FDD LTE Key Performance Indicator s Description Guide

Learning Objectives
After having learned this training course, you should be able to understand:

Basic knowledge of FDD LTE KPIs Categories of FDD LTE KPIs Formulas of FDD LTE KPIs

Contents

Basic Knowledge of FDD LTE KPIs FDD LTE eNodeB KPIs FDD LTE Network KPIs Other FDD LTE KPIs

References

Counters, Definitions, and Numbering Conventions

For more information, refer to the NetNumen M31(LTE) NE Management System eNod eB Performance Counter Reference at the following URL: http://tsm.zte.com.cn/tsm/FileCenter/File.aspx?Mode=read&FileID=30341574

Formulas

For more information, refer to the NetNumen M31(LTE) NE Management System eNod eB Key Performance Indicator Reference at the following URL: http://tsm.zte.com.cn/tsm/FileCenter/File.aspx?Mode=read&FileID=30341576

Counter Example

Counter
Measurement Object Cell Type Measurement Number 37320 Measurement Type Statistics of RRC Connection Establishment Counter Name Number of Successful mtAccess RRC Connection Establishment Counter Sequence Number C373200000

Example

Counter Structure
Measurement Object
ID: 780-799 Cell Type eNodeB Type eNodeB IP Link Type Cell Pair Type

Measurement Type
ID: 373320-373399 Statistics of RRC connection establishment Statistics of E-RAB Statistics of time Type of a cell pair Other

Counter
ID: C + Measurement Type Number + XXX Number of successful mtAccess RRC connection establishment Number of unsuccessful mtAccess RRC connection establishment (timer timeout) Other

KPI/PI
KPI ID: 310500-310799 RRC connection establishment success rate Other

PI ID: 310500-310799 Paging congestion rate Other

Reporting Procedure

This reporting procedure is described as follows: Step 1 The RNLC subsystem reports data to the OAM using the event-triggered reporting mechanism. Step 2 The CMAC, RNLU, BRS, or OSS subsystem reports data to the OAM every 10 seconds. Step 3 The OAM synchronizes data with the OMC for each measurement object every 15 minutes.

KPI Overview
Description Mobile subscriber (Serviceability + Reliability): An FDD LTE can provide a mobile subscriber with high-quality, reliable, and long-term services. Mobile operator (Serviceability + Reliability + Traffic): An FDD LTE can provide as many mobile subscribers as possible with high-quality, reliable, and longterm services. Dependency System performance: UE + eNodeB + Transport + EPC Applicable environments: Bandwidth configuration, radio access, mobility speed, service type, and so on Categories FDD LTE eNodeB KPIs FDD LTE Network KPIs

Contents

Basic Knowledge of FDD LTE KPIs FDD LTE eNodeB KPIs FDD LTE Network KPIs Other FDD LTE KPIs

FDD LTE eNodeB KPIs

Latency

C-Plane Latency U-Plane Latency

Throughput (Mono-UE)

Peak UE Data Rate Average UE Data Rate Cell Edge UE Data Rate

Cell Capacity (Multi-UE)

Peak Cell Throughput Average Cell Throughput Cell-Edge UE Throughput

U-Plane Latency
Round Trip Time
Combined EPC Application Server eNodeB UE

ping

Measured Round Trip Time

Dependency Radio frequency: Under cell, mid-cell, and cell edge Neighbor cell loading: Unloaded and loaded Scheduling algorithm: Pre-scheduled and non-scheduled Ping size: 32 B, 1000 B, and 1500 B

C-Plane Latency
State Transition Time

Signaling Procedure

Throughput (Mono-UE)
Peak UE Data Rate

Transm ission

E.g. Gateways, Routers, Firewalls

Term inal

BS D o w n lin ko rU p lin kD a t aS t r e a m

Application Server at NGMN network edge

Assumption Loading: Single cell, unloaded UE speed: Stationary mode Location: Under cell (good RF conditions) MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO

Dependency Operating bandwidth UE category UE UL/DL RB limitation

Throughput (Mono-UE)
Average UE Data Rate

Assumption MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO Loading: 1 UE in the serving cell, neighbor cell loaded UE speed: Stationary mode Location: Uniformly-distributed locations over the signal quality range

Dependency Operating bandwidth UE category UE UL/DL RB limitation UE distribution in each signal range

Throughput (Mono-UE)
Cell-Edge UE Data Rate

Assumption MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO Loading: 1 UE in the serving cell, neighbor cell loaded UE speed: Stationary mode Location: Cell edge

Dependency Operating bandwidth UE category UE UL/DL RB limitation UE SINR at the cell edge

Throughput (Multi-UE)
Peak Cell Throughput

Assumption MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO Loading: Multiple cells, loaded (70%) UE Speed: Stationary mode Location: Good RF conditions

Dependency UE category: TBS limitation, 64QAM in UL UE UL/DL RB limitation Environment: Dense urban, urban, suburban, and rural Scheduling algorithm

Throughput (Multi-UE)
Average Cell Throughput Typical UE Distribution Model

Assumption MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO Loading: Multiple cells, loaded UE speed: Stationary Mode Location: Uniformly-distributed over the signal quality range

Dependency UE category (TBS limitation, 64QAM in UL) UE UL/DL RB limitation UE distribution in each signal range, scheduling algorithm

Throughput (Multi-UE)
Cell-Edge UE Throughput

Assumption MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO Loading: Multiple cells, loaded (70%) UE speed: Stationary mode Location: Cell edge

Dependency UE category (TBS limitation, 64QAM in UL) UE UL/DL RB limitation Environment: Dense urban, urban, suburban, and rural Scheduling algorithm UE SINR at the cell edge

Contents

Basic Knowledge of FDD LTE KPIs FDD LTE eNodeB KPIs FDD LTE Network KPIs Other FDD LTE KPIs

FDD LTE Network KPIs

Accessibility RRC Establishment Success Rate E-RAB Setup Success Rate Retainability E-RAB Drop Rate Mobility Handover Success Rate Availability Cell Availability Integrity Packet Loss Rate DL PDCP SDU Latency

Accessibility
Theoretical limit: 100% Ideal value in a commercial network: > 98%
Initial E-RAB Accessibility Accessibility (Category )

Attach Success Rate

Detach Success Rate

KPI

RRC Connection Setup Success Rate

S1-SIG Establish Success Rate

Initial E-RAB Setup Success Rate

RRC Re-Establishment Success Rate

Added E -RAB Accessibility

Paging Success Rate

Air Interface

RRC Connection Setup Success Rate

S1-SIG Establish Success Rate

UE Context Setup Success Rate

Added E -RAB Setup Success Rate

Contention -Based PRACH Performance

Contention -Free PRACH Performance

KPI

E-RAB Setup Success Rate

Call Setup Success Rate, Call Barring Rate

E-RAB Block Rate (per QCI )

RRC Establishment Success Rate

Description

This KPI shows the probability for a subscriber to be provided with an RRC connection upon request.

Signaling Procedure UE

EUTRAN

RRCConnectionRequest

RRCConnectionSetup

RRCConnectionSetupComplete

Formula

RRC Establishment Success Rate = Number of successful RRC connection establishment / (Number of successful RRC connection establishment + Number of failed RRC connection establishment) * 100%

E-RAB Setup Success Rate

Description

This KPI shows the probability for a subscriber to be provided with an E-RAB request including initial and added context setup procedures.

Signaling Procedure eNB

INITIAL CONTEXT SETUP REQUEST

MME

eNB

MME

1 2

E-RAB SETUP REQUEST

3 4

INITIAL CONTEXT SETUP RESPONSE

E-RAB SETUP RESPONSE

Formula

E-RAB Setup Success Rate = (Number of successful initial E-RAB + Number of successful added E-RAB establishment) / (Number of successful initial E-RAB establishment + Number of failed initial E-R AB establishment + Number of successful added E-RAB establishment + Number of failed added E-RAB establishment) * 100%

Retainability
Theoretical limit: 0% Ideal value in a commercial network: < 2%

Retainability (Category)

KPI
UE E-RAB Retainability
How Often

UE E-RAB Retainability Per QCI

RRC Drop Rate

E-RAB Drop Rate (Per QCI )

Percentage

Active E-RAB Drop Rate

E-RAB Duration Time per QCI Drop

E-RAB UL Data Volume Per QCI Drop

QoS Before Drop

E-RAB DL Data Volume Per QCI Drop

E-RAB Drop Rate

Description

This KPI shows the probability for an a subscriber to loss the E-RAB, such as an event b eing released by the eNodeB due to overload control.

Signaling Procedure E-RAB Released by the eNodeB))

eNB MME

E-RAB RELEASE INDICATION

Formula

E-RAB Drop Rate = Number of Abnormally Released E-RAB / Number of Successfully Established E-RAB * 100%

Mobility
Successful Handover Categories

Intra-Freq Intra-eNodeB

Intra-Freq Inter-eNodeB

Inter-Freq Inter-eNodeB

Inter-Freq Intra-eNodeB

LTE to UMTS

UMTS to LTE

LTE to GSM

GSM to LTE

Inter- RAT HO Success Rate X2 Based HO S 1 Based HO

All Incoming HO Per Cell Contention Based HO Contention Free HO

All Outgoing HO Per Cell

Intra-RAT HO Success Rate

All Incoming HO Per Cell Pair

All Outgoing HO Per Cell Pair

Troubleshooting

Mobility
Handover Handover Preparation Success Rate Theoretical limit: 100% Ideal value in a commercial network 99% (Step 4 Step 6)

Handover Execution Success Rate Theoretical limit: 100% Ideal value in a commercial network: 98% HO In: Step 6 Step 11 HO Out: Step 7 Step 17

Mobility Success Rate Theoretical limit: 100% Ideal value in a commercial network: 97%

MobilitySR= HO. Pr eparationS R HO.ExecutionS R

Handover Preparation and Execution

Handover Outgoing handover preparation Outgoing handover execution Start Point The source eNodeB decides to perform a handover. The source eNodeB sends the RRC Connection Reconfiguration message to the UE. The destination eNodeB receives the Handover Request message from the source eNodeB. The destination eNodeB receives the RRC Connection Reconfiguration Complete message from the UE. Stop Point The source eNodeB sends the RRC Connection Reconfiguration message to the UE. The source eNodeB receives the UE Context Release message from the destination eNodeB.

Incoming handover preparation

The destination eNodeB returns the Handover Response message to the source eNodeB.

Incoming handover execution

The destination eNodeB sends the UE Context Release message to the source eNodeB.

Cell Handover and Cell Pair Handover

The outgoing handover preparation success rate for each handover pair measures the hando ver preparation from the serving cell to a certain neighbor cell. Here are two typical examples:

In the event of the handover from cell A to cell B, the Intra-eNodeB Intra-freq Outgoing Handover Preparation Success Rate per Cell pair is measured in cell A. In the event of the handover from cell B to cell A, the Intra-eNodeB Intra-freq Outgoing Handover Preparation Success Rate per Cell pair is measured in cell B.

The outgoing handover success rate contains all outgoing handover preparation successes fr om the serving cell to all neighbor cells. The measurement of these two KPIs can facilitate us in sifting two cells that suffer most hando ver preparation failures, performing neighbor cell optimization, and even deleting unusable nei ghbor cells.

Intra-eNodeB Handover Success Rate

Description

The intra-eNodeB handover success rate measures the service continuity when a subs criber is on the move. This KPI is perceptible to the subscribers, depending on system hand over processing capabilities and network planning. Formulas

Intra-frequency handover success rate = Number of intra-frequency handover succes ses / Number of intra-frequency handover requests * 100% Inter-frequency handover success rate = Number of inter-frequency handover succes ses / Number of inter-frequency handover requests * 100%

Intra-eNodeB Handover Success Rate

Note: If the eNodeB receives the RRC Connection Reconfiguration Complete message in step 4, it i ndicates that the handover is successful. If the eNodeB receives the RRC Connection Reestablishment Request message in step 6, it i ndicates that the handover is unsuccessful.

Inter-eNodeB X2-Interface Handover Success Rate

Description The inter-eNodeB X2-interface handover success rate measures the handover successes when the UE moves between the eNodeBs over the X2 interface. This KPI is perceptible to the subscribers being o n the move, depending on system handover processing capabilities and network planning.

Formulas

Outgoing intra-frequency X2-interface handover success rate = Number of outgoing intra-freque ncy X2-interface handover successes / Number of outgoing intra-frequency X2-interface handov er attempts (serving cell) * 100% Incoming intra-frequency X2-interface handover success rate = Number of incoming intra-freque ncy X2-interface handover successes / Number of incoming intra-frequency X2-interface handov er attempts (serving cell) * 100% Outgoing inter-frequency X2-interface handover success rate = Number of outgoing inter-freque ncy X2-interface handover successes / Number of outgoing inter-frequency X2-interface handov er attempts (serving cell) * 100% Incoming inter-frequency X2-interface handover success rate = Number of incoming inter-freque ncy X2-interface handover successes / Number of incoming inter-frequency X2-interface handov er attempts (serving cell) * 100%

Inter-eNodeB X2-Interface Handover Success Rate

Inter-eNodeB S1-Interface Handover Success Rate

Description When the eNodeB decides to perform a handover according to the UE measurement report and me anwhile the destination cell is not connected to the eNodeB through the X2 interface, the inter-eNod eB S1-interface handover success rate measures the S1-interface handover performed through the EPC. This KPI is perceptible to the subscribers being on the move, depending on system handover processing capabilities and network planning. Formulas

Outgoing intra-frequency S1-interface handover success rate = Number of outgoing intra-freque ncy S1-interface handover successes / Number of outgoing intra-frequency S1-interface hando ver attempts (serving cell) * 100% Incoming intra-frequency S1-interface handover success rate = Number of incoming intra-frequ ency S1-interface handover successes / Number of incoming intra-frequency S1-interface hand over attempts (serving cell) * 100% Outgoing inter-frequency S1-interface handover success rate = Number of outgoing inter-frequ ency S1-interface handover successes / Number of outgoing inter-frequency S1-interface hand over attempts (serving cell) * 100% Incoming inter-frequency S1-interface handover success rate = Number of incoming inter-frequ ency S1-interface handover successes / Number of incoming inter-frequency S1-interface hand over attempts (serving cell) * 100%

Inter-eNodeB S1-Interface Handover Success Rate

Inter-System Handover Success Rate

The inter-system handover success rate consists of both incoming and outgoing handover between the LTE network and t he CDMA network, between the LTE network and the UMTS network, between the LTE network and the GSM network. Thi s example shows a handover from the CDMA network to the LTE network.

Availability
Availability Cell availability Theoretical limit: 100% Ideal value in a commercial network: > 99.995%

Important and Demanding

measuremen t_period- RRU.CellUnavailableT ime.[cause ]
cause

CellAvailability =
Dependency Software + Hardware

measuremen t _ period

100

Unavailable Time = Unplanned downtime only (excluding planned downtime) Physical meaning 365 * 24 * 60 * (1-99.995%) = 26.28 min cell out-of-service time

Cell Availability

Description The cell availability measures the ratio of in-service time to measurement granularity time. The inservice time indicates the time interval between cell establishment and cell deletion. By counting the cell in-service time, this KPI forms a foundation for analyzing system failures and measuring system stability.

Signaling Procedure

Formulas Cell Availability = In-Service Time / Measurement Granularity Time Cell Availability = C373230700 / Measurement Granularity Time * 100%

Integrity
UL/DL Packet Loss Rate DL PDCP SDU Latency
QCI Resour ce Type Priority Packet Latency Budget (NOTE 1) Packet Error Loss Rate (NOTE 2) 10-2 Example Services

1 (NOTE 3) 2 (NOTE 3) 3 (NOTE 3) 4 (NOTE 3) 5 (NOTE 3) 6 (NOTE 4) 7 (NOTE 3) 8 (NOTE 5) NonGBR

100 ms

Conversational Voice

4 GBR 3

150 ms 50 ms

10-3 10-3

Conversational Video (Live Streaming) Real Time Gaming

300 ms

10-6

Non-Conversational Video (Buffered Streaming)

100 ms

10-6

IMS Signalling

300 ms

10-6

Video (Buffered Streaming) TCP-based (e.g., www, e-mail, chat, ftp, p2p file sharing, progressive video, etc.) Voice, Video (Live Streaming) Interactive Gaming Video (Buffered Streaming) TCP-based (e.g., www, e-mail, chat, ftp, p2p file File sharing, progressive video, etc.

100 ms

10-3

8 300 ms 10-6

9 (NOTE 6)

Downlink PDCP SDU Latency

Description This KPI indicates average downlink PDCP SDU latency based on the QCI type, from the time when a PDCP SDU reaches the eNodeB, to the time when the UE receives this PDCP SDU, th at is to say, all fragments of this PDCP SDU receives a successful HARQ response.

Signaling Procedure

Formula Average Downlink PDCP SDU Latency = Total Latency of All PDCP SDUs / Number of All PDCP SDUs

Downlink IP Packet Latency

Description This KPI indicates average downlink IP packet latency, from the time when the eNodeB receives the IP packet through the S1 or X2 interface, to the time when the first fragment of this IP packet is transmitted by the eNodeB through the air interface. It measures the time interval at which the service is processed by the eNodeB, which forms a strong foundation for network optimization.

Signaling Procedure

Formula Average Downlink IP Packet Latency (QCI 1 9) = Total Downlink IP Packet Latency (QCI 1 9 ) / N umber of All PDCP SDUs (Downlink QCI 1 9)

Uplink/Downlink Packet Loss Rate

Description
This KPI measures the ratio of discarded PDCP SDUs to received PDCP SDUs due to the timeout of the TimeDisCard timer, when no, partial, or all fragments are transmitted through the eNodeB or air interface. It should be noted that PDCP SDUs vary from QCI to QCI (1 9), from uplink to downlink.

Counters
Number of All PDCP SDUs Uplink: When the PDCP layer of the eNodeB receives the PDCP SDU from the UE, this counter is incremented by 1. Downlink: When the PDCP layer of the eNodeB sends the PDCP SDU to the RLC layer, this counter is incremented by 1. Number of Discarded PDCP SDUs Uplink: When the PDCP layer of the eNodeB receives the PDCP SDU from the UE, the SN is not consecutive. Downlink: When the PDCP layer of the eNodeB sends the PDCP SDU to the RLC layer, the SN is not consecutive.

Formula
Air Interface or eNodeB Packet Loss Rate = Number of Discarded Packets over the Air Interface or eNodeB / Number of All Packets over the Air Interface or eNodeB

Contents

Basic Knowledge of FDD LTE KPIs FDD LTE eNodeB KPIs FDD LTE Network KPIs Other FDD LTE KPIs

Traffic Type - Number of RRC Connections

Description This KPI counts the number of RRC connections, which is used to measure the UE access to the system for call hold. It involves the following two counters:

Average number of RRC connections Maximum number of RRC connections.

Signaling Procedure

When the eNodeB receives the RRC Establishment Complete message, the number of RRC connections is in cremented by 1. When the eNodeB triggers the RRC establishment successfully due to the handover, the number of RRC con nections is incremented by 1. When the eNodeB releases the RRC connection, the number of RRC connections is decremented by 1. When the eNodeB reestablishes the RRC connection in another cell, the number of RRC connections is decre mented by 1. When the eNodeB reestablishes the RRC connection back to the serving cell, the number of RRC connection s is incremented by 1.

Formulas

Maximum Number of RRC Connections (Sampling Counter) Average Number of RRC Connections (Sampling Counter)

Traffic Type Average Number of QCI-Based UEs

Description This KPI measures the average number of E-RAB connections, which evaluates network traffic. It s hould be noted that this KPI varies from QCI to QCI.

Signaling Procedure (Sampling Counter)

Initial E-RAB establishment success Added E-RAB establishment success Incoming E-RAB handover success Incoming E-RAB modification success (changing the new QCI to the old QCI) Outgoing E-RAB handover success Outgoing E-RAB modification success (changing the old QCI to the new QCI) E-RAB release

Formula

This KPI is counted by averaging the measured values of all sampling points within a specific meas urement cycle.

Traffic Type Number of Activated QCI-Based UEs

Description This KPI counts the number of UEs in the cache during a specified time period. It can be m easured based on a specific QCI, uplink or downlink, average number or maximum number .

This KPI indicates the number of in-service UEs in the system, which forms a foundation fo r evaluating system capacity and capabilities.

Signaling Procedure N/A

Formula

This KPI is counted when any data is present in the E-RAB cache of QCI i every sampling cycle (100 ms), as defined in the 3GPP TS.

Resource Allocation
Cell & System (Category ) Resources

Total UL PRB Usage

PDCCH Usage

UL -C PRB Usage

UL -U PRB Usage

PRACH Usage

PRACH Propagation Delay

PRACH Message Load

UL PRB Usage per QCI

Paging Congestion Rate

Total D L PRB Usage

Avg CPU Load

Avg DSP Load

DL -C PRB Usage

DL -U PRB Usage

Avg DL Transmission Power

DL PRB Usage per QCI

Avg UL Interference Per PRB