours]
833


Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
30%
Min(A)
30%
Max(A)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

Neighbourhood cause
Importance Function

Co-site
Adjacent
No
No
10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
No
Yes
30%+30%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Yes
Yes
60%+40%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Where:
overlapping.
causes.

They are active,

834

AT332_TRR_E0
zone.
conditions:

[Neighbours]
CDMA Carriers: This option is available when a Wi-Fi network is being co-planned with a UMTS, CDMA, or TDSCDMA network. This option enables you to select the CDMA carrier(s) that you want Atoll to consider as potential
neighbours of Wi-Fi cells. You may choose one or more carriers. Atoll will allocate only the cells using the selected

neighbour.

This means that the signal received from A is greater than the minimum required (calculated from the C/N
threshold), and is the highest one.
835

The signal level received from A exceeds the minimum required (calculated from the C/N threshold) and is
within a margin from the highest signal level.
to 0 dB.
2nd case: The margin is other than 0 dB. SB is the area where:
level.
SA SB
SA
SA SB

Neighbour cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
100 %

distance conditions
d1 ---------d max
distance.

Neighbour cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
IF

IF
transmitter.
d Di = 1 ---------d max

836

AT332_TRR_E0
Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

cause
IF

No
10%+50%{10%(Di)+90%(O)}
Yes
60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.

Co-channel interference is the main reason for overall network quality degradation in Wi-Fi. In order to improve network
performance, the Wi-Fi AFP tries to minimise co- and adjacent channel interference as much as possible while respecting any
constraints input to it. The main constraints are the resources available for allocation, i.e., the number of frequencies with
The following describes the AFPs automatic planning method for frequencies in Wi-Fi networks, which takes into account
They are active,

zone.

837


You can modify these weights in your Wi-Fi document. The absolute values of the constraint weights are calculated from the
% Neighbour
% IM
% Dis tan ce

$
TX i ic TX j jc
TX i ic TX j jc
= rO
TX i ic TX j jc
Where r O
TX i ic TX j jc
Neighbour Neighbour
TX i ic TX j jc
TX ic TX j jc
+ i
IM
IM
page 803.
TX i ic TX j jc
TX i ic TX j jc
Neighbour
is
TX i ic TX j jc
IM
TX i ic TX j jc
TX i ic TX j jc
IM
= r CCO
TX i ic TX j jc
TX i ic TX j jc
IM CC
and IM CC
TX i ic TX j jc
IM CC
TX i ic TX j jc
+ r ACO
TX i ic TX j jc
IM AC
TX i ic TX j jc
TX i ic TX j jc
and r ACO
TX i ic TX j jc
Dis tan ce
TX i ic TX j jc
them. Dis tan ce
QRF
TX i ic TX j jc
= 1$
TX i ic TX j jc

cell has relations:
838

AT332_TRR_E0
TX ic
i
QRF
$ Total = 1
TX ic TX jc
i
j
TX jc
j
above, i.e.,
$ Total =
TX i ic
$ Total
TX i ic

Calculates the cost (as described above) of the initial frequency plan,
solution.
11.5.4 Appendices
S TX ic
i
TX ic TX jc
i
j
IM CC
TX j jc
TX i ic
+ M Quality
n DL
C DL
---------------------
----------------------------------------------------TX i ic
TX i ic
10
10
C DL
10 Log 10
+ 10
T Min
= ------------------------------------------------------------------------------------------------------------------------------------------------S TX ic
i

S TX ic
i
TX i ic TX j jc
IM AC
TX j jc
TX i ic
TX i ic
+M
+f
n
C DL
Quality
ACS
FB
DL
--------------------
-----------------------------------------------------------------------------------TX i ic
TX i ic
10
10
C DL
10 Log 10
+ 10
T Min
= ----------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX ic
i
TX i ic
Here S TX ic is the best server coverage area of the cell TXi(ic), that comprises all the pixels where CNR DL
i
calculated in "Service Area Calculation" on page 820. S TX ic

i
TX i ic
the given condition is true. C DL

TX i ic
respectively, n DL
TX j jc
and C DL
Condition
TX i ic
T Min
as
is the best server coverage area of the cell TXi(ic) where
are the received downlink signal levels from the cells TXi(ic) and TXj(jc)
the downlink noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 808, M Quality is
TX i ic
the quality margin used for the interference matrices calculation, and f ACS FB is the adjacent channel suppression factor
defined for the frequency band of the cell TXi(ic).

TX i ic TX j jc
839

TX ic TX jc
i
j
Dis tan ce
2
D Reuse
= Log --------------------------------
TX ic TX jc
j
D i

--------------------------------------------------------2
Log D Reuse
if D
TX ic TX jc
i
j
Otherwise
as follows:
D
D
TX ic TX jc
i
j
TX i ic TX j jc
them. d
= d
TX i ic TX j jc
TX ic TX jc
i
j
1 + x cos cos 2
TX i ic TX j jc
TX i ic TX j jc


840
Chapter 12
LPWA Networks

842

Chapter 12: LPWA Networks
AT332_TRR_E0
12 LPWA Networks
This chapter describes all the calculations performed in Atoll LPWA documents. The first part of this chapter lists all the input
parameters in the LPWA documents, their significance, location in the Atoll GUI, and their usage. It also contains the lists of
the formulas used for the calculations.
signal quality coverage predictions, and calculations on subscriber lists. The calculation algorithms used by these calculation
processes are available in the next part.
signal levels, noise, and interference for downlink and uplink, and radio resource management algorithms.
cells.
All the calculation algorithms in this section are described for:
Mi: A pixel (coverage predictions) or subscriber (calculations on subscriber lists)

covered/served by the studied cell TXi(ic).
12.1 Definitions
Name
Value
Unit
Description
1.38 x 10-23
J/K
Boltzmanns constant
290
Ambient temperature
n0
dBm/Hz

noise
M PC
Global parameter
dB
CNR Min
Global parametera
dB

W Total
Channel configuration parameter
None
Total bandwidth of the frequency

band
W Used DL
None
Downlink used bandwidth of the

frequency band
W Used UL
None
Uplink used bandwidth of the

frequency band
W Channel DL
None
Width of one downlink channel in a

frequency band
W Channel UL
None
Width of one uplink channel in a

frequency band
N Channel DL
None
Number of downlink channels in a

frequency band
N Channel UL
None
Number of uplink channels in a

frequency band
843

Name
Value
Unit
Description
F Start FB DL
MHz
DL Start frequency of the frequency

band
F Start FB UL
MHz
UL Start frequency of the frequency

band
Inter Tech
Network parameter
dB
reduction factor
Bearer parameter
None
Bearer index
Mod B
Bearer parameter
None
CR B
Bearer parameter
None
Bearer parameter
bps/Hz
Bearer efficiency
TB
Bearer parameter
dB
dB
N Ant TX
None

in transmission
N Ant RX
None

in reception
TX
Antenna parameter
dB
TX
dB
Transmitter loss
P DL
Cell parameter
dBm
Power
TL DL
Cell parameter
TL UL
Cell parameter
Uplink traffic load
TL DL Max
Cell parameter
TL UL Max
Cell parameter
NR UL
Cell parameter
dB
Uplink noise rise
N Users Max
Cell parameter
None
N Users DL
Cell parameter
None

cell in downlink
N Users UL
Cell parameter
None

cell in uplink
T AMS
Cell parameter
dB
T MU MIMO
Cell parameter
dB
T Min
Cell parameter
dB
Minimum C/N threshold
G MU MIMO
Cell parameter
None
Uplink MU-MIMO gain
Inter Tech
Cell parameter
dB
Inter Tech
Cell parameter
dB
G SU MIMO
Max
Cell LPWA equipment parameter
None
G Div UL
Cell LPWA equipment parameter
dB
Service parameter
None
Service priority
B DL Highest
Service parameter
None

the downlink
f IRF
nf
G
L
TX
NR DL
NR UL
844
Forsk <$year

AT332_TRR_E0
Name
Value
Unit
Description
B UL Highest
Service parameter
None

the uplink
B DL Lowest
Service parameter
None

the downlink
B UL Lowest
Service parameter
None

the uplink
f Act
UL
Service parameter
f Act
DL
Service parameter
TPD Min UL
Service parameter
kbps

uplink
TPD Min DL
Service parameter
kbps

downlink
TPD Max UL
Service parameter
kbps

uplink
TPD Max DL
Service parameter
kbps

downlink
UL
Service parameter
kbps

uplink
TP Average
DL
Service parameter
kbps

downlink
TP Offset
Service parameter
kbps
Throughput offset
f TP Scaling
Service parameter
Scaling factor
L Body
Service parameter
dB
Body loss
P Min
Terminal parameter
dBm
P Max
Terminal parameter
dBm
nf
Terminal parameter
dB
Terminal parameter
dB
Terminal parameter
dB
Terminal loss
N Ant TX
Terminal parameter
None

in transmission
N Ant RX
Terminal parameter
None

in reception
G SU MIMO
Max
Terminal LPWA equipment parameter
None
G Div DL
Terminal LPWA equipment parameter
dB

diversity gain
UL
Clutter parameter
dB
G Div
DL
Clutter parameter
dB
f SU MIMO
Clutter parameter
None
SU-MIMO gain factor
L Indoor
Clutter parameter
dB
Indoor loss
L Path
dB
Path loss
None

TP Average
G Div
F
ICP DL
Network parameter
845

Forsk <$year
Name
Value
Unit
Description
M Shadowing Model

dB
M Shadowing C I

dB
a.


Name
TX i ic
C DL
EIRP
Value
Unit
Description
dBm
Received signal level
dBm
EIRP of a cell
Value
Unit
Description
W TXi ic
Used DL
n 0 + 10 Log ----------------------TX i ic
W Total
dBm
dBm
Unit
Description
dBm
Interference generated by an
interfering cell
dB

downlink traffic load
interference
Unit
Description
dB
Downlink C/N for a cell
TX i ic

L
Mi
Mi
Mi
Mi
L Ant L Body
TX i ic
TX i ic
P DL
+G
TX i
TX i

Name
TX i ic
n 0 DL
TX i ic
TX i ic
n 0 DL + nf
n DL
Mi

Name
TX j jc
I DL
TX jc
j
f TL DL
Inter Tech
I DL
Value
TX j jc
C DL
TX i ic TX j jc
+ fO
TX j jc
Inter Tech
+ f TL DL + I DL
TX j jc
10 Log TL DL
TX k
P DL Rec
--------------------------------------
F TX i ic TX k
TX k ICP DL

Name
TX i ic
CNR DL
Value
TX i ic
C DL
TX i ic
With MIMO: CNR DL
846
TX i ic
n DL
Mi
DL
+ G Div DL + G Div

AT332_TRR_E0

Name
Value
TX i ic
TX i ic
CINR DL
C DL
TXj jc
IDL
-----------------10
10 Log
10
All TX j jc
Unit
TX ic
i
n DL
+ I Inter Tech + -------------------10
DL
10
TX i ic
Mi
TX jc
TX i ic
Downlink C/(I+N) for a cell
DL
+ G Div DL + G Div
With MIMO: CINR DL
I + N DL
+ NRInter Tech
DL
dB
Description
TX ic
I j
n DL
DL
------------------
---------------------
10
10
Inter
Tech
+ NRInter Tech
10
+I
10 Log
+ 10
DL
DL
All TX j jc
dBm
Unit
Description
dBm
dBm
Value
Unit
Description
W TXi ic
Used UL
n 0 + 10 Log ----------------------TX i ic
W Total
dBm
dBm
Unit
Description
dB
Unit
Description
dB

Name
Value
M
Mi
C UL

L
TX i
Mi
Mi
With P
Mi
Mi
L Ant L Body
P
EIRP UL
TX
Mi
+G
Mi
Mi
Mi
Mi
Mi
= P Eff after
power control

Name
TX i ic
n 0 UL
TX i ic
n UL
TX i ic
n 0 UL + nf
TX i ic

Name
Value
C UL n UL
TX i ic
Mi
i
CNR UL
Mi
TX i ic
UL
With MIMO: CNR UL + G Div UL + G Div

Name
Value
TX i ic
Mi
Mi
CINR UL
CNR UL NR UL
Mi
TX i ic
UL
847

Forsk <$year
12.2.10 Channel Throughput, Cell Capacity, and Per-user Throughput

Calculation
Name
Value
Unit
Description
kbps

throughput
kbps

throughput
kbps

throughput
kbps
kbps
kbps
kbps

user
kbps

per user
kbps

user
kbps

throughput
kbps

throughput
kbps

throughput
kbps
kbps
TX i ic
W Channel DL
M
CTP P DL
With MIMO (AMS):
Mi
Max
=
B
DL
M
i
B DL
Mi
DL
TX i ic
TX i ic
T AMS
if CNR DL
M
Mi
i
i
Mi
Mi
f TP Scaling
- TP Offset
CTP E DL -----------------------100
CTP E DL
CTP A DL
Mi
Cap P DL
Mi
Mi
TX i ic
Mi
TX i ic
CTP P DL TL DL Max N Channel DL

M
Mi
i
i
Cap P DL 1 BLER BDL
Mi
Mi
f TP Scaling
- TP Offset
Cap E DL -----------------------100
Cap E DL
Cap A DL
Mi
Mi
Mi
Cap P DL
----------------------TX i ic
N Users DL
Mi
PUTP P DL
PUTP E DL
Mi
PUTP A DL
Cap E DL
----------------------TX ic
i
N Users DL
Mi
Mi
i
PUTP E DL
M
f TP Scaling
i
- TPOffset
-----------------------100
TX i ic
W Channel UL
With MIMO (AMS):
Mi
B UL
Mi
Max
Mi
B UL
TX i ic
CTP P UL
Mi
B UL
if CNR DL
TX i ic
T AMS

TX ic
i
R UL
Mi
B UL
TX ic
--------------------------------- G MUi MIMO

D Frame
M
CTP E UL
Mi
CTP A UL
Mi
Mi
i
CTP E UL
M
f TP Scaling
i
- TP Offset
-----------------------100
TX i ic
Mi
TX i ic
Cap P UL
CTP P UL TL UL Max N Channel UL
i
i
Cap E UL
848
i
i

AT332_TRR_E0
Name
M
Value
M
i
Cap E UL
Cap A UL
kbps
kbps

user
kbps

user
kbps

user
Cap P UL
----------------------TX i ic
N Users UL
PUTP P UL
Description
M
f TP Scaling
i
------------------------- TP Offset
100
Mi
Unit
Mi
Cap E UL
----------------------TX i ic
N Users UL
Mi
PUTP E UL
Mi
Mi
Mi
f TP Scaling
100
Mi
PUTP A UL

TX ic
i
Downlink signal level C DL
Path loss L Path
,G

for which calculated path loss matrices are available.
depends on the signal level of the best server. The bar graph displays cells whose received signal levels are higher than their
C/N thresholds and are within a 30 dB margin from the highest signal level.
You can use a value other than 30 dB for the margin from the highest signal level, for example a smaller value for improving
the calculation speed. For more information on defining a different value for this margin, see the Administrator Manual.

Atoll displays the best server signal level and interference from other cells.
Interference level bar graphs show the interference levels in decreasing order. The maximum number of bars in the graph
depends on the highest interference level. The bar graph displays cells whose C/N are higher than the minimum interferer C/
N threshold and whose interference levels are within a 30 dB margin from the highest interference level.
You can use a value other than 30 dB for the margin from the highest interference level, for example a smaller value for
Manual.
849

Forsk <$year

12.3.2.1 Signal Level Coverage Predictions
The following coverage predictions are based on the received signal levels:
Overlapping Zones
For these calculations, Atoll calculates the received signal level, then determines the selected display parameter on each pixel
inside the cells calculation area. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver.
L
Mi
,G
Mi
Mi
Mi
, L Ant , and L Body are not considered in the calculations performed for the signal level based coverage predictions.
For more information on signal level calculations, see "Signal Level Calculation (DL)" on page 854


All Servers
TX i ic
TX ic
TX ic
i
or L i

TX i ic
TX ic
TX ic
i
or L i
AND
TX i ic
C DL
TX j jc
Best C DL M
ji
If M = 0 dB, Atoll considers pixels where the received signal level from TXi(ic) is the highest.
If M = 2 dB, Atoll considers pixels where the received signal level from TXi(ic) is either the highest or within a 2 dB
margin from the highest.
signal levels from the cells which are 2nd best servers.

TX i ic
AND
TX i ic
C DL
850
TX jc
nd
j
2 Best C DL
ji
TX ic
TX ic
i
or L i

AT332_TRR_E0
values.
If M = 0 dB, Atoll considers pixels where the received signal level from TXi(ic) is the second highest.
If M = 2 dB, Atoll considers pixels where the received signal level from TXi(ic) is either the second highest or within
a 2 dB margin from the second highest.
signal levels from the cells which are 3rd best servers.

such as:

level.
Path Loss (dB)
Total Losses (dB)
highest signal level) and evaluates the path loss from this cell.
highest signal level) and evaluates the total losses from this cell.
of these cells).

The following coverage predictions are based on the received signal levels and noise, and take into account the receiver
characteristics ( L
Mi
,G
Mi
Mi
Mi

For these calculations, Atoll calculates the received signal level or C/N level at each pixel. Each pixel within the calculation area
of TXi(ic) is considered a non-interfering receiver. The properties of the non-interfering probe receiver are set by selecting a
terminal, a mobility type, and a service.




page 864.
851

Forsk <$year
options:
Signal Level (DL) (dBm)

C/N Level (DL) (dB)
options:

C/N Level (UL) (dB)


Mi
,G
Mi
Mi
Mi
are based on the uplink noise rise values.

For more information on throughput calculations, see:
"Channel Throughput, Cell Capacity, and Per-user Throughput Calculation" on page 865.


page 864.
options:
C/(I+N) Level (DL) (dB)

Total Noise (I+N) (DL) (dBm)
options:
852
Bearer (DL)
Service

AT332_TRR_E0
options:

display options:
Quality indicators available in the document (Quality Indicators table): Atoll calculates the downlink C/(I+N) levels
quality indicator from the quality graphs defined in the LPWA equipment of the selected terminal.
options:

options:
Bearer (UL)
Service
options:

display options:
Quality indicators available in the document (Quality Indicators table): Atoll calculates the uplink C/(I+N) levels
quality indicator from the quality graphs defined in the LPWA equipment of the best serving cell.

station.
853

Forsk <$year

"Throughput Calculation" on page 865.

The following sections describe all the calculation algorithms used in point analysis, calculation of coverage predictions, and
calculations on subscriber lists.
12.4.1 Signal Level and Quality Calculations

The following sections describe how signal levels, noise and interference, C/N, and C/(I+N) ratios are calculated on the
downlink and uplink.


Input
TX i ic
P DL
TX i
TX i
: Transmission power of the cell TXi(ic).
: Transmitter antenna gain for the antenna used by the transmitter TXi. ( G
TX
TX i
TX i
TX i
= G Ant ).
= L Total DL ).
TX i
is selected.
Mi
Mi

Mi
Mi
Mi
854
Mi

AT332_TRR_E0
Calculations
TX ic
i
C DL
= EIRP
TX ic
i
L Ant L Body
EIRP
TX ic
i
TX ic
i
= P DL
+G
TX
TX
Output
TX i ic
C DL
: Received signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.

For determining the C/N and C/(I+N), Atoll calculates the downlink noise over the channel bandwidth used by the cell. The
used bandwidth depends on the number of used subcarriers. The downlink noise comprises thermal noise and the noise figure
Input
W Used DL : Downlink used bandwidth defined for the channel configuration of a cell TXi(ic).
W Total : Total bandwidth defined for the channel configuration of a cell TXi(ic).
nf
TX i ic
TX i ic
Mi
Calculations
n 0 = 10 Log K T 1000 = 174 dBm/Hz
TX i ic
n 0 DL
W TXi ic
Used DL
= n 0 + 10 Log ----------------------TX i ic
W Total
TX i ic
n DL
TX i ic
= n 0 DL + nf
Mi
Output
TX i ic
n DL

cells TXi(ic) and TXj(jc), and on the traffic loads of the interfering cells TXj(jc).
Input
TX j jc
C DL
: Received signal level from the cell TXi(ic) as explained in "Signal Level Calculation (DL)" on page 854.
855

Forsk <$year
information, see "Shadow Fading Model" on page 90). As the received signal levels from interferers already include
M Shadowing Model , M Shadowing C I is added to the signal levels from interferers in order to achieve the ratio
TX jc
j
TX jc
j
= C DL
C DL
+ M Shadowing C I
is selected.
TX j jc
TL DL
Calculations
Interference (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows:
TX jc
j
I DL
TX jc
j
= C DL
TX jc
j
Inter Tech
+ f TL DL + I DL
Calculation for the interference reduction factor due to traffic load is explained below:
Interference reduction due to interfering cells traffic load:
The interference reduction factor due to the interfering cells traffic load is calculated as follows:
TX jc
TX jc
j
j
f TL DL = 10 Log TL DL

TX
k
P DL Rec
--------------------------------------
=
F TX ic TX
i
k
TX k ICP DL
Inter Tech
I DL
TX k
F TX ic TX
i
k
ICP DL

TX k
power from LTE cells, preamble power from WiMAX cells, and downlink cell power from Wi-Fi or LPWA cells.
Output
TX j jc
I DL
I DL
: Downlink interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc).
Inter Tech

Input
TX i ic
C DL
856
TX i ic
n DL
T B : Bearer selection thresholds of the bearers defined in the LPWA equipment used by Mis terminal.
TX i ic
M

AT332_TRR_E0
or mobile Mi.
or mobile Mi.
TX i ic
Mi
mobile Mi.
i
BLER B DL : Downlink block error rate read from the graphs available in the LPWA equipment assigned to the

Calculations
The C/N for a cell TXi(ic) are calculated as follows for any pixel, subscriber, or mobile Mi:
TX i ic
CNR DL
TX i ic
= C DL
TX i ic
n DL
The bearers available for selection in the pixel, subscriber, or mobile Mis LPWA equipment are the ones:
Whose selection thresholds are less than the C/N at Mi: T B CNR DL
TX ic
i
Mi
If the cells channel configuration supports AMS, the STTD/MRC or SU-MIMO diversity gain, G Div DL ,
corresponding to the bearer is applied to its selection threshold. The gain is read from the properties of the LPWA
TX ic
i
i
i
equipment assigned to the pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX , Mobility M i , BLER B DL .
DL
applied. Therefore, the bearers available for selection are all the bearers defined in the LPWA equipment for which
Mi
Mi
TX i ic
DL
T B G Div DL G Div CNR DL
Once the bearer is known, the C/N calculated above become:
TX ic
i
TX ic
i
= CNR DL
CNR DL
DL
+ G Div DL + G Div
Mi
Output
TX i ic
CNR DL
: C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.

857

Forsk <$year
adjacent channel overlap between the studied and the interfering cells, and the traffic loads of the interfering cells. Finally,
Atoll takes the ratio of the signal level and the sum of the total interference from other cells and the downlink noise (as
calculated in "Noise Calculation (DL)" on page 855).
Input
TX i ic
C DL

TX i ic
n DL
I DL
TX j jc
: Interference from any cell TXj(jc) calculated for a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) as
explained in "Interference Calculation (DL)" on page 855.

Inter Tech
NRDL
T B : Bearer selection thresholds of the bearers defined in the LPWA equipment used by Mis terminal.
TX i ic
Mi
Mi
or mobile Mi.
Mi
or mobile Mi.
TX i ic
mobile Mi.
i
BLER BDL : Downlink block error rate read from the graphs available in the LPWA equipment assigned to the


Inter Tech
I DL
Calculations
The downlink C/(I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX i ic
CINR DL
TX i ic
= C DL
TXj jc
TX i ic
n DL
IDL
Inter
Tech
Inter
Tech
------------------
10 + I DL
+ -------------------+ NR DL
10 Log
10
10
10
All TXj jc
The Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX ic
TX i ic
I + N DL
TX j jc
n DL
I DL
---------------------
Inter Tech
Inter Tech
10
-----------------= 10 Log
+ 10
+ NR DL
10 + I DL
10
All TX j jc
The bearers available for selection in the pixel, subscriber, or mobile Mis LPWA equipment are the ones:
858

AT332_TRR_E0
Whose selection thresholds are less than the downlink C/(I+N) at Mi: T B CINR DL
TX i ic
Mi
Mi
If the cells channel configuration supports AMS, the STTD/MRC or SU-MIMO diversity gain, G Div DL ,
TX ic
i
i
i
equipment assigned to the pixel, subscriber, or mobile Mi for N Ant TX , N Ant RX , Mobility M i , BLER B DL .

DL
Mi
Mi
TX i ic
DL
T B G Div DL G Div CINR DL
TX i ic
CINR DL
TX i ic
= CINR DL
Mi
DL
+ G Div DL + G Div
Output
TX i ic
CINR DL
: Downlink C/(I+N) from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i ic
I + N DL
: Total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by a cell TXi(ic).
Mi

Input
Mi
control.
Mi
TX i
TX i
: Transmitter antenna gain for the antenna used by the transmitter TXi ( G
TX i
TX i
TX i
= G Ant ).
= L Total UL ).
TX i
TX i
is selected.
859

Forsk <$year

i
Mi
Mi
Mi
Calculations
Mi
Mi
TX i
TX i
Mi
Mi
L Ant L Body
Mi
EIRP UL = P
With P
Mi
Mi
+G
Mi
Mi
Mi
Mi
Mi
Output
Mi

used bandwidth depends on the number of used subcarriers. The uplink noise comprises thermal noise and the noise figure
Input
W Used UL : Uplink used bandwidth defined for the channel configuration of a cell TXi(ic).
W Total : Total bandwidth defined for the channel configuration of a cell TXi(ic).
nf
TX i ic
TX i ic
TX i ic
Calculations
n 0 = 10 Log K T 1000 = 174 dBm/Hz
TX i ic
n 0 UL
W TXi ic
Used UL
= n 0 + 10 Log ----------------------TX i ic
W Total
TX i ic
n UL
860
TX i ic
= n 0 UL + nf
TX i ic

AT332_TRR_E0
Output
TX ic
i
n UL

Input
Mi
TX i ic
n UL
M PC : Power control margin defined in the global network settings (1 dB).
T B : Bearer selection thresholds of the bearers defined in the LPWA equipment used by the cell TXi(ic).
TX i ic
M
Mi
Mi
Mi
or mobile Mi.
Mi
or mobile Mi.
Mi
mobile Mi.
TX i ic
i
BLER B UL : Uplink block error rate read from the graphs available in the LPWA equipment assigned to the cell TXi(ic).

Calculations
Mi
Mi
TX i ic
CNR UL = C UL n UL
The bearers available for selection in the cell TXi(ic)s LPWA equipment are the ones:
Mi
Mi
TX ic
i
M
TX ic
i
i
i

UL
861

M
TX ic
i
Forsk <$year
M
UL

Mi
Mi
TX i ic
UL

TX i ic
bearer.
If with P
Mi
Mi
Mi
TX i ic
Mi
+ M PC , where T
B UL
TX i ic
Mi
B UL
is the bearer selection threshold, from the LPWA
M
M
TX ic
Mi
Mi
i
i
i
P Eff = Max PMax CNR UL T M + M PC P Min
i
UL
Mi
Mi

Output
Mi

uplink C/N by the uplink noise rise value of the cell.
Input
Mi
CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic) as calculated in "C/N Calculation (UL)"
on page 861.
862
TX i ic
NRUL
T B : Bearer selection thresholds of the bearers defined in the LPWA equipment used bythe cell TXi(ic).
TX i ic
Mi
Mi
Mi

AT332_TRR_E0
M
or mobile Mi.
M
or mobile Mi.
Mi
mobile Mi.
TX i ic
i
BLER B UL : Uplink block error rate read from the graphs available in the LPWA equipment assigned to the cell TXi(ic).
Calculations
Mi
TX i ic
Mi
CINR UL = CNR UL NRUL
The bearers available for selection in the cell TXi(ic)s LPWA equipment are the ones:
Whose selection thresholds are less than the uplink C/(I+N) at Mi: T B CINR UL
Mi
Mi
TX i ic
M
TX ic
i
i
i

UL
Mi
TX i ic
UL
Mi

Mi
Mi
TX i ic
UL

TX i ic
Where G Div UL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer.
Mi
Mi
TX i ic
UL

TX i ic
bearer.
863

Forsk <$year
If with P
TX ic
i
M
i
B
UL
+ M PC , where T
TX ic
i
M
i
B
UL
is the bearer selection threshold, from the LPWA
Mi
M i TX i ic
Mi
Mi
P Eff = Max PMax CINR UL T M + M PC P Min
i
UL
Mi
Mi

Output
Mi
CINR UL : Uplink C/(I+N) from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
Mi
M

In LPWA, best server refers to a cell ("serving transmitter"-"reference cell" pair) from which a pixel, subscriber, or mobile Mi
gets the highest signal level. This calculation also determines whether the pixel, subscriber, or mobile Mi is within the coverage
area of any transmitter or not.
Input
TX i ic
C DL
: Downlink signal level received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Signal
Level Calculation (DL)" on page 854 using the terminal and service parameters ( L
Mi
,G
Mi
Mi
Mi
, L Ant , and L Body ) of Mi.
Calculations
The best server of any pixel, subscriber, or mobile Mi, BS M , is the cell from which the received downlink signal level is the
i
highest among all the cells. The best server is determined as follows:
BS M = TX i ic
i
TX ic
TX i ic
i
= Best
C DL
C
All TX i ic DL
Here ic is the cell of the transmitter TXi with the highest power. However, if more than one cell of the same transmitter covers
the pixel, subscriber, or mobile, the final reference cell ic might be different from the initial cell ic (the one with the highest
power). In coverage prediction calculations and in calculations on subsriber lists, the cell of the highest priority layer is
selected as the serving (reference) cell.
Output

i

In LPWA, a pixel, subscriber, or mobile Mi can be covered by a cell (as calculated in "Best Server Determination" on page 864)
TXi(ic) if the downlink C/N from the cell at the pixel, subscriber, or mobile is greater than or equal to the minimum C/N
864

AT332_TRR_E0
Input
TX ic
i
CNR DL
: Downlink C/N from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "C/N Calculation (DL)"
on page 856.
TX ic
i
T Min
: Min C/N threshold defined for the cell TXi(ic).
Calculations
TX ic
i
CNR DL
TX ic
i
T Min
Output

False: Otherwise.

Calculation of throughputs is explained in "Channel Throughput, Cell Capacity, and Per-user Throughput Calculation" on
page 865.
12.4.4.1 Channel Throughput, Cell Capacity, and Per-user Throughput Calculation

Channel throughputs are calculated for the width of one channel. Cell capacities are calculated by multiplying the channel
throughputs by the numbers of downlink and uplink channels per cell and are upper-bound by the maximum downlink and
uplink traffic loads. Per-user throughputs are calculated by dividing the cell capacities by the average number of connected
users, downlink or uplink, defined for the cell.
Input
TX i ic
W Channel DL : Width of one downlink channel of the cell TXi(ic).
W Channel UL : Width of one uplink channel of the cell TXi(ic).
N Channel DL : Number of downlink channels of the cell TXi(ic).
N Channel UL : Number of uplink channels of the cell TXi(ic).

M : Bearer efficiency (bps/Hz) of the bearer assigned to the pixel, subscriber, or mobile Mi in the uplink in "C/(I+N)
TX i ic
TX ic
i
TX i ic
TX i ic
TX i ic
Mi
B DL
: Bearer efficiency (bps/Hz) of the bearer assigned to the pixel, subscriber, or mobile Mi in the downlink in "C/
i
B UL
and Bearer Calculation (UL)" on page 862.

TX i ic
CNR DL
: Downlink C/N the cell TXi(ic) as calculated in "C/N Calculation (DL)" on page 856.
i
i
BLER B DL : Downlink block error rate read from the BLER vs. CINR DL
TX i ic
TX i ic
TX i ic
TX ic
graph available in the LPWA equipment
865

Forsk <$year
i
i
BLER BUL : Uplink block error rate read from the BLER vs. CINR UL graph available in the LPWA equipment assigned

M
Mi.
Mi
TX i ic
TX i ic
Calculations
Downlink:
TX i ic
Mi
Peak MAC Channel Throughput: CTP P DL = W Channel DL
Mi
B DL
MIMO SU-MIMO Gain:

Max
If the channel configuration supports AMS, SU-MIMO gain G SU MIMO is applied to the bearer efficiency. The gain is
read from the properties of the LPWA equipment assigned to the pixel, subscriber, or mobile Mi for:
TX ic
i
Mi
or mobile Mi.
B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink as explained in "C/(I+N) and Bearer
Mi
i
BLER B DL : Downlink block error rate read from the graphs available in the LPWA equipment assigned to the

TX i ic
terminal used by the pixel, subscriber, or mobile Mi. BLER is determined for CINR DL
Mi
B DL
TX i ic
Max
Mi
TX i ic
T AMS
B DL
M
i
i
i
Mi
Mi
Mi
f TP Scaling
- TP Offset
Application Channel Throughput: CTP A DL = CTP E DL -----------------------100
Peak MAC Cell Capacity: Cap P DL = CTP P DL TL DL Max N Channel DL
i
i
i
Mi
TX i ic
Mi
866
TX i ic
M
Mi
Mi
f TP Scaling
- TP Offset
= Cap E DL -----------------------100
Mi
Mi
Mi
Cap P DL
= ----------------------TX i ic
N Users DL

AT332_TRR_E0
Effective MAC Throughput per User:
Cap E DL
= ----------------------TX ic
i
N Users DL
i
PUTP E DL
i
PUTP A DL
i
PUTP E DL
M
f TP Scaling
i
- TP Offset
-----------------------100
Uplink:
TX i ic
Mi
Peak MAC Channel Throughput: CTP P UL = W Channel UL

B
Mi
UL
MIMO SU-MIMO Gain:

Max
If the channel configuration supports AMS, SU-MIMO gain G SU MIMO is applied to the bearer efficiency. The gain is
read from the properties of the LPWA equipment assigned to the cell TXi(ic) for:
Mi
or mobile Mi.
TX i ic
B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink as explained in "C/(I+N) and Bearer
Mi
i
BLER B UL : Uplink block error rate read from the graphs available in the LPWA equipment assigned to the cell

Mi

Mi
B UL
TX i ic
Max
Mi
TX i ic
T AMS
B UL
TX i ic
CNR DL
TX i ic
TX i ic
TX i ic
T MU MIMO and N Ant RX 2 , the MU-MIMO gain G MU MIMO is applied to the channel throughput. The
MU-MIMO gain is read from the properties of the cell TXi(ic).

Mi
Mi
TX i ic
i
i
i
i
CTP A UL
i
CTP E UL
Mi
M
f TP Scaling
i
- TP Offset
-----------------------100
Peak MAC Cell Capacity: Cap P UL = CTP P UL TL UL Max N Channel UL
i
i
i
Mi
Application Cell Capacity:
i
Cap A UL
Peak MAC Throughput per User:
TX i ic
Mi
i
PUTP P UL
Mi
i
Cap E UL
M
TX i ic
M
f TP Scaling
i
- TP Offset
-----------------------100
Mi
Cap P UL
= ----------------------TX i ic
N Users UL
867

Forsk <$year
Effective MAC Throughput per User:
i
PUTP E UL
Cap E UL
= ----------------------TX ic
i
N Users UL
M
i
PUTP A UL
i
PUTP E UL
M
f TP Scaling
i
- TPOffset
-----------------------100
Output
Mi
Mi
M
Mi
Mi
Mi
Mi
M
Mi
Mi
Mi
Mi
M
Mi
Mi
Mi
Mi
M



868
They are active,


AT332_TRR_E0

zone.
discarded.

[Neighbours]

869

Forsk <$year
decreasing rank.
neighbour.
i.
[Neighbours]
i

i
i
The received signal level is greater than or equal to the signal level threshold. The received signal level
TX i ic
( C DL
TX i ic
) and the signal level threshold are calculated from CNR DL

TX i ic
value of the noise ( n DL
, respectively, by adding the
) to them.
TX ic
i
S TX ic is the surface area covered by TXi(ic) within C DL

i
TX i ic
and T Min
TX ic
i
+ HO Start and C DL
+ HO End . HOStart is the
margin with respect to the best signal level at which the handover starts, and HO End is the margin with
respect to the best signal level at which the handover ends.

j
870
TX i ic
If a global value of the C/N threshold ( T Min
) is set in the coverage conditions dialog
box, for each cell, Atoll uses the higher of the two values, i.e., global value and the
value defined for that cell.
gives the largest possible C/N coverage areas for the cells.
Atoll

AT332_TRR_E0
S TX ic S TX jc
i
j
- 100 ),
i

S TX ic S TX jc
i
j
TXj(jc) is considered a neighbour of TXi(ic) if --------------------------------------- 100 % Min Coverage Area .
S TX ic
i
Neighbourhood cause
When
Importance value
Existing neighbour

Existing importance
Exceptional pair
100 %
Co-site cell

selected
Adjacent cell

selected

relationship
transmitter.
d Di = 1 ---------d max
of antennas.
871

Forsk <$year
[Neighbours]

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
30%
Min(A)
30%
Max(A)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

Neighbourhood cause
Importance Function

No
10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
No
Yes
30%+30%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Yes
Yes
60%+40%{10%(Di)+30%(O)
+60%(A)}+1%+9%(Di)
Co-site
Adjacent
No
Where:
overlapping.
causes.

872

AT332_TRR_E0
They are active,


zone.
conditions:

[Neighbours]
CDMA Carriers: This option is available when a LPWA network is being co-planned with a UMTS, CDMA, or TDSCDMA network. This option enables you to select the CDMA carrier(s) that you want Atoll to consider as potential
neighbours of LPWA cells. You may choose one or more carriers. Atoll will allocate only the cells using the selected

neighbour.
873

Forsk <$year
This means that the signal received from A is greater than the minimum required (calculated from the C/N
threshold), and is the highest one.
The signal level received from A exceeds the minimum required (calculated from the C/N threshold) and is
within a margin from the highest signal level.
to 0 dB.
2nd case: The margin is other than 0 dB. SB is the area where:
level.
SA SB
- 100 ) and compares this value with the %
Atoll calculates the percentage of the coverage area overlap ( ----------------SA
SA SB

Neighbour cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
100 %

distance conditions
d1 ---------d max
distance.

Neighbour cause
When
Importance value
Existing neighbour
Existing importance
Exceptional pair
100 %
IF

IF
transmitter.
d
Di = 1 ----------d max
874

AT332_TRR_E0

Factor
Min importance
Default value
Max importance
Default value
Min(Di)
1%
Max(Di)
10%
Min(O)
10%
Max(O)
60%
Co-site factor (C)
Min(C)
60%
Max(C)
100%

cause
IF

No
10%+50%{10%(Di)+90%(O)}
Yes
60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.
875

876
Forsk <$year
Chapter 13
ACP Module
"Objectives" on page 879
"Quality Predictions and Antenna Masking Methods" on

page 884
"Configuration" on page 886
"Multi-RAT and Co-planning Support" on page 888
"Optimisation Methodology" on page 889
"Load Balancing Objective" on page 895
"EMF Exposure" on page 904
"Shadowing Margin and Indoor Coverage" on page 907
"Multi-Storey Optimisation" on page 907
"ACP Software Data Flow" on page 910

878

Chapter 13: ACP Module
AT332_TRR_E0
13 ACP Module
The ACP module optimises key network parameters to improve coverage and quality. It can also select the best sites from a
list of candidate sites.
ACP uses user-defined objectives to evaluate the quality and implementation cost of a network reconfiguration.
It uses an efficient global search algorithm to test many network configurations and propose the reconfigurations which best
meet the objectives. ACP presents the changes ordered from the most to the least beneficial, allowing phased implementation
or implementation of just a subset of the suggested changes.
Currently, ACP supports the following single-RAT radio access technologies: GSM, UMTS, CDMA2000, LTE, WiMAX, and Wi-Fi.
ACP also supports 3GPP and 3GPP2 multi-RAT documents as well as co-planning.
13.1 Objectives
Atoll ACP uses user-defined objectives to evaluate the quality and cost of the network reconfiguration or site selection. In
addition, a cost objective can be taken into account to reduce the expected implementation cost.
13.1.1 Quality Objective

Each quality objective is a logical combination of defined rules used to evaluate specific quality indicators, which are evaluated
in a given zone and for a given traffic profile. An objective can combine several quality indicators from different technology
layers. Each quality indicator is technology-dependent, and is consistent with the corresponding coverage predictions in Atoll.
13.1.1.1 Definition and Evaluation

ACP calculates the quality objective using the user-defined resolution within the borders of the computation zone.
It calculates the basic quality indicators (RSCP, EcIo, CINR, overlap, etc.) on each pixel of the computation zone. Quality maps
covering the computation zone are provided for the initial network (before reconfiguration) and final network (after
reconfiguration).
Each objective is measured on a defined target zone. The target zone can be either the computation zone, the focus zone, a
hot spot, or a zone defined as a group of clutter classes. The objective is calculated only on the subset of pixels belonging to
this zone.
An objective can also be weighted according to traffic or weighted on a given zone. The defined weight enables you to assign
a different importance to different pixels. When using traffic weighting, the pixel weights are taken from Atoll traffic maps.
When using zone weighting, the pixel weights are taken from a weight defined with each zone. Both types of weighting can
be used at the same time, in which case the zone weight is taken as a supplementary factor to the traffic weight. For more
information on how weights are applied, see "Optimisation Methodology" on page 889.
An objective is defined by both a set of rules and a target. A pixel is said to be "covered" by the set of rules when it fulfils all
the rules according to their logical relationship (OR, AND). A rule is a single quality indicator on a single technology layer
fulfilling a defined threshold.
An example of combined rules is:
(UMTS 2100 - RSCP > -85dBm OR LTE 2010 - C/N > 20dB)
The target for the objective defines the required percentage of pixels in the target zone (after applying any defined weight)
which must fulfil the rule. For example if the target is 90%, the objective is said to be fulfilled if 90% of the pixels are covered
by the objective rule.
This is described by the following formula:
Cov Obj =
i 1 Th1 Qual 1 i OR 1 Th2 Qual 2 i
i pixels of zone
where,
1 Th is the step function: 1 Th x = 1 if x Th and 1 Th x = 0 if x Th
Qual k i is the basic quality measurement on pixel i
i is the normalised traffic density on pixel "i" such as Max i = 1
In case of uniform weighting, all i are equal to 1.
In case of non-uniform weights, weights vary between 0 and 1 with Max i = 1 .
879

Thresholds on rules can be defined separately per zone or per clutter groups.
The target threshold can be defined as absolute or relative compared to initial status:
for example as 90% (absolute) or as increase of 5% over current coverage (relative).
13.1.1.2 Progressive Thresholds

Progressive thresholds allow ACP to evaluate the amount of improvement or degradation of each objective, leading to more
intelligent decisions on improvements that may cause degradations elsewhere in the network.
Progressive thresholds are proposed by default for some quality parameters. When this feature is supported for a quality
parameter, the Progressive Threshold check box is enabled in the Thresholds Definition dialog box. You can disable it by
setting the useProgressiveThreshold option to 0 in the [ACPCore] section of the ACP.ini file.
Figure 13.1: Thresholds Definition dialog box
When the Progressive Threshold check box is cleared, a step function (1/0) is used and ACP objectives are evaluated
on a "fixed-threshold basis", on each pixel and in a logical manner, whether the objectives are met or not. The step
function is described in "Definition and Evaluation" on page 879 and it is used by default in all objectives.
When the Progressive Threshold check box is selected, a progressive function is used instead with a weighting varying
from 1 to 0, from a maximum to a minimum. Below are the main curve parameters used by the progressive function:
The options below can be added in the ACP.ini file in order to modify maximum and minimum threshold values
(see full option names and default values in the Administrator Manual):
[ACPCore]
<Technology>.<Objective>.th.min
<Technology>.<Objective>.th.max
Threshold defined in the objectives properties on the Objectives tab of the ACP setup. This value is used as the
transition between two signal level ranges (below and above threshold), each range having its own hard coded
modelling.
Main curve parameters can produce several shapes according to the user-defined values. Below is a typical example
with the signal level type objective (RSCP) and the default Min/Threshold/Max values (-120dBm/-90dBm/-60dBm).
Figure 13.2: Progressive Thresholds function
13.1.1.3 Target Filtering

ACP allows you to filter pixels on which the target percentage will be evaluated according to defined filter conditions. When
using a filter, the target percentage coverage is not evaluated on all pixels of the target zone, but only on pixels of the target
zone which are not filtered out.
For example, you could calculate an objective only on the pixels of a zone for which there is no coverage in a given technology:
880

AT332_TRR_E0
Target: 90% of pixel with UMTS RSCP > -95dB for which GSM Signal Level < -95dBm
The 90% target will be applied only to the subset of pixels for which the GSM signal level is below -90dBm
13.1.2 ACP Quality Indicators

ACP defines a set of basic quality indicators. These quality indicators are used when defining a rule to form complex objectives.
For each quality indicator, ACP uses the same formulas as used elsewhere in Atoll. This ensures that the measured values are
the same in ACP and Atoll predictions.
13.1.2.1 GSM Quality Indicators

The quality indicators used by ACP give the same maps and results as the reference prediction in Atoll. In the following list,
each quality indicator defined in GSM is followed in italics by the reference prediction in Atoll (with the relevant "Field"
setting), if any:
BCCH Signal Level (Coverage by Signal Level (DL) + "Best Signal Level (dBm)") used as a measure of raw network
coverage.
CINR Co-channel (Coverage by C/I Level (DL) + "C/I Level (dB)")
Overlap (Overlapping Zones (DL) + "Number of Servers") to define cell dominance and decrease the level of
interference between cells while allowing a level of cell overlap.
Best Server Distance
1st-2nd Difference
1st-Nth Difference
ACP manages interference quality in the network by measuring signal pollution: a limited number of overlapping cells are
allowed in order to allow for coverage continuity and handover capability. The number should be consistent with the
frequency reuse ratio used for the network.
13.1.2.2 UMTS Quality Indicators

each quality indicator defined in UMTS is followed in italics by the reference prediction in Atoll (with the relevant "Field"
setting), if any:
RSCP (Coverage by Signal Level (DL) + "Best Signal Level (dBm)") used as a measure of raw network coverage.
EcIo (Pilot Quality Analysis (DL) with "Ec/Io (dB)") to define the service area zone.
RSSI (Total Noise Level Analysis (DL) + "Max Noise Level (dBm)")
Overlap (Overlapping Zones (DL) + "Number of Servers") to measure pilot pollution as well as soft handover quality.
1st-2nd Difference
1st-Nth Difference
13.1.2.3 CDMA2000 Quality Indicators

each quality indicator defined in CDMA2000 is followed in italics by the reference prediction in Atoll (with the relevant "Field"
setting), if any:
Signal Level (Coverage by Signal Level (DL) + "Best Signal Level (dBm)") used as a measure of raw network coverage.
EcIo (Pilot Quality Analysis (DL) with "Ec/Io (dB)") to define the service area zone.
Overlap (Overlapping Zones (DL) + "Number of Servers") to measure pilot pollution as well as soft handover quality.
1st-2nd Difference
1st-Nth Difference
ACP handles CDMA2000 similarly to UMTS. The main difference is that the formula for deriving signal level and EcIo differs
between 1xRTT and 1xEv-DO:
In 1xRTT, formulas are similar to the ones in UMTS, taking into account the pilot power as the basis for signal level
computation.
In 1xEv-DO, the pilot is transmitted at full cell power. The cell max power is thus used as the basis of the signal level
computation, as well as the Ec/Io computation.
13.1.2.4 LTE Quality Indicators

each quality indicator defined in LTE is followed in italics by the reference prediction in Atoll (with the relevant "Field" setting),
if any:
881

RS C (Effective Signal Analysis (DL) + "RS Signal Level (DL) (dBm)") used as a measure of raw network coverage.
RS C/N (Effective Signal Analysis (DL) + "RS C/N Level (DL) (dB)") used as a measure of raw network coverage.
RSRP (Effective Signal Analysis (DL) + "RSRP Level (DL) (dBm)") used as a measure of raw network coverage.
RS CINR (Coverage by C/(I+N) Level (DL) + "RS C/(I+N) Level (DL) (dB)") to measure and control interference.
RSRQ (Coverage by C/(I+N) Level (DL) + "RSRQ Level (DL) (dB)") to measure and control interference.
RSSI (Coverage by C/(I+N) Level (DL) + "RSSI Level (DL) (dBm)") to measure and control interference.
PDSCH CINR (Coverage by C/(I+N) Level (DL) + "PDSCH C/(I+N) Level (DL) (dB)") to measure and control interference.
RLC Peak Rate (Coverage by Throughput (DL) + "Peak RLC Channel Throughput (DL) (kbps)") to measure and control
rate.
Overlap (Overlapping Zones (DL) + "Number of Servers") to better control cell dominance.
1st-2nd Difference
1st-Nth Difference
Because RS CINR, RSRQ, RSSI, PDSCH CINR, and RLC Peak Rate depend strongly on the frequency plan, two methods are
currently provided by ACP:
Using the current frequency plan: The existing frequency plan is taken into account when calculating RS CINR, RSRQ,
RSSI, PDSCH CINR, and RLC Peak Rate. Currently the frequency plan is not dynamically recalculated while changing
network parameters. In some cases this may lead to suboptimal reconfiguration, in which case it is recommended to
perform one or several ACP -> AFP cycles.
Ignoring the current frequency plan: All the network cells are assumed to be on the same channel.
ACP Calculation Method for CINR

The ACP calculation method for CINR (RS CINR and PDSCH CINR) applies to interferences on physical channels (RS, PDSCH, or
PDCCH).
It is recommended to use the following settings:
In the Atoll.ini initialisation file:
If SameItf_PDSCH_RS_PDCCH is set to 0 (default value), the Consider frequency plan option must be selected in
objective parameters of the ACP setup.
If SameItf_PDSCH_RS_PDCCH is set to 1, the Ignoring frequency plan option must be selected in objective
parameters of the ACP setup.
For more details about the SameItf_PDSCH_RS_PDCCH option, see the Administrator Manual.
The SameItf_PDSCH_RS_PDCCH must not be defined in the ACP.ini initialisation file.

Use the same configuration for RS CINR and PDSCH CINR objective parameters.
Calculation method
option in ACP
objective parameters
(RS CINR and PDSCH
CINR)
PCI Allocation
considered in
the ACP
interference
algorithm
= 0 (Method 1)
Synchronised transmission/
reception (default or option
not defined))
Consider frequency
plan
Yes
= 1 (Method 2)
Unsynchronised
transmission/reception
(former Atoll algorithm)
Ignoring frequency
plan
No
Atoll.ini file
[LTE]
SameItf_PDSCH_RS_PDCCH
ICIC
Yes, if
activated in
cells
13.1.2.5 WiMAX Quality Indicators

each quality indicator defined in WiMAX is followed in italics by the reference prediction in Atoll (with the relevant "Field"
setting), if any:
882
Preamble C (Effective Signal Analysis (DL) + "Preamble Signal Level (DL) (dBm)") used as a measure of raw network
coverage.
Preamble C/N (Effective Signal Analysis (DL) + "Preamble C/N Level (DL) (dB)") used as a measure of raw network
coverage.
Preamble CINR (Effective Signal Analysis (DL) + "Preamble C/(I+N) Level (DL) (dB)") to measure and control
interference and signal quality.

AT332_TRR_E0
Overlap (Overlapping Zones (DL) + "Number of Servers") to better control cell dominance.
1st-2nd Difference
1st-Nth Difference
Because Preamble CINR depends strongly on the frequency plan, two methods are currently provided by ACP:
Using the current frequency plan: The existing frequency and segmentation plan are taken into account when
calculating the CINR. Currently the frequency plan and segmentation plans are not dynamically recalculated while
changing network parameters. In some cases this may lead to suboptimal reconfiguration, in which case it is
recommended to perform one or several ACP -> AFP cycles.
Ignoring the current frequency plan and segmentation: All the network cells are assumed to be on the same channel.
13.1.2.6 Quality Indicator Parameters and Reference Maps

The parameters that define how each quality indicator is calculated are under "Parameters" on the Objectives tab, for
example which service and terminal to use to define body loss and other losses (terminal antenna gain and loss). Additionally,
you can consider shadowing in the calculation.
For most quality indicators, you can specify a reference prediction from among the predictions already calculated. By using a
reference prediction, you can ensure that the quality indicator will be calculated the same as the reference prediction,
enabling comparison of the quality map with the Atoll coverage prediction.
13.1.2.7 Advanced Objective Configuration

By combining several rules to define one objective, you can define more advanced objectives. For example:
Example
UMTS Overlap > 0
AND UMTS Overlap < 4
(UMTS RSCP > -90dBm AND UMTS EcIo > -12dB)
(LTE RS C > -85dBm AND LTE RS CINR > 4dB)
OR
Description
Pilot Pollution avoidance (UMTS)
Coverage offered by at least one technology
By defining a filter, you can even more advanced objectives by applying the rules only to certain pixels. For example:
Example
GSM BCCH > -90dBm
FOR Pixels where:
(UMTS RSCP < -100 AND Overlap < 2
Description
Possibility of Inter-technology handover
UMTS->GSM
13.1.2.8 Cost Objective

ACP also takes cost objectives into consideration. There are two modes of operation:
Cost limit: The total cost of the reconfiguration will not exceed a given maximum cost.
Trade-off between quality and cost: ACP will select the changes which have the most benefit for the least cost.
You can also assign different sets of costs for different site classes. Each site class can be assigned a different set of costs. You
can automatically create a set of site classes and their associated costs by defining some options in the [ACPGeneralPage]
section of the ACP.ini file. For more information, see the Administrator Manual.
Sites are assigned to a site class either manually or automatically. You can assign them automatically by defining a custom
field in the Sites table in Atoll and then defining the custom field in the ACP.ini using the "site.costClass" option.
[ACPCustomFieldExtraction]
site.costClass=SITECLASS
# The name of the custom column in SITE table used to define the 'cost class'.
# 'cost class' is used to define precisely the cost of changes applied to a site.
The site class defined in the custom field in the Sites table will be assigned automatically to each site in the database when a
new ACP setup is created. If a new candidate site is created in ACP and is co-located with an existing site, it will inherit the site
class of the existing site. If it is not co-located with an existing site, the site class is set to Default and can be changed manually.
13.1.3 Atoll and ACP Predictions Matching

ACP coverage predictions try as much as possible to match the Atoll coverage predictions (e.g. ACPs "EcIo" prediction versus
Atolls "Pilot Quality Analysis (DL)" prediction). Coverage predictions are therefore similar in most cases, in spite of the variety
of potentially conflictuous conditions such as varying resolutions, etc.
883

Generally speaking, ACP and Atoll coverage predictions will match except in corner cases which are difficult to identify and
manage. When there are differences, they will be at pixel level and are negligible (e.g. small map shifts, etc). At the scale of
overall maps, they still match pretty well despite small cosmetic mismatches in some very specific corner cases.
Hence, potential mismatches between ACP and Atoll core predictions may appear according the following parameter settings:
Resolution:
The best match between ACP and Atoll coverage predictions is obtained when the ACP resolution matches the
path loss resolution. When the resolution of the optimisation is different from the resolution of the path loss
matrices, ACP performs a bilinear interpolation by using the four closest path loss values and interpolating.
The most acute match between ACP and Atoll coverage predictions is obtained when the ACP resolution, the path
loss resolution, the Atoll prediction resolution, and the raster resolution are identical.
Setup Preferences > Calculation setting (on the Preferences tab of the ACP Properties dialog box):
In "High speed" mode, ACP reduces the number of cells it monitors for each pixel, some of which may only create
a bit of interference at first, and later create significantly more interference after antenna parameters are changed
during optimisation.
The "Automatic Candidate Positioning" functionality (New Candidate Setup dialog box >
Action button) can be impacted in "High speed" mode.
In "High precision" mode, ACP increases the number of cells it monitors for each pixel, thereby reducing potential
inconsistencies with Atoll coverage predictions (for more information, refer to the "Configuring Default Settings"
section in the User Manual).
When the Multi-Storey extension is enabled, the coverage predictions calculated by ACP may differ slightly from
Multi-Storey predictions due to different methodologies used by ACP and the Atoll platforms. ACP uses a mix
combining a radial method for lower storeys (based on Atoll's "CalculateGrid" API) and a systematic method for upper
storeys where few evaluation points are present (using Atoll's "CalculateSubscribers" API).
13.2 Quality Predictions and Antenna Masking Methods

ACP needs to correctly assess how well a reconfigured network will meet quality objectives when performing an antenna
reconfiguration such as changing the antenna model, tilt, or azimuth. ACP assesses this change by calculating how the path
loss matrices change when the antenna is modified. This process is strongly dependent on the type of propagation model used
originally to produce the path loss matrices.
ACP distinguishes between two categories of propagation models: native and non-native. For native propagation models, ACP
selects the "Optimised" antenna masking method by default. For non-native propagation models, ACP proposes four different
antenna masking methods: "Basic", "Improved", "Full Path Loss", and "Antenna Correction".
The antenna masking method is not used for site selection and antenna height
optimisation. These types of reconfiguration are performed by direct path loss
calculation. In addition, reconfiguring power is performed by direct scaling of existing
path loss matrices and therefore does not use either an antenna masking method or
recalculation of the path loss matrices.
13.2.1 Optimised Masking Method

The "Optimised" antenna masking method is used for propagation models which are native to Atoll: the Standard Propagation
Model (SPM), Cost Hata, CrossWave, etc. ACP performs an unmasking operation with the current antenna pattern, followed
by remasking with the new antenna pattern. The "Optimised" antenna masking method ensures that the ACP prediction
correlates strongly with the propagation model calculation.
This calculation depends strongly on the horizontal and vertical emission angles between a transmitter and the receiving pixel.
The "Optimised" antenna masking method provides accurate prediction of the emission angles, using one of two internal
antenna masking methods:
Direct calculation: ACP calculates incidence angles by direct calculation using the raster data.
Delegating to the model: ACP calculates incidence angles by delegating the calculation to the propagation model,
providing that the propagation model implements the appropriate antenna masking methods of Atolls API.
ACP automatically selects which internal antenna masking method to use for each native propagation model:
884
CrossWave: ACP delegates the calculation to model the propagation model.

All others native models: ACP calculates directly.

AT332_TRR_E0
You can define the internal antenna masking method used by setting the appropriate
option in the ACP.ini file. For information on modifying the ACP.ini file, see the
13.2.2 Antenna Masking Methods for Non-Native Models

For non-native propagation models, ACP proposes four antenna masking methods:
"Basic Masking Method" on page 885

"Improved Masking Method" on page 885
"Full Path Loss Masking Method" on page 885
"Antenna Correction Masking Method" on page 886
13.2.2.1 Basic Masking Method

The "Basic" antenna masking method is similar to the "Optimised" method with direct calculation, but with a few additional
parameters. You can set the following parameters for the default antenna masking method:
Parameter
Description
Antenna pattern interpolation The antenna gain calculation method for deriving the antenna gain from a set of angles of
incidence. Either:
Native 3D interpolation method: The method used by Atoll. For more information
on Atolls method for 3D interpolation, see the Technical Guide
Linear interpolation method: A simple linear method with a smoothing parameter:
G = A hor azi + A ver elev smooth
Direct view
When selected, the angle of incidence will be the direct Tx-Rx angle
Use clutter height
Specify whether clutter heights should be applied along the profiles between transmitter
and receiver. Clutter heights are either extracted from the clutter height file, or from
default clutter heights based on the clutter class file.
Receiver on top of clutter
Specify whether the receiver should be considered to be on top of the clutter or not.
Operator-specific propagation models can often be modelled correctly using the "Basic" antenna masking method.
13.2.2.2 Improved Masking Method

The "Improved" antenna masking method performs antenna masking by delegating the calculation of the angles of incidence
to the propagation model. If the propagation model does not implement the appropriate methods of Atolls API, the
"Improved" antenna masking method is not available.
You can adjust the following parameter when using the "Improved" antenna masking method:
Parameter
Description
Antenna pattern interpolation The antenna gain calculation method for deriving the antenna gain from a set of angles of
incidence. Either:
Native 3D interpolation method: The method used by Atoll. For more information
on Atolls method for 3D interpolation, see the Technical Reference Guide
Linear interpolation method: A simple linear method with a smoothing parameter:
G = A hor azi + A ver elev smooth
The "Improved" antenna masking method usually gives accurate results.
13.2.2.3 Full Path Loss Masking Method

With the "Full Path Loss" antenna masking method, ACP recalculates all path loss matrices for all combinations of parameters
which are tested. This is a fall-back method for complex propagation models not accurately modelled by the "Basic" or
"Improved" antenna masking methods, for example, for complex ray tracing propagation models.
When using the pre-calculated antenna masking method, Atoll ACP first calculates new path loss matrices for every possible
combination of antenna parameters which needs to be tested. The optimisation process then uses these pre-calculated path
loss matrices to determine how attenuation changes when an antenna is modified.
ACP does not calculate all path loss matrices for all possible combinations, for example, five possible changes in electrical tilt
and five possible changes in azimuth, i.e., 25 path loss matrices. ACP only calculates the path loss matrices for the changes
885

which need to be evaluated by the optimisation algorithm. By pre-calculating only this subset, ACP reduces the number of
path loss matrices to be calculated and the calculation time.
If a change is tested on a transmitter that was not taken into consideration when the path loss matrices were calculated, ACP
recalculates the path loss matrices for that change only.
The main drawback of the pre-calculated antenna masking method is the lengthy pre-calculation required and the disk space
required to store the path loss matrices. Therefore, the following are recommended:
Use the pre-calculated antenna masking method only when necessary. If the "Basic" or "Improved" antenna masking
method gives accurate ACP predictions that are in line with Atoll, use one of these methods instead.
When using the pre-calculated antenna masking method, limit the number of parameters covered. For example, only
enable two or three azimuth options. Also carefully design your antenna groups.
Use a path loss storage directory which is dedicated to your project region. This ensures that future optimisations in
that region will have path loss matrices that have already been calculated.
Power optimisation and site selection (without reconfiguration) do not require

recalculation of the path loss matrices.
Antenna height reconfiguration as well as new candidates always use a method
similar to the "Full Path Loss" antenna masking method to calculate missing path
loss matrices.
13.2.2.4 Antenna Correction Masking Method

The "Antenna Correction" antenna masking method performs antenna masking by delegating both the calculation of the
angles of incidence and antenna 3D interpolation to the propagation model. Antenna losses recovered by ACP may include
antenna correction and 3D antenna extrapolation.
If the propagation model does not implement the appropriate methods of Atoll's API, the "Antenna Correction" antenna
masking method is not available. If available, the "Antenna Correction" method is the recommended one an it usually gives
the best results.
13.2.2.5 Antenna Masking and Repeaters, Remote Antennas, Secondary Antennas

ACP fully supports repeaters, cascaded repeaters, and remote antennas. The repeater or remote antenna can be reconfigured
for the "coverage side". However the "donor side" of repeaters can not be reconfigured.
ACP correctly takes into account the path loss produced by transmitters using secondary antennas.
13.2.3 CrossWave Propagation Model

ACP supports the CrossWave propagation model as a native model using the "Optimised" antenna masking method and
delegating the calculation of the angles of incidence to the model.
However, the clutter height files and DTM must be accurate in Atoll so that ACP can access the terrain profile (even when you
have configured CrossWave to directly access building vectors).
13.2.4 Aster Propagation Model

ACP supports the Aster propagation model as a model using the following antenna masking methods:
"Basic Masking Method" on page 885

"Improved Masking Method" on page 885
"Full Path Loss Masking Method" on page 885. This is the recommended method (default).
"Antenna Correction Masking Method" on page 886
13.3 Configuration
13.3.1 Configuring an Optimisation Setup
Setting up the reconfiguration parameters is straightforward. For each parameter change, a range for the parameter can be
specified, for example:
886
Maximum variation for azimuth

Minimum/maximum range for electrical tilt, mechanical tilt, power, height, etc.

AT332_TRR_E0
You have the option of locking height and azimuth optimisation per site. In other words, when a change to antenna height or
azimuth is made to one transmitter on a site, the same change is made to all transmitters of a site. In the case of the azimuth,
refers to the rotation of the mast, and for antenna height, all antennas will be all displaced to the same height.
By default, azimuth locking is disabled, while height locking is enabled by default for all co-localised transmitters on the same
site.
13.3.1.1 Antenna Setup

Electrical tilt and antenna model optimisation require correct antenna modelling. The concepts on which Atoll ACP antenna
modelling are based are the following:
Antenna Element: An antenna element groups all instances of an antenna, belonging to the same frequency band,
with different electrical tilts.
Physical Antenna: A physical antenna is a multi-band antenna, grouping all antenna elements from different
frequency bands which are physically the same antenna.
Antenna Groups (Optional): An antenna group is a user-defined subset of the physical antenna enabling you to select
antenna model reconfiguration to be done within this subset.
Modelling the antennas normally only needs to be done once. Atoll enables you to carry this out in several different ways:
Using the Physical Antenna field of the Antenna table: You can assign the same name in the Physical Antenna field
in the Antenna table to antennas belonging to the same physical antenna, independently of the frequency band they
use. ACP will then use this information to automatically create all antenna elements and physical antennas. The
physical antenna name is displayed in the "Model" column of the Antenna Pattern Table.
You can also create a custom column in the Antennas table to automatically link antenna elements of a multi-band
physical antenna which have the same electrical tilt. You must identify this column using the "antenna.etilt.share"
option in the ACP.ini file. For more information, see the Administrator Manual.
This is the preferred method, as ACP will then automatically create all antenna elements and physical antennas each
time a setup is created.
Manually defining Antenna Elements and Physical Antennas: You can manually define antenna elements and
physical or use a REGEX expression. For more information on manually defining antenna elements and physical
antennas, see the User Manual.
Detecting automatically the "Freq. Band" field in the ACP "Antenna Pattern Table":
If a FREQUENCY custom field exists in Atoll, the value it contains will be extracted.
If a FREQUENCY custom field does not exist in Atoll or exists and is less than or equal to 0, then a frequency will
be determined in the following order:
If "antennaPattern" is referenced by a transmitter, the frequency defined by transmitters FREQBAND is used.

FREQBAND is the "Frequency" field (in GSM) or "Start Frequency" field (in UMTS and LTE) in the Frequency
Bands table available from Parameters > Network Settings > Frequencies > Bands.
If the project contains a single FREQBAND, then the frequency defined by this FREQBAND is used
If "antennaPattern" defines a [FMIN-FMAX] range, the first FREQBAND defining a frequency within this range
is used. Else, the hard-coded value (935, 2110, or 1805) contained in the [FMIN-FMAX] range is used. If all fails,
then the value is set to 0 and the cell remains empty.
13.3.1.2 Additional Electrical Tilt (AEDT)

ACP supports additional electrical downtilt (AEDT) processing. AEDT is used when antenna patterns are not available for
changes in electrical tilts. The patterns are derived by Atoll ACP using geometric down-tilts of the original antenna pattern.
You can enable AEDT support in ACP by setting the following option in the ACP.ini file:
[ACPAntennaPage]
enableAedt=1
When you have activated AEDT support, new columns appear in the Antenna Pattern table on the Antenna > Patterns vertical
tab to allow you to configure which antenna uses AEDT and the range of allowed electrical tilt.
You can use the following ACP.ini options to reference custom columns in the Antennas table. ACP will use the data entered
in these custom fields to set the default values in a new optimisation setup.
[ACPCustomFieldExtraction]
antenna.aedt.use=ACP_AEDT_USE
antenna.etilt.min=ACP_ETILT_MIN
antenna.etilt.max=ACP_ETILT_MAX
887

For more information on the available options in the ACP.ini file, see the Administrator Manual.
13.3.1.3 Relative Electrical Tilt Values

By default, ACP allows the reconfiguration of electrical tilt parameters based on absolute values. The same default settings
apply to mechanical tilt parameters.
The following option allows you to display the electrical tilt values in the Transmitters table (on the Reconfiguration tab) as
relative values, rather than absolute ones:
[ACPReconfPage]
tx.etilt.asRelative=1
The following option can be used to create an additional constraint on the Reconfiguration tab that will be applied to electrical
tilt changes. This constraint enables the user to define a range of electrical tilt changes within a defined number of degrees
above or below the current electrical tilt. The following example forces ACP to find an optimal electrical tilt 4 degrees higher
than or 4 degrees below the current electrical tilt, for all transmitters.
[ACPReconfPage]
tx.etilt.deltaLimitConstraint=4
13.4 Multi-RAT and Co-planning Support

ACP fully supports multi-frequency band, multi-RAT, and co-planning documents.
13.4.1 Multi-RAT and Co-planning Modes

When working in co-planning mode with several Atoll documents, ACP enables you to import the other Atoll project into one
ACP setup. The ACP setup then becomes a multi-RAT setup.
The benefits of using ACP in multi-RAT mode are:
You can define multi-frequency band/multi-RAT combined objectives

You can automatically synchronise shared multi-band antennas. This ensures that any antenna reconfiguration is
properly taken into account in all impacted technologies.
Shared site location is automatically managed for site location and site candidates. In a multi-RAT document, you can
upgrade existing sites with a new radio access technology, and take the upgrade cost into consideration.
ACP automatically detects multi-technology sites and shared multi-band antennas, provided that the co-located transmitters
are within a user-definable inter-antenna distance (default = 1m). It also automatically detects co-located sites and antennas,
including secondary antennas, using the following algorithm:
Parameter
Co-located site
Description
User-definable inter-site distance (default = 2m) using the ACP.ini option below:
[ACPTplReconfPage]
site.min.distance.colocated = 2
Co-located Antenna
(i.e. Transmitters)
888
User-definable inter-antenna distance (default=1m) using the ACP.ini option below:

[ACPTplReconfPage]
tx.min.distance.colocated = 1
Antenna height within 1 metre
Antenna azimuth within 2 degrees
Mechanical tilt within 1 degree

Same physical antenna when the antenna defines this field

AT332_TRR_E0
Occasionally, problems in the Atoll database can mean that ACP does not
recognise that sites or antennas are co-located. If this happens, you can manually
set the sites or antennas to be co-located, although you should also review the
database to correct any errors there.
The normal way of detecting linked transmitters is to use the "Shared Antenna"
field in the Transmitters table (SHAREDMAST). If at least one transmitter defines
a "Shared Antenna", then the following logic is used:
Detection of co-located sites as sites located within a user-definable inter-site

distance (default = 2m). All transmitters from co-located sites with the same
shared antenna are linked if they have a different frequency band, or if the
technology is different.
Sanity check is performed to validate that antenna parameters are
consistent: same position, same azimuth, same mechanical tilt, and same
antenna height (within a user-definable inter-antenna distance [default = 1m]
for position, within 2 degrees for azimuth, within 1 degree for tilt, and within
1 metre for antenna height).
When two linked transmitters are not consistent, ACP will issue a nonblocking warning.
When two transmitters are linked, their values in the Current columns on
the Reconfiguration > Transmitters vertical tab are highlighted in red.
If the "Shared Antenna" field is not used by at least one transmitter, then ACP will
use another mode where it automatically detects the linked transmitter using the
same criteria as the one used in sanity check (within a user-definable interantenna distance [default = 1m] , within 2 degrees for azimuth, etc.).
13.4.2 Technology Layer Definition

ACP sees each radio access technology as one or several technology layers which are defined according to the following rules:
Technology
Technology Layer Definition
Example
GSM
Each independent frequency band is seen as a

separate technology layer
GSM 900
GSM 1800
UMTS and CDMA
Each carrier is seen as a separate technology layer
Carrier 10562 of 2110 MHz band

Carrier 10587 of 2110 MHz band
Carrier 2937of 925 MHz band
LTE and WiMAX
Each independent frequency band is seen as a

separate technology layer
2010 MHz band

900 MHz band
When defining objectives, each rule is associated with a single technology layer.
Each quality indicator is evaluated for the technology layer to which it is assigned, however you can group quality indicators
from different technology layers within a same objective.
When you are using ACP with more than one technology layer (and, therefore, in multi-RAT projects as well), you can put more
emphasis on some technology layers by modifying the global weight of the objectives of each technology layer.
ACP optimises the quality objectives for all technology layers. All are considered the same;
there is not, for example, one target technology layer and one or more constraint
technology layers. If one needs to be optimised without degrading others, you need to:
Define a heavier weight on the objectives related to the "target" technology layer,
Or use a coverage target for the objectives of the "constraint" technology layers
which are relative to the current coverage (where a successful optimisation would
be defined as "no coverage decrease").
13.5 Optimisation Methodology

When ACP performs an optimisation, all objectives (quality and cost) are combined into a single global score function. This
global score function is used as the basis for the search algorithm, which means the algorithm attempts to find the best
parameter combination to minimise the global score function.
889

In the final step, a sorting algorithm provides an implementation plan where the most useful changes (in terms of minimizing
the score) are done first and the least useful changes are done last.
13.5.1 Search Algorithm

The possible number of configurations grows exponentially with the number of sectors to optimise. The naive search method,
consisting of enumerating all possible solutions, very quickly becomes unmanageable. For example, selecting the best
antenna among 10 possible antennas on a 100-sector network, leads to a search space of 10100 possible solutions, i.e., more
than the number of atoms in the universe.
ACP uses a Tabu-based search algorithm with fast convergence. In short, this algorithm performs local greedy optimisation,
while allowing for the exploration of new locations in the search space. When a transmitter has been allowed a parameter
change, any new change is forbidden during a certain number of iterations.
A number of additional techniques are used to improve the basic process, such as randomisation, diversification, aspiration,
and long-term search. Knowledge of the particular nature of the network (cell neighbour relations, for example) is also used
to improve the process and make it efficient.
This search algorithm uses the concept of iterations: each iteration consists of one parameter change on one of the sectors
or sites of the network. The number of iterations is a key parameter of the optimisation, and should be high enough to ensure
that the search space is properly covered. Usually a few times the number of entities to optimise is sufficient, although this
parameter still depends strongly on the size of the network and the quality of the initial network.
It should be noted that a given sector might be modified in several steps, i.e., the final change might be the result of several
different iterations. Some iterations might also cancel each other, i.e., a sector is returned to its initial state at the second
iteration.
ACP recommends a number of iterations. The recommended number of iterations is
calculated by multiplying the number of entities to optimise by two. By defining a number
of iterations equal to or greater than the recommended number, you ensure that the
search space is explored correctly.
13.5.2 Tuning Algorithm

ACP search algorithm includes a tuning phase between the search and sorting phases (corresponding to optimisation and
finalisation phases in the ACP GUI).
The tuning phase consists in improving the best solution found during the search phase. It performs a local optimisation of
the network while preventing useless changes from being done.
During the tuning phase, ACP proceeds as in the Tabu-based search phase, but without using a Tabu list, randomisation, etc.
It simply finds the best neighbour candidate to move after each iteration, i.e. the change which mostly improves the score
function STOT(x). The tuning phase stops as soon as the score function (see "Global Score Function" on page 892), can no
longer be improved.
The tuning phase is fully transparent to the end-user and it provides the following benefits:
Removes the changes with insignificant benefits, i.e. changes with less than 1% of the benefit provided by the best
single change.
Finds the local optima around the best solution of the optimisation phase. Due to the behaviour of the optimisation
phase, and to the possibility of early stopping, it can happen that ACP finds the non-optimal local optima.
Allows the Tabu-based search phase to concentrate more on spanning the solution phase, without having to also
perform local optimisation around the best candidate solution.
Provides a better automatic stop condition for the Tabu-based search phase, and better management of early manual
stops by the user.
Will be useful in future releases for better management of multiple solution findings corresponding to different
qualities or cost trade-offs, i.e. different points on the Pareto surface of multi-objective optimisation problems.
This feature is fully transparent. The requested number of iterations is used in both phases. By default, about 2/3 of the
iterations are used in the Tabu-based search phase, and 1/3 in the tuning phase. These ratios can differ when an early stop
(automatic or manual) is performed during the Tabu-based search phase.
On the Graph tab of the Optimisation window, a vertical bar is displayed to show the switch point between optimisation and
tuning phases.
890

AT332_TRR_E0
Figure 13.3: Graph tab of the Optimisation window
13.5.3 Sorting Algorithm

After applying the search and tuning algorithms, ACP proposes a solution consisting of a number of changes to be applied to
the initial network. A change is typically a modification to an antenna parameter or (for candidate sites) deploying a site or
(for existing sites) sector or removing a site or sector.
ACP then uses a sorting algorithm to create an ordered implementation plan. The sorting algorithm recursively finds the best
change to apply among all remaining changes. The best change is the one which improves the total cost function the best:
The first changes proposed have more benefits (in terms of the trade-off between quality and cost) than later changes.
It is then possible to select a subset of the total number of changes by selecting only the N first changes.
13.5.4 Principles of the Pre-optimisation of Activated Sites

The site selection algorithm considers the sites that are selected and activated. These sites are directly pre-optimised when
they are activated, which makes their insertion in an existing network more efficient.
Each candidate site is put in competition with others for the best candidate selection; this candidate evaluation is based on
the ACP score function and, to make sure that the best configuration is evaluated, site parameters are reconfigured according
to specific rules.
In other words, each candidate is pre-optimised before the algorithm makes the optimal choice. This avoids a potentially good
site to be rejected due to the initial configuration, for example caused by an incoherent or not adapted initial configuration
(tilts, azimuths, and so on).
The specific reconfiguration rules that apply only in this process are the following:
For each parameter (tilt, azimuth), all values within the min/max range are evaluated according to the defined step. However,
for performance purpose and to reduce the number of site configurations to evaluate, the values are locked between sectors
within the site. For example, all sectors are allocated the same tilt value for a configuration X (0,0,0 then 10,10,10, and so on),
and all azimuths are locked within the same site.
A second-stage reconfiguration can still occurs once the site is activated to reconfigure each sector with different parameter
values.
The benefits are:
A lower number of sites is selected for the same quality improvement or a higher quality improvement is obtained for
the same number of sites.
An optimal insertion in an existing network: a candidate site having a higher probability to be inserted while improving
the overall quality.
891

13.5.5 Global Score Function

13.5.5.1 Search Algorithm
The global score function used as a basis for the search algorithm is created by a linear combination of the cost objective and
every quality objective.
The global score function is in the following form:
C TOT x =
ai fi Qi x + k fc C x
i quality obj
Where:
i is an index spanning all quality objectives defined
n is the network configuration to be tested
Q i x is the "ith" quality objective evaluation
a i is a weight factor associated with the "ith" quality objective, and applies different importance on the different
quality objectives for the different technology layers.

C x is the (financial) cost associated with configuration "x"
f i is a one-dimensional function expressing the individual given cost for the "ith" quality objective measurement.
The coverage costs are null if the target coverage is reached: f i x = 0 for x T arg etCov
f c is a one-dimensional function mapping the network financial cost suitable to be used alongside the quality
objective costs.
Example
C TOT x = a obj1 f obj1 Cov obj1 x + a obj2 f obj2 Cov obj2 x

Where: Cov obj1 x (or Cov obj2 x ) is the percentage of coverage over a specified threshold for configuration "x".
For example, in UMTS:
1
Cov obj1 = --N
i 1 Th E c i Thresh Ec
i pixels
Where: 1 Th is the step function.
1
Cov obj2 = --N
i 1 Th E c I o i Thresh EcIo
i pixels
Where:
i is the normalised traffic density on pixel "i" such as Max i = 1
In case of uniform weighting, all i are equal to 1.
In case of non-uniform weights, weights vary between 0 and 1 with Max i = 1 .
E c and E c I o are the basic quality measurements on one pixel as described earlier.
f obj1 and f obj2 are the one-dimensional mapping functions expressing the individual cost for a coverage figure or
network quality:
f obj1 cov = 0 for cov T arg etRSCP

f obj2 cov = 0 for cov T arg etEcIo
13.5.5.2 Tuning Algorithm

The global score function used as a basis for the tuning algorithm is the following:
S TOT ( x ) CTOT ( x ) p f n N ( x )
892

AT332_TRR_E0
Where:
CTOT (x)
N (x)
is the network configuration to be tested

is the global score function described earlier
is the number of changes performed in the network from the initial configuration
is a weight factor derived automatically to have an appropriate scaling of the new term with
chosen such that the "cost" of one change is equal to
q%
of the score function improvement provided by the
best individual change in all the proposed changes. By default
fn
CTOT (x) . It is
q 1. 25%
is a one-dimensional function expressing an individual cost for a given number of changes. By default, in initial
implementations, it is the identity function:
fn y y
13.5.6 Weighting
Several types of weight are applied during the calculation of the global score function.
On a geographical level (used to calculate the weights i in the above formulas):
Traffic weighting: Each pixel can have an importance proportional to the traffic supported on it.
Zone weighting: Each pixel within a defined zone (computation, focus, hot-spot zones or clutter-group zone) can have
an additional weight which increases or decreases the importance of the zone.
On a global level (forming the weights a i in the above formulas):
Quality objective weighting: Within a technology layer, each quality objective can be given more or less importance
as compared to other quality objectives of that technology layer.
13.5.7 Controlling the Optimisation

Although the ACP optimisation process is designed to be as automatic as possible, there are a couple of parameters that
require some consideration:
Number of iterations: This option defines the number of iterations in the search algorithm.
Resolution: The resolution defines the size of the pixels used to measure the quality objectives.
These parameters affect the quality and speed of the optimisation.

If the resolution is high, ACP does a better job of sampling the network zone, but takes longer to run. If the resolution is low,
the sampling is more approximate but the speed is highly increased. As a suggestion, ACP provides information on the total
number of pixels, as well as the average number of pixels per site.
Similarly, if the number of iterations is high, the optimisation will likely find a better solution but will take longer to run.
The following table gives typical values to be used for a good optimisation:
Parameter
Typical Value
Number of iterations
Around 1 to 2 per item (antenna, azimuth, tilt, cell pilot

power, etc.) to be optimised
Resolution
The average number of positions per site is between 300

and 3000.
You can start with a low resolution first, before using a higher resolution for more
accurate results. When ACP is running an optimisation, the tabs of the Optimisation
dialog box provide feedback which can help you to decide to stop the optimisation early
if the overall network quality seems to have improved enough.
13.5.8 Implementation Plan

The sorting process in the implementation plan is based on a global score function which includes quality objectives and the
cost of changes (i.e. it takes into account the quality settings as well as the Cost Control settings if cost of changes is
considered, with the selection of Maximum Cost or Quality/Cost tradeoff options).
893

C TOT x =
ai fi Qi x + k fc C x
i quality obj
Figure 13.4: Setup > Properties > Optimisation tab > Cost Control dialog box
Only the data displayed on the Change Details tab is actually separated for quality objectives and cost of changes:
Quality Improvement Ratio %: this column shows the ratio of attained quality VS the maximum quality when all
changes are made (the displayed values range from 0% and 100%). This ratio allows you to know the relative gain of
each change. It takes into account the coverage and quality objectives (i.e. LTE RS Coverage and LTE RS CINR) and, if
used, the financial cost. However, the load balancing objective is not considered.
Total Cost: this column shows the associated cost of changes.
The data is separated as it makes more sense to display understandable values in 2 columns rather than display a "Score
improvement ratio %" which would be difficult to understand.
One consequence of this is that if you obtain 2 changes providing the same quality improvement ratio, then the first one in
the list will be the one with the lowest associated cost.
Figure 13.5: Setup > Optimisation > Properties > Change Details tab
894

AT332_TRR_E0
13.5.9 Memory Usage and Optimisation Resolution

The administrator can set an option in the ACP.ini file to set a limit on the amount of memory that ACP can use. If this is the
case, you might reach the set limit when using a high resolution and Atoll will stop the ACP optimisation early. If this happens,
you can rerun the optimisation by decreasing the resolution or decreasing the size of the computation zone. As well, when
performing an optimisation on a large area, you can limit memory usage by performing the optimisation in several steps, each
time on a different portion of the entire area. For more information on ACP memory limits, see the Administrator Guide.
13.5.10 Internal Data Management and Performance

13.5.10.1 Memory Usage
For each tested network change, ACP recalculates how each objective improves or degrades. Each objective is the
combination of one or several quality indicators evaluated on several hundred thousand or even millions of pixels depending
on the resolution and network size.
ACP needs to store, for each pixel, a list of the neighbouring cells in order to be able to find new best servers and calculate
interference levels.
Given the amount of data processed, it would be prohibitive in terms of the amount of time necessary to read path loss
matrices from the disk on each iteration. For each pixel, ACP keeps the list of neighbour cells and their attribute (path loss
attenuation, for example) in memory in a format optimised for fast processing.
These internal data structures are created during the loading phase of an ACP run. Any memory issue that could occur when
optimising a large network should happen during this initial phase, as allocations to memory are very limited during the
solution search.
You can change how ACP manages the data it loads into memory by setting certain options in the ACP.ini file. For example,
you can set the number of cells that ACP monitors by setting the "maxMonitorCell" option in the ACP.ini file. You can use the
"threshLevelMonitorCell" option to define the best server signal threshold (dB) of the cell in order for the cell to be monitored.
For more information on these options, see the Administrator Guide.
In practice, however, it is usually sufficient to change the mode of operation, thereby performing a trade-off between memory
usage, and the accuracy of the optimisation (especially on measures related to interference):
High speed: The cell list is shortened to reduce memory use, and the algorithm is optimised to improve speed.
Normal: The normal mode with a balanced trade-off between speed, memory use, and accuracy.
High precision: The high precision mode results in higher memory usage and a lower speed, but offers the highest
accuracy by monitoring a longer list of cells.
When doing site selection in Greenfield scenarios where a lot of candidate site are defined
close to each other, it is recommended to use the High precision mode in order to ensure
that all neighbour candidate sites are well monitored by ACP.
13.5.10.2 Disk Space Usage

To reduce disk space usage, the user can define in Windows a private storage directory for ACP with compression set to ON.
This also holds for the path loss directory.
To better use ACP and avoid lengthy recalculations after rollback changes, specify a Shared directory for path loss matrix
storage in the predictionss Properties dialog box.
13.6 Load Balancing Objective

13.6.1 Principle Used in ACP
One obvious approach for load control and balancing within ACP is to compute the new actual cell loads for each tested
reconfiguration. By detecting when the cell loads become imbalanced or in excess of given cell resources, it is possible to avoid
such reconfiguration.
The overall difficulty is the calculation of cell loads given the precise inputs required for traffic, service, etc., (but which are
often imperfect in practice), as well as the complex and lengthy calculation involved which would require full Monte Carlo
simulations for each tested reconfiguration.
Another approach is to consider that the actual cell loads are fixed, i.e. fixing the level of interference generated in the
network corresponding to a target load, and then make sure that the supported traffic is maximised and well balanced. This
approach avoids the complexity of cell load calculation through complex Monte Carlo simulation, and we believe it is more
robust to imperfect inputs. In particular this approach would still provide useful insights on load imbalance from only a partial
895

traffic model (considering for example only 1 or 2 services typically representative of the traffic distribution). This is the
approach used in ACP.
Since the cell power load is fixed in ACP calculation, another method is needed to insure that the real cell load does not
increase beyond capacity and is correctly balanced between cells.
The ACP approach is basically to perform cell capacity load balancing. The cell capacity load is not the actual cell load derived
from a set of realistic traffic maps and services using a Monte Carlo simulation with power convergence loop, but is only a
capacity indicator computed by assuming a fixed cell load and a traffic model which could be simplified model as long as it is
representative of the actual traffic distribution.
For UMTS R99, the cell capacity load is related to transmitted power:
A total transmitted power is computed over the whole Best Server area by adding transmitted power for each pixel.
Each pixel transmitted power is computed by using the load factor equation, but still assuming that the cells have a
fixed load for the purpose of interference calculation. Each pixel power is scaled with traffic density distribution. The
ratio of this total cell power over the maximum cell power is the cell capacity load, and is expressed in % of the
available resources.
The ACP equalises the cell capacity load, avoiding that some cells use excessive power.
Other technologies use the same principles, however with a different definition of cell capacity load.
The key reason why cell capacity load balancing is a suitable approach is that cell capacity load is correlated to the actual cell
load. When the cell capacity loads are being successfully balanced, they tend to converge loosely towards the actual cell load.
Any imbalance in the actual cell load is thus reflected by imbalance in the capacity cell load.
Moreover, by focusing on a what-if scenario where cell loads are set to an average target load (using for example a target
network load of 70%), then the computed capacity cell loads measure directly how much room is available or missing in actual
cell capacity to support this target network load.
13.6.2 Optimisation Principle

The calculation is performed in three steps:
Assignment of the requested traffic to the various cells on the network, across technology layers.
Computation of cell capacity load from the assigned traffic.
Derivation of the score function of the load balancing objective from the cell capacity loads.
13.6.2.1 Traffic Capture for Load Balancing

ACP is designed to perform load balancing across multiple technology layers. This means that the requested traffic will be
shared across the available technology layers within the allowed technologies for this service.
The requested traffic for each service is assigned to cells according to the following rules:
Candidate cells for assignment of a pixel traffic are selected among all best server cells in all different technology
layers, and for which:
The technology layer is allowed for the service.

The service traffic capture conditions are fulfilled.
The traffic is assigned partially to each of these candidate cells, such that the cells capacity load is minimal.
The procedure uses a simulation process where the pixel traffic is added gradually to the network, and the cell capacity loads
are updated after each assignment, followed by a convergence loop to reach a minimum state.
This process basically simulates a network where the traffic is dynamically assigned to technology layers in order to equalise
cell loads. For example, if a pixel traffic requested can be assigned to Cell1 from TechnologyLayer1 and to Cell2 from
TechnologyLayer2, it will be assigned to the Cell with the minimum cell load.
The end result of this process is to distribute the traffic across cells and technology layers in such a way that overlapping cells
from different technology layers tend to be equally loaded. For example, in UMTS the cells belonging to same transmitter
usually have the same footprint, and as such the computed load will tend to be equal across these cells; for cells which do not
fully overlap, the process still tends to equalise technology layer loads as much as possible.
In other words, this process simulates a perfect call admission control procedure whose purpose is to perfectly equalise the
requested traffic among all cells and technology layers in the network: a new call is always assigned to the technology layer
and cell having the minimum load. Technology layers with high capacity (for example LTE vs GSM) tend to acquire more traffic
automatically.
13.6.2.2 Cell Capacity Load Calculation

The traffic assignment stage basically balance the traffic request across the different cells, while computing the cells capacity
loads. The exact method for computing the contribution of a pixel to the cell capacity load is technology-dependent.
The cell capacity load is the sum of all pixel contributions, scaled with assigned traffic to the cells technology layer:
896

AT332_TRR_E0
Li =
T k i k
Pixel k Cell i
Where: k is the pixel load ratio and is T k i is the traffic assigned to pixel "k" and cell "i".
The unit of capacity load is a percentage (%).
In UMTS R99
The load factor equation used is the following:
Eb No Io
1
= ----------- -----------------------P max A tt G proc
Where:
is the pixel load ratio
P max is the maximum cell power
E b N o is the target EbNo for the given service
I o is the total noise and interference
A tt is the attenuation towards the cell, including antenna gain and losses
G proc is the service processing gain (Gproc = 3.84e6/Tputserv)
In LTE
Given the pixel SINR, one derives first the best Bearer which can be assigned to a call originating from that pixel, then derives
the maximum possible throughput Tputmax which could be provided to that pixel-originated call (given bandwith, etc). The
pixel load ratio is then the ratio of resources used by the service on this pixel, and is given by:
Tput serv
= ------------------Tput max
Where Tput serv is the average requested throughput for the service
13.6.2.3 Load Balancing Score Function

The Load Balancing Score Function being minimised to drive ACP optimisation is the following:
Score = QI 1 + b 1
Where:
QI is the Load Quality Index, whose minimisation reduces both the average cell load and load imbalance
(explained below with formulas)
is the Traffic Captured ratio which measures how much traffic is potentially served in the target zone, due to
traffic capture condition being fulfilled
b is a scaling factor to give more or less weighting to the traffic captured ratio (default is 1)
This score function will be considered in the ACP global score function when load balancing is activated. This is described in
"Impact on the Global Score Function" on page 901.
13.6.2.4 Load Quality Index

The Load Quality Index is defined as:
QI = Mean QI l
Where QI l = l + a l and is the Quality Index computed for technology layer "l"
Where:
l is the weighted average of the cell capacity load
l is the weighted standard deviation of the cell capacity load
a is a scaling factor to give more or less weighting to the standard deviation, i.e. load imbalance (default =1)
897

1
l = --------------
Wli Li
W
li Celli Layerl
i
l =
1
2
-------------
Wli Li l
Wli Celli Layerl
i
Where W li is a weighting factor applied on each cell load, for technology layer "l", and is used to reduce the effect
or completely deselect a cell in the calculation:
Cells inside the target zone are considered as having a weight of 1, and cells outside the target zone are
allocated a weight of 0.
Inactive cells are not considered in the calculation (in term of average/standard deviation values and number
of cells).
The overall Load Quality index is an average of the all the technology layers Load Quality
indexes, i.e. each technology layer gets identical importance.
13.6.2.5 Captured Traffic Ratio

The captured traffic ratio is defined as:
T ass
= -------T req
Where:
T req is the total traffic requested in the target zone
T ass is the total traffic assigned in the target zone
Increasing scaling factor "b" leads to increase the total traffic assigned when the Score function is being minimised.
See definition of "b" in "Load Balancing Score Function" on page 897.
13.6.2.6 Introduction of Load Balancing as a Quality Indicator

The Load Quality Index can be rewritten as follows:
l
QI l = l 1 + a ---- = l 1 + aD l
l
Where:
D l = ----l is the Load Dispersion measure for technology layer "l"

l
the Load Balance (B) is defined as B = 1 D l
This load dispersion parameter directly measures the load imbalance. Hence, minimising the Load Quality Index can be seen
as a way of decreasing both the Load Dispersion (thus improving the Load Balance) and the Average Load (thus improving the
network capacity).
A default Load Balance Target value is defined for the Load Balance in the user interface (e.g. 80%). When this target is
reached, the Load dispersion part of the Load Quality index is set to null, since the objective is above target. This Load Balance
target allows a margin, in other words a low level of imbalance, for example 10% or 20% without further optimisation. Hence,
when the Load Balance target is reached, only the average load ( l ) is considered for minimisation in the score function.
The minimisation of the Quality index can also be seen as minimisation of the number of overloaded cells. Let's make the
assumption that the cell capacity loads are drawn from a Normal random distribution (Gaussian process). It is completely
described by both mean and standard deviations.
898
By decreasing the Quality Index with "a = 1", we then decrease the load of the 85th percentile of the distribution.
By using "a = 2", the load decrease is for the 98th percentile of the distribution.

AT332_TRR_E0
Figure 13.6. Cell Load Distribution

Percentiles represent the area under the normal curve, increasing from left to right.
As the bottleneck in network capacity for a given quality is often given by its most loaded cells, we directly increase the
capacity by focusing on the high distribution percentile. In the most general cases, the cell load distribution is not derived from
a Normal distribution, however the argument still holds: decreasing the + a will focus on the capacity network
bottleneck.
13.6.3 Quality Figures Used for Graphs and Statistics Results

ACP provides graphs for the Load Balance and the Average Load values, in terms of quality figure for an easy understanding
by users. The goal is to provide a percentage value for an improvement and a graph which increases when the quality indicator
increases.
13.6.3.1 Load Balance

The figure is designed to show 100% for perfect load balance, and to show 0% for total imbalance. However, the formula used
for calculation, B = 1 --- , must be adapted due to negative values obtained when . Consequently, the formulas used
for the Load Balance quality figure are:
LB = 1 0,8 --
when
1
LB = --------------------------- 2
1 + 4 ---
when
It is displayed as a curve on the Graphs tab, with initial and final values available on the Statistics tab. Note that the Load
balance value will tend to increase (and the graph to go up) when cell capacity loads are better equalised.
13.6.3.2 Average Load

The quality figure is given in terms of improvement (%) from the initial average load. The formula used for the quality figure is
0
----- 1
Where 0 is the initial value and is the final value, or 0 the value for step N and the value for step N+1.
It is displayed as a curve on the Graphs tab and the final improvement can be found on the Statistics tab.
In both cases, a 100% improvement means a decrease by 2 of the average load, 200% a decrease by 3, -50% an increase by 2,
-75% an increase by 4, etc.
899

This quality figure will tend to increase when the average load decreases, which is
expected during optimisation.
13.6.4 Optimisation Results

13.6.4.1 Load Balancing Tab
The Average Load and Load Balance quality figures are shown on the Load Balancing tab for any specified zone. They are
based on the cell capacity loads which are displayed on the right on the Load Balancing tab. Hence, from the list of cell capacity
loads, it is possible to recompute these quality figures using the formulas described in the previous section.
The graph also directly shows the cumulative distribution of cell capacity loads, thus providing the ratio of cell capacity load
being smaller than some load value.
Figure 13.7: Setup > Optimisation > Properties > Load Balancing tab
The initial/final load balance values displayed on the Statistics tab and the Load Balancing tab will be identical if:
The same technology layers are considered:

In the Load Balancing page of the setup properties Objectives tab
Next to For Technology Layer on the optimisation properties Load Balancing tab
The same zone is considered:

Next to Evaluation Zone on the setup properties Optimisation tab
Next to For Zone on the optimisation properties Load Balancing tab
The values displayed on the Statistics tab are calculated for the cells based on the selected technology layers or hetnet layers
and located within the target zone (more precisely for the cells that are actually considered for load balancing).
However the capacity load statistics displayed on the Load Balancing tab are calculated based on the technology layers or
hetnet layers and the zones that are currently selected in the dialog box. Therefore, the displayed initial/final load balance
values can be different on each tab.
13.6.4.2 Graphs
The graph representing the Load Balance quality figure shows the progress of this quality figure for each iteration.
900

AT332_TRR_E0
Figure 13.8: Graphs for Quality Figures
Figure 13.9: Statistics results (extract)
13.6.5 Impact on the Global Score Function

The Load Balancing Score Function is added to the ACP global score function, which already includes quality objectives and
the cost of changes.
The weighting of the load balance objective versus other quality objectives (e.g. coverage) can be modified from the user
interface (Objectives tab > Load Balancing > Weight parameter).
It is calibrated in such a way that X% improvement of the Load quality index is equivalent to an identical improvement on other
objectives. For example, 1% improvement of the Load quality index is equivalent to 1% improvement of the RSCP coverage.
13.7 Throughput Objective

The LTE throughput objective is designed to optimise the throughput (which means maximisation) obtained through all
network layers, with consideration of layer priority management. This objective requires traffic data as input, supports
uniform distribution or Atoll traffic maps, represented with the following data types: user ratio, service, terminal, and
mobility.
This objective allows and offers:
Support of layer priority (HetNet layers)

Management of traffic demand, with a traffic capture in line with best server selection. The goal is to assign in real
time users to their best server according to layers and priorities. Both uniform distribution and Atoll traffic maps are
supported.
Integration in the ACP global score function
Production of output results (tables, predictions)
901

13.7.1 Optimisation Principle

The throughput objective is built on traffic map, which is shared with load balancing. Both objectives are designed to work in
parallel with different target zones. The objective goal is the optimisation of cumulative throughput:
A first step is based on an accurate calculation per pixel of the throughput value considering the service, terminal, and
mobility parameters and a user ratio.
A second step consists in cumulating the throughput per cell, per layer, and per technology layer.
Finally, an integration in the ACP global score function with existing quality objectives is performed.
13.7.2 Difference between the RLC Peak Rate and Throughput

Objectives
The RLC Peak Rate objective offers throughput calculation, however the following limitations must be considered:
Calculation is based on a unique service, terminal, and mobility. The objective is then not compatible with a multiservice, terminal, and mobility traffic map; indeed multi-service or multi-terminal means also multi-technologies.
Live user density cannot be used, unless using a weighting map associated with this objective. Note that a such map
often contains only information of relative density of users, which is are statistical and not deterministic data.
The RLC Peak Rate objective is for LTE only and cannot be integrated in a multi-technology objective taking into
consideration multi-technology terminal, services priorities, and potential traffic overflow between layers and
technologies.
In conclusion, the throughput objective benefits from calculation algorithm developed for RLC Peak Rate but its application is
extended in a multi-service, terminal, and mobility context.
13.7.3 Traffic Capture

Traffic capture is the process to extract users from a traffic map based on parameters (services, mobility, terminal, ratio).
For each service, traffic capture is performed for a given compatible technology, within an area defined by a quality parameter
(for example RSRP) and limited by a threshold value (traffic capture conditions). The traffic capture is updated in real time
when any change in the network is triggered b the ACP optimisation (tilt, site addition). For example, the throughput
calculation is impacted when the number of connected users is updated with different radio conditions.
13.7.3.1 Basic Configuration without Atoll Traffic Map

The default terminal and mobility must be defined, this couple being unique. Only services can be filtered out and have userdefined densities.
With a basic configuration, ACP produces a uniform distribution of users within the computation zone and offers the capability
to scale the traffic.
Figure 13.10: Setup properties > Objectives tab > Capacity page (no traffic map defined)
13.7.3.2 Advanced Configuration with Atoll Traffic Map

Service, terminal, and mobility data are extracted from traffic map. The All (Defined by traffic map) mobility option (as shown
in Figure 13.11 on page 903)preserves initial data from Atoll map.
902

AT332_TRR_E0
The Capacity page allows either to use the extracted traffic data or to filter out selected user types (service, terminal,
mobility), and optionally define a scale factor.
Figure 13.11: Setup properties > Objectives tab > Capacity page (with selected traffic map)
13.7.4 Throughput Score Function

The throughput objective performs a throughput maximisation without needing a % target parameter. The calculation is done
on all pixels, each pixel having its best server on layer X (small, macro, and so on) and technology layer (LTE frequency band).
The calculation is repeated for each service and terminal, and then aggregated to produce a global score.
The formula for the absolute score is:
UserRatio TH
N 1 ----------------------------------------------------------------
THmax
UserRatio
Where:
N is the total number of pixel in target zone.
TH corresponds to the RLC Peak Rate Channel Throughput on pixel; calculation uses the RLC Peak Rate ACP quality
parameter for specific service/terminal/mobility. Best server cell is selected through layers according to radio
conditions, priority, CIO,HO margin, and so on.
THmax is a constant value used for normalisation.
UserRatio is a value extracted from traffic map.
The score can be expressed under a similar form than quality objective (score decreasing from 1 to 0).
1
Score = 1 --N
i pixels
TH i -
a i ----------------
THmax
Where:
a i is the user ratio on pixel i.
TH i is the throughput value on pixel i.
THmax is a constant value used for normalisation.
The ACP score function tends to decrease the score by increasing the cumulated throughput, which means that user ratio and
associated throughput tend to increase as well. To reach this goal, the ACP algorithm triggers reconfiguration, site activation,
or any action leading to improve the CINR or user throughput (such as down-tilt or small cell activation).
13.7.5 Impact on the Global Score Function

The throughput score function is added to the ACP global score function, which already includes quality objectives, load
balancing (if defined), and change costs. The weighting of the throughput objective versus other quality objectives (for
example, coverage) can be modified using the Weight parameter in the Throughput page of the setup properties (Objectives
tab > Capacity > Throughput).
903

13.7.6 Optimisation Results

In the Statistics tab of the optimisation results, a section is dedicated to the throughput to display the improvement (% gain)
of the cumulated throughput. When the initial value is equal to 0 (which means that all cells are initially deactivated), 100%
gain is displayed by default.
The throughput value is generated by the score function calculation. The score function aggregates the throughput value only
for the selected layers considered for the objective (for example, the objective can optimise only the small cell layer) or for
the target zone selected for the objective. These settings are configured in the Throughput page of the setup properties
(Objectives tab > Capacity > Throughput).
Other numerical results can be found in the following tabs of the optimisation results:
Throughput tab, which provides throughput per cell, throughput improvement for the target zone used in the
objective, and typical statistics (initial/final number of cells). Cell results can be filtered by technology layer but are
always displayed for the target zone used in the objective.
Sectors tab, which provides cell throughput values.
13.8 EMF Exposure

EMF exposure is defined as the total electromagnetic field measured at a given location. Although the exact limit on the
acceptable level of EMF exposure varies by jurisdiction, it is typically a few Vm.
ACP can analyse and optimise the EMF exposure in the network in order to reduce excessive electromagnetic radiation in
populated areas. Using an internal propagation model specific to EMF exposure, ACP calculates the EMF exposure in two
dimensions (for open areas such as parks or roads) or in three dimensions (for buildings). Additionally, with buildings, you can
choose to measure the exposure only at the front faade, where the EMF exposure will be the greatest.
The internal propagation model evaluates the field strength in V/m using a model based on "free space" propagation, but it
can take diffraction into account when required.
For regulation requirements, a "worst case" mode can be used for the EMF exposure calculation where EMF predictions are
very pessimistic, providing the highest EMF values which would ever likely be seen in the real world. This mode is useful to
ensure that an unacceptable level of EMF exposure is never reached in sensitive areas such as schools, hospitals, etc.
13.8.1 Concepts of ACP EMF Exposure

13.8.1.1 Propagation Classes
The internal propagation model calculates EMF exposure using propagation classes which are retrieved from input files. Each
propagation class is either opaque, meaning that the signal experiences diffraction losses at the edge of the object but does
not go completely through, or transparent, meaning that the signal passes through it (with perhaps some losses) and does not
experience diffraction loss. The propagation classes have the following parameters:
Penetration loss (dB): The loss occurring when the signal enters the object.
Linear loss (dB/m): A linear loss applied for each meter within an object. The loss is applied only after a given number
of meters, specified by the "Linear loss start distance (m)" parameter.
Distribution of evaluation points: Field strength measurements are made on a set of points within an object. The
measurement points can be distributed in either a 3D pattern or in a 2D pattern. For a two-dimensional distribution,
the points can be placed either at the bottom (for example, in a park) or at the top (for example, for a bridge) to better
reflect where people will be.
The following default propagation classes are provided:
Open: The Open propagation class is for areas without obstacles, such as an open area or water. An open area can
also be an elevated area such as a bridge. Such areas are transparent, with no diffraction loss.
Vegetation: The Vegetation propagation class is used for areas covered with vegetation, such as parks. They can be
considered as transparent but with a certain degree of diffraction loss.
Building: The Building propagation class is used for opaque objects such as buildings. The signal experiences some loss
when going through and also suffer from diffraction loss.
It is possible to define new propagation classes, for example to differentiate between similar items with different penetration
loss characteristics (for example, glass buildings, stone buildings, wood constructions, etc.) or for differentiating items on
which EMF evaluation should be done (for example, habitation versus monuments).
Currently, user-defined classes are always of the type "opaque".
904

AT332_TRR_E0
13.8.1.2 Terrain Profile

To measure EMF exposure, ACP does not need any specific terrain modelling, but instead it uses all the geo data available in
the Atoll project:
Geo raster data: Raster data give a grid-based representation of the terrain with a defined resolution. The raster files
needed are DTM (Digital Terrain Model) representing the ground altitude, clutter classes representing the type of
terrain and clutter heights (also called a digital height model) representing individual heights (for example, building
heights).
Geo vector data: Geo vector data model the buildings and their height, in the form of one or several ArcView SHP files
defining numerous polygons.
ACP uses the geo data to create a 3D representation of the terrain in the form of a fine raster of pixels with a default resolution
of 2 meters. For each pixel in this raster representation, both the height and propagation class information are encoded:
For geo vectors, each polygon is associated with a single propagation class and a height. If a geo vector contains more
than one polygon, ACP uses the associated DBF file to map the polygons to propagation classes and heights.
For geo rasters, each clutter class is associated with a single propagation class. The height is obtained from the clutter
height raster file. If no clutter height file is present, the default clutter class height is used.
For areas covered both by vector data and raster data, only the geo vector data are used. Geo raster data are only used for
the areas not covered by geo vector data.
It is recommended to always provide either geo vector data or clutter heights raster data
to have the most accurate EMF exposure prediction.
13.8.1.3 Distribution of Evaluation Points

ACP uses the internal terrain representation to specify where to set evaluation points for EMF exposure evaluation.
It is possible to distribute evaluation points separately on each propagation class and for each terrain entry. For example, you
can distribute evaluation points on one geo vector entry for one subset of polygons, but not on another vector entry. Similarly
you distribute evaluation points on only selected clutter classes.
Evaluation points are distributed either in a 2D or 3D pattern depending on the propagation class:
Open: For the Open class (e.g., bridges, roads, open spaces, etc.) evaluation points are distributed in 2D on the top of
the class height.
Vegetation: For the Vegetation class (e.g., parks, forests, etc.) evaluation points are distributed in 2D at the bottom
of the class height.
Building: For the Building class evaluation points are distributed in 3D. In this case it is also possible to specify the
maximum indoor distance on which to measure EMF exposure, and or to restrict the prediction on the building front
faade.
13.8.1.4 The Contribution of Transmitter Power to EMF Exposure

ACP takes the maximum power transmission on all the carriers and channels used by transmitters in the network into
consideration in its calculations. This ensures that all provided results are for a fully loaded network, thereby giving a worst
case calculation of EMF exposure.
ACP also takes into consideration the antenna radiation pattern by creating a 3D interpolation from the 2D horizontal and
vertical cross-sections of the antenna radiation pattern. The method is similar to the one used elsewhere in Atoll.
13.8.1.5 Worst-case Mode

ACP allows you to consider a worst case scenario, where any opaque propagation class such as Building becomes fully
transparent to electromagnetic waves. The EMF exposure is therefore evaluated as if the area was completely free space.
This mode gives you a pessimistic view of the actual exposure since, in the real world, penetration loss through obstacles as
well as diffraction and reflection around obstacles tends to strongly decrease the signal strength compared to a completely
free space model.
This mode is useful when optimising the networks while ensuring that regulatory limits are never exceeded.
13.8.2 General Workflow

ACP creates a representation of the terrain in 3D. It then distributes a number of evaluation points in this representation
according to the parameters set in the setup.
905

ACP evaluates EMF exposure on each of these evaluation points using an integrated propagation model specially tailored for
the evaluation of strong signals in the vicinity of the antennas. This modelling is based on free space formulas which are
applicable starting at a few meters from the antennas up to a few hundred meters. It takes into account the antenna gain and
attenuation patterns towards each evaluation point, as well as some diffraction and indoor losses.
The overall EMF exposure calculation is obtained by adding the electromagnetic signal level generated by each technology
involved (GSM, UMTS, CDMA, LTE, WiMAX) and considering all carriers and channels used.
To predict the overall EMF exposure, ACP calculates the quadratic sum of all channels in all multi-RAT technologies:
E =
E GSM + E UMTS + E LTE
The following parameters are used in the calculation:

o
BcchPower
GSM:
UMTS:
MaxPowerCelli
TRX
ACP takes into account the maximum transmission powers of cells or transmitters to simulate "worst case" scenarios.
Hence, the following conditions must be met for the most reliable results:
GSM: the number of TRXs must be correctly referenced in the Atoll database, and
UMTS and LTE: all the cells which are to be considered must be present and activated in the Atoll database.
It is highly recommended to activate all technology layers.
In UMTS, even if only one technology layer is activated in the ACP interface for a given
frequency band, the other carriers will also be included in the calculation.
13.8.3 EMF Exposure Calculation

The calculation of EMF exposure is based on the following formula giving the electromagnetic E field (in Volt/meter) at
distance d , in free space far field:
30 EIRP
E = ---------------------d
Where EIRP = P tx G tx and P tx is the transmitted power in Watts and G tx is the antenna gain.
Reason for Using the Free Space Far Field Model
In the free space far field model, the plane wave power density is given (in Watts per square meter) by:
WM
P tx G tx
= -------------2
4d
The power received by a theoretical ideal antenna with an effective aperture of A er and a gain of G rx is given by the Friis
transmission formula for wavelength :
P rx = P
Wm
A er
Where
A = ------ G rx
4
By combining the Friis formula with the expression of E previously defined, the relationship between the EMF exposure level
(in dBVm) and the received signal power level (in dBm) when considering a receiving antenna with 0 dB gain at frequency F
(in MHz) is:
P rx dBm = E dB V m + 42,8 20log 10 F MHz

For frequencies around 1 GHz, the significant EMF exposure level (> 0.1 Vm) corresponds to a received signal level greater
than -37 dBm. Propagation models designed for coverage analysis typically deal with signal levels usually lower than -40 dBm.
They model complex phenomena such as diffraction, reflection, multipath transmission, etc., either deterministically (ray
tracing) or empirically.
906

AT332_TRR_E0
For example, urban empirical models such as Cost-Hata models are typically an extension of the Friis formula where the
2
distance denominator 1 d is replaced with 1 d , with n being a value from 3 to 5.

In addition they only measure the signal level in a 2D horizontal plane, and not in 3D horizontal and vertical planes.
ACP uses a simple propagation model dedicated to cope with the requirements of EMF exposure evaluation. This model is
based on the free space far field formula since "line of sight" exposure can cause significant EMF exposure, resulting in a
potential health hazard (when exposure is above a few tenths of V/m). In non-line of sight situations or far away from the base
station (i.e., beyond a few hundred meters), distance, diffraction, and reflection phenomena decrease the signal strength very
rapidly. The signal strength then becomes smaller than the range of interest (a few tenths of V/m).
The Far-field Restriction
The far field area is usually defined by the area beyond a distance related to antenna size D :
d far field = 2
When the largest dimension D of the antenna is less than the wave length ( ).
2
d far field = 2D
When the largest dimension D of the antenna is greater than the wave length ( ).
The far field starts at around 10 to 20 meters from the antenna.
However, the far field formula usually leads to good field estimates starting at a distance of around 5 meters from the
antenna.
This is in part because antennas are formed of several stacked dipoles (for example 8 to 10) with low coupling between them.
In practice, the total EM signal can be obtained by adding the signal generated by each dipole, each one having a far field
starting at a distance of 2 (or less than 1 meter for a typical frequency of interest).
13.9 Shadowing Margin and Indoor Coverage

ACP supports both indoor coverage and a shadowing margin.
When indoor coverage is used, an additional indoor loss is added to all pixels marked as being indoors. This indoor loss is
clutter-dependent. By default, all pixels are considered as being indoors, but it is possible to specify which clutter classes
should be considered as indoors.
When the shadowing margin has been enabled, a shadowing margin is added to the basic quality measurement. This margin
usually depends on:
Cell coverage probability, according to the log normal distribution function.

The model standard deviation which is clutter-dependent and defined separately for different quality measures.
The shadowing margin is applied in the same way as it is in Atoll coverage predictions, for example, in calculating the macrodiversity gain in UMTS.
13.10 Multi-Storey Optimisation

ACP includes a multi-storey extension where evaluation points are distributed on all floors of buildings defined in a clutter
heights map. This enables ACP to take all floors of the building into account during optimisation.
ACP then proceeds with its optimisation algorithm as usual. All 3-D points participate in the objective optimisation.
13.10.1 Path Loss Calculation and Data Caching

The ACP Multi-storey add-in calculates the path losses from each transmitter to the points distributed in three dimensions by
calling certain methods of Atoll's API on the propagation model (CalculateSubscriber and CalculateGrid).
ACP first detects if path loss matrices created by Atolls MultiStorey add-in are present, and in that case reuses them if
possible. If matrices are not present, it recalculates path loss matrices itself for different heights. If the number of points
distributed for a given height is low, then ACP uses a point-to-point calculation instead of full path loss matrices (i.e., it uses
API CalculateSubscriber instead of CalculateGrid).
ACP stores the path loss attenuation to the multi-storey evaluation points in the ACP storage directory. On further ACP runs,
there is usually little or no need for path loss recalculation, even after modifying parameters such as resolution, etc.
907

ACP then calculates the angles of incidence which are used for the antenna masking method. The calculation method depends
on the propagation model:
Direct calculation at the required height when not using "delegation to the propagation model"
Angle estimation from the original angle of incidence calculated at ground level and taking into account geometrical
considerations, when using "delegation to the propagation model" (for example, when using a propagation model
such as Crosswave).
13.10.2 Pixel Weighting

The total weight associated with an x/y pixel (derived from traffic and zone weighting) is either shared equally among all
vertical evaluation points present at this pixel, or multiplied by the number of vertical points at that pixel.
For example, if a pixel presents a weight of 1 and a total of 5 points at that location (one point at ground level and four
additional points, one every 10 meters), each point either takes a weight of 0.2 (when vertical weight sharing is selected) or
1 (when vertical weight sharing is not selected).
13.10.3 Results
All statistical results provided take into account both 2-D and 3-D points, through the weighting process described earlier.
However all predictions provided by ACP relate only to the ground layer by default.
Viewing detailed results for 3D points is done by creating "Objective Analysis" and "Quality Analysis" ACP predictions. A new
tab becomes available in the predictions properties showing results at different heights.
Three options are provided:
Display at ground level (defined receiver height): Only the prediction values seen at the ground pixels are shown.
Display min. values seen at a given position: When several points are present for a pixel (1 ground level point + one
or several multi-storey points at different heights), the minimum value of those points is shown.
Display values at given storey: Only the points at the given storey are displayed
Example:
Signal Level at ground level
Signal Level at 3rd storey
13.10.4 Notes
ACP distributes multi-storey evaluation points only where clutter heights are present. By default, no point is created
using the default clutter class height when only a clutter class file is present. To enable taking the default clutter class
height into account, you can define the following option in the ACP.ini file:
[ACPCore]
multistorey.onlyDHM=true
The actual heights used for multi-storey evaluation depend on the receiver height defined in Atoll. If a receiver height
of 1.5 m is used, and a vertical step of 2 storeys (with a storey defined in this example as being 3 m), then the actual
heights are 1.5m, 7.5m, 13.5m, etc. This is the same as the process used by the Multi-storey Prediction add-in.
Predictions calculated by ACP might differ slightly from multi-storey predictions due to different methodologies used
by the two tools:
908
ACP uses a mix of radial method (CalculateGrid) for lower storeys and "systematic" method (CalculateSubscribers)
for upper storeys where few evaluation points are present.
Atoll uses in general a "radial" method.

AT332_TRR_E0
13.10.5 Concepts of ACP EMF Exposure

13.10.5.1 Propagation Classes
The internal propagation model calculates EMF exposure using propagation classes which are retrieved from input files. Each
propagation class is either opaque, meaning that the signal experiences diffraction losses at the edge of the object but does
not go completely through, or transparent, meaning that the signal passes through it (with perhaps some losses) and does not
experience diffraction loss. The propagation classes have the following parameters:
Penetration loss (dB): The loss occurring when the signal enters the object.
Linear loss (dB/m): A linear loss applied for each meter within an object. The loss is applied only after a given number
of meters, specified by the "Linear loss start distance (m)" parameter.
Distribution of evaluation points: Field strength measurements are made on a set of points within an object. The
measurement points can be distributed in either a 3D pattern or in a 2D pattern. For a two-dimensional distribution,
the points can be placed either at the bottom (for example, in a park) or at the top (for example, for a bridge) to better
reflect where people will be.
The following default propagation classes are provided:
Open: The Open propagation class is for areas without obstacles, such as an open area or water. An open area can
also be an elevated area such as a bridge. Such areas are transparent, with no diffraction loss.
Vegetation: The Vegetation propagation class is used for areas covered with vegetation, such as parks. They can be
considered as transparent but with a certain degree of diffraction loss.
Building: The Building propagation class is used for opaque objects such as buildings. The signal experiences some loss
when going through and also suffer from diffraction loss.
It is possible to define new propagation classes, for example to differentiate between similar items with different penetration
loss characteristics (for example, glass buildings, stone buildings, wood constructions, etc.) or for differentiating items on
which EMF evaluation should be done (for example, habitation versus monuments).
Currently, user-defined classes are always of the type "opaque".
13.10.5.2 Terrain Profile

To measure EMF exposure, ACP does not need any specific terrain modelling, but instead it uses all the geo data available in
the Atoll project:
Geo raster data: Raster data give a grid-based representation of the terrain with a defined resolution. The raster files
needed are DTM (Digital Terrain Model) representing the ground altitude, clutter classes representing the type of
terrain and clutter heights (also called a digital height model) representing individual heights (for example, building
heights).
Geo vector data: Geo vector data model the buildings and their height, in the form of one or several ArcView SHP files
defining numerous polygons.
ACP uses the geo data to create a 3D representation of the terrain in the form of a fine raster of pixels with a default resolution
of 2 meters. For each pixel in this raster representation, both the height and propagation class information are encoded:
For geo vectors, each polygon is associated with a single propagation class and a height. If a geo vector contains more
than one polygon, ACP uses the associated DBF file to map the polygons to propagation classes and heights.
For geo rasters, each clutter class is associated with a single propagation class. The height is obtained from the clutter
height raster file. If no clutter height file is present, the default clutter class height is used.
For areas covered both by vector data and raster data, only the geo vector data are used. Geo raster data are only used for
the areas not covered by geo vector data.
It is recommended to always provide either geo vector data or clutter heights raster data
to have the most accurate EMF exposure prediction.
13.10.5.3 Distribution of Evaluation Points

ACP uses the internal terrain representation to specify where to set evaluation points for EMF exposure evaluation.
It is possible to distribute evaluation points separately on each propagation class and for each terrain entry. For example, you
can distribute evaluation points on one geo vector entry for one subset of polygons, but not on another vector entry. Similarly
you distribute evaluation points on only selected clutter classes.
909

Evaluation points are distributed either in a 2D or 3D pattern depending on the propagation class:
Open: For the Open class (e.g., bridges, roads, open spaces, etc.) evaluation points are distributed in 2D on the top of
the class height.
Vegetation: For the Vegetation class (e.g., parks, forests, etc.) evaluation points are distributed in 2D at the bottom
of the class height.
Building: For the Building class evaluation points are distributed in 3D. In this case it is also possible to specify the
maximum indoor distance on which to measure EMF exposure, and or to restrict the prediction on the building front
faade.
13.10.5.4 The Contribution of Transmitter Power to EMF Exposure

ACP takes the maximum power transmission on all the carriers and channels used by transmitters in the network into
consideration in its calculations. This ensures that all provided results are for a fully loaded network, thereby giving a worst
case calculation of EMF exposure.
ACP also takes into consideration the antenna radiation pattern by creating a 3D interpolation from the 2D horizontal and
vertical cross-sections of the antenna radiation pattern. The method is similar to the one used elsewhere in Atoll.
13.10.5.5 Worst-case Mode

ACP allows you to consider a worst case scenario, where any opaque propagation class such as Building becomes fully
transparent to electromagnetic waves. The EMF exposure is therefore evaluated as if the area was completely free space.
This mode gives you a pessimistic view of the actual exposure since, in the real world, penetration loss through obstacles as
well as diffraction and reflection around obstacles tends to strongly decrease the signal strength compared to a completely
free space model.
This mode is useful when optimising the networks while ensuring that regulatory limits are never exceeded.
13.10.6 General Workflow

ACP creates a representation of the terrain in 3D. It then distributes a number of evaluation points in this representation
according to the parameters set in the setup.
ACP evaluates EMF exposure on each of these evaluation points using an integrated propagation model specially tailored for
the evaluation of strong signals in the vicinity of the antennas. This modelling is based on free space formulas which are
applicable starting at a few meters from the antennas up to a few hundred meters. It takes into account the antenna gain and
attenuation patterns towards each evaluation point, as well as some diffraction and indoor losses.
The overall EMF exposure calculation is obtained by adding the electromagnetic signal level generated by each technology
involved (GSM, UMTS, CDMA, LTE, WiMAX) and considering all carriers and channels used.
13.11 ACP Software Data Flow

Understanding the ACP software data flow will help understand how the module works and some of its internal constraints.
Here are some of the concepts related to the data flow:
910
Data Model Extraction: When first run (i.e., when the user selects New from the ACP - Automatic Cell Planning
context menu), ACP extracts all relevant information from the current Atoll project and builds its internal data model.
This data model is saved in a "Setup" node and enables each optimisation setup to be reviewed or replayed later. The
data model also contains information identifying the version used to generate it, meaning that the results produced
by a previous release can in general be reloaded or replayed.
Data Model Content: The data model includes all necessary data from the Atoll database; essentially all antennas,
transmitters, sites, cells, and service definitions. The data model does not include raster information such as clutter,
DEM, DHM maps, nor does it contain the path losses matrices. The raster data and path loss matrices are accessed
directly by the core optimisation engine during ACP calculations.
Setup: The ACP Setup dialog box allows you to view and modify the optimisation parameters. The data model is not
accessible using the Setup dialog box.
Results: After an optimisation run, ACP produces a result model which is found under the original setup in an
Optimisation node. Using the Optimisation node, you can view the results and generate analysis maps that can be
displayed directly in Atolls map window. You can also commit the set of recommended changes directly into the Atoll
database.
Optimisation Engine: The optimisation engine is the core algorithm that performs the optimisation on a defined
setup. It works using the extracted internal data model in the Setup node, but also uses direct access to raster and
path loss information.

AT332_TRR_E0
Because ACP uses this internal data model, it is important to understand that:
An optimisation runs on the data model stored in the setup node. If changes are introduced into the Atoll database
later (such as changes to the antennas, cells, site, etc.), these changes are not taken into account in any existing setup
node. The network configuration is essentially frozen in the setup node in the state it was in when the setup was
created. A new setup needs to be created in order for the changes to be taken into account.
Because the path loss information is not stored in the setup node, but is instead accessed directly by the core
optimisation engine, it can happen that there is a mismatch between stored path loss matrices and the data model in
the setup node (for example, after modifying transmitters directly in Atoll).
ACP manages cases of data mismatch by using the concept of a locked setup node. No optimisation can be run on a
locked setup node unless the path loss information is consistent with the internal data model of the setup. Setup
nodes are automatically unlocked when the path loss information and the internal data model once again match.
This behaviour is particularly true when new settings produced by an optimisation run are committed to Atoll. The setup node
is locked after a commit. It will be unlocked if the Atoll project is rolled back to its initial state.
ACP enables you to duplicate an existing setup node while at the same time updating its internal data model to be
consistent with the current state of the Atoll project.
911
Head Office
7 rue des Briquetiers
31700 Blagnac, France
Tel: +33 562 747 210
Fax: +33 562 747 211
AT332_TRR_E0
US Office
China Office
200 South Wacker Drive Suite 3100

Chicago, IL 60606, USA
Tel: +1 312 674 4800
Fax: +1 312 674 4847
Suite 302, 3/F, West Tower, Jiadu Commercial Building,

No. 66 Jianzhong Road, Tianhe Hi-Tech Industrial Zone,
Guangzhou, 510665, P. R. of China
Tel: +86 20 8553 8938 Fax: +86 20 8553 8285
www.forsk.com
October 2016

Atoll 3.3.2 Technical Reference Guide Radio

Uploaded by

Copyright:

Available Formats

Atoll 3.3.2 Technical Reference Guide Radio

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Atoll 3.3.2 Technical Reference Guide Radio

Uploaded by

Copyright:

Available Formats

Version 3.3.

Technical Reference Guide

Atoll 3.3.2 Technical Reference Guide for Radio Networks

Atoll 3.3.2 Technical Reference Guide for Radio Networks

Atoll 3.3.2 Technical Reference Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21

Antennas and Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25

Antenna Pattern Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Power Received From Secondary Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Transmitter Radio Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

Repeaters and Remote Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

Beamforming Smart Antenna Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Grid-of-Beams Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Adaptive Beam Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

Statistical Smart Antenna Gain Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

Radio Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57

Path Loss Calculation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

List of Default Propagation Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Okumura-Hata and Cost-Hata Propagation Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Atoll 3.3.2 Technical Reference Guide for Radio Networks

2016 Forsk. All Rights Reserved.

Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

ITU 529-3 Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

Standard Propagation Model (SPM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65

WLL Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75

ITU-R P.526-5 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76

ITU-R P.370-7 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76

Erceg-Greenstein (SUI) Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78

ITU-R P.1546-2 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80

Sakagami Extended Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84

Free Space Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86

Shadow Fading Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90

Atoll 3.3.2 Technical Reference Guide for Radio Networks

Path Loss Matrices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

GSM GPRS EDGE Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125

Signal Level Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Interference-based DL Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

GPRS/EDGE Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

Atoll 3.3.2 Technical Reference Guide for Radio Networks

2016 Forsk. All Rights Reserved.

Codec Mode Selection and CQI Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Atoll 3.3.2 Technical Reference Guide for Radio Networks

Max C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

Traffic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Network Dimensioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Key Performance Indicators Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

Atoll 3.3.2 Technical Reference Guide for Radio Networks

Blocking Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2016 Forsk. All Rights Reserved.

UMTS HSPA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209

General Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Atoll 3.3.2 Technical Reference Guide for Radio Networks

Circuit Switched Service (i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224

UMTS HSPA Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

Atoll 3.3.2 Technical Reference Guide for Radio Networks

2016 Forsk. All Rights Reserved.

HSUPA Prediction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Primary Scrambling Code Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .