Physical Layer Implementation of a class of

ZigBee Baseband Transceiver using FPGA

Thesis Submitted to the

National Institute of Technology Rourkela
for award of the degree

of

Master of Technology (Research)

by

Bijaya Kumar Muni

(Roll No: 610EC602)

Under the guidance of

Prof. Sarat Kumar Patra

National Institute of Technology Rourkela

Department of Electronics and Communication Engineering

National Institute of Technology Rourkela
Rourkela, Odisha, 769008, India
January 2013
 Dept. of Electronics and Communication Engineering
National Institute of Technology Rourkela
Rourkela-769 008, Odisha, India.

January 29, 2013

Certificate

This is to certify that the work in the thesis entitled “Physical Layer Implementation
of a class of ZigBee Baseband Transceiver using FPGA” by Bijaya Kumar
Muni is a record of an original research work carried out under my supervision and guid-
ance in partial fulfillment of the requirements for the award of the degree of Master of
Technolgy (R) in Electronics and Communication Engineering. Neither this thesis nor any
part of it has been submitted for any degree or academic award elsewhere.

Dr. Sarat Kumar Patra

Professor
To my Parents

iii
Acknowledgements

It has been a great experience to work under the esteemed supervision of Prof. Sarat

Kumar Patra. I am very much privileged to have him as my mentor. I would like to

thank him from the bottom of my heart for his involvement, guidance, most importantly

his support and encouragement throughout the project work.

In the same breath, I would also like to thank Prof. D. P. Acharya, Prof. Samit Ari

and Prof B. D. Subudhi, for their suggestions and comments during the project work. I am

indebted to prof. K K Mahapatra for his involvement and helpful discussions throughout

the work. I am thankful to Prof. S. Meher, Head of the Dept., for his suggestion and

cooperation.

I thank all faculty members of Dept. of Electronics & Communication Engineering,

National Institute of Technology Rourkela, Rourkela who have been a great source of

inspiration for my research work.

I would also like to acknowledge ANURAG Lab, Defence Research Development Or-

ganization (DRDO), Hyderbad, for the financial support which lent completeness of my

Master research. I would like to thank Scientist Mr. M.Madhav Kumar of ANURAG Lab

for his support and guidance during research work. I would also like to thank Scientist

Vijay Kumar Agrawal for his cooperation.

My sincere thanks to all "Mobile Communication" group fellow members for their knowl-

edge sharing and filling enthusiasm in me to make my project work enjoyable. It is my

immense pleasure to thank my parents and family members for their constant moral sup-

port and inspiration.

iv
Abstract

ZigBee and IEEE 802.15.4 standard for wireless technology, developed in 2003 were de-

signed for interconnection of data communication using low data rate, low power and low

complexity short range communication in a wireless personal area network (WPAN). Later

in 2006, it was enhanced for market applicability to remove ambiguities in implementation

for low data rate and short range wireless networks with high battery life. This technology

supports cost effective, low power, wireless network monitoring and control products based

on open global standard.

This thesis presents a FPGA implementation of Baseband physical layer for ZigBee. It

presents the designs, implementation, verification and validation.

The ZigBee baseband transceiver proposed in this thesis is based on IEEE 802.15.4

where the transceiver uses OQPSK modulation. DS spread spectrum and half sine pulse

shaping is used for coding and baseband processing respectively. The transceiver is initially

simulated in matlab software using Simulink and next it was simulated in verilog HDL by

the mentor graphics modelsim simulator. Subsequently the baseband transceiver system

was realized on Virtex 5 FPGA using ISE design environment.

Further a new form of baseband transceiver was designed using PN sequence gener-

ated by Residue number system (RNS). The performance of the transceiver using RNS

system was first analyzed through matlab simulation. Following this the transceiver was

implemented on Virtex 5 FPGA in ISE design environment.

v
Contents

Certificate i

Acknowledgements iv

Abstract v

List of Tables x

List of Figures xii

Abbreviations xv

1 Introduction 1

1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.2 Background of Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.3 Objective of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.4 Thesis Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 ZigBee for IEEE 802.15.4 Review 7

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.2 ZigBee Data Communication . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.2.1 ZigBee Network Device . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.2.2 Topologies in ZigBee . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3 IEEE 802.15.4 Physical Layer Specifications . . . . . . . . . . . . . . . . . . 11

2.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.3.2 Data rate and frequency of operation . . . . . . . . . . . . . . . . . . 14

2.3.3 Channel Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

vi
 2.3.4 ZigBee protocol Architecture . . . . . . . . . . . . . . . . . . . . . . 16

2.3.5 PHY Layer Services . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.3.6 MAC Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.3.7 Data transfer between devices . . . . . . . . . . . . . . . . . . . . . . 22

2.3.8 Data Transfer Methods . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.4 ZigBee 2.4 GHz PHY Specifications . . . . . . . . . . . . . . . . . . . . . . 26

2.4.1 Modulation and Spreading . . . . . . . . . . . . . . . . . . . . . . . 26

2.4.2 Reference Modulator . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

2.5 Summary and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3 FPGA Implementation of 2.4 GHz ZigBee Transceiver 30

3.1 ZigBee Transmitter in Simulink . . . . . . . . . . . . . . . . . . . . . . . . . 30

3.1.1 Symbol to Chip Mapping . . . . . . . . . . . . . . . . . . . . . . . . 32

3.1.2 Serial to Parallel Conversion . . . . . . . . . . . . . . . . . . . . . . 34

3.1.3 Half sine pulse shaping . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3.1.4 RF Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

3.2 ZigBee Receiver in Simulink . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.2.1 Receiver Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

3.2.2 Demodulator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

3.2.3 Half sine pulse shape . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

3.2.4 Low Pass Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3.2.5 Sample and Hold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

3.2.6 Parallel to serial conversion . . . . . . . . . . . . . . . . . . . . . . . 42

3.2.7 Despreading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

3.3 Baseband Transmitter Design for FPGA . . . . . . . . . . . . . . . . . . . . 45

3.3.1 Input data stream . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.3.2 Symbol to chip conversion . . . . . . . . . . . . . . . . . . . . . . . . 47

vii
 3.3.3 Serial to Parallel Conversion . . . . . . . . . . . . . . . . . . . . . . 48

3.3.4 ZigBee transmitter Pulse Shaping . . . . . . . . . . . . . . . . . . . 50

3.4 Baseband Receiver Design for FPGA . . . . . . . . . . . . . . . . . . . . . . 52

3.4.1 Pulse shape detection . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.4.2 Parallel data(even and odd stream) . . . . . . . . . . . . . . . . . . 53

3.4.3 Parallel to serial conversion . . . . . . . . . . . . . . . . . . . . . . . 54

3.4.4 Chip to symbol mapping at receiver . . . . . . . . . . . . . . . . . . 56

3.5 FPGA Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

3.5.1 VIRTEX 5 FPGA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

3.5.2 Digital Clock Manager(DCM) . . . . . . . . . . . . . . . . . . . . . . 61

3.5.3 Implementation on VIRTEX 5 FPGA . . . . . . . . . . . . . . . . . 61

3.5.4 Synthesis Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

3.6 Summary and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

4 ZigBee Transceiver Design with RNS Based PN sequence 69

4.1 Residue Number System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

4.1.1 Chinese Remainder Theorem . . . . . . . . . . . . . . . . . . . . . . 70

4.1.2 Arithmetic in RNS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

4.1.3 Structure of RNS Conversion . . . . . . . . . . . . . . . . . . . . . . 73

4.2 RNS Based PN Sequence for ZigBee Spreading and Despreading . . . . . . 74

4.2.1 RNS Generated PN sequence . . . . . . . . . . . . . . . . . . . . . . 75

4.3 Performance of RNS based PN sequence in ZigBee Transceiver . . . . . . . 75

4.4 RNS Based Baseband Transceiver in FPGA . . . . . . . . . . . . . . . . . . 77

4.4.1 Simulation Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

4.5 Implementation on VIRTEX 5 FPGA . . . . . . . . . . . . . . . . . . . . . 82

4.5.1 Synthesis Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

4.6 Summary and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

viii
5 Conclusions and Future Scope of Work 88

5.1 ZigBee Transceiver Design using FPGA . . . . . . . . . . . . . . . . . . . . 88

5.1.1 Design Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

5.2 Limitation of work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

5.3 Future Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Bibliography 91

Publication 94

Biography 96

ix
List of Tables

2.1 ZigBee Terminology in contrast to IEEE 802.15.4 . . . . . . . . . . . . . . . 8

2.2 Data rate and frequency of operation . . . . . . . . . . . . . . . . . . . . . . 15

2.3 IEEE 802.15.4 Channel assignment . . . . . . . . . . . . . . . . . . . . . . . 15

2.4 Symbol to Chip Mapping for 2.4 GHz Band . . . . . . . . . . . . . . . . . . 28

3.1 I/O Description Datain Module . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.2 I/O Description pnsequence module . . . . . . . . . . . . . . . . . . . . . . 48

3.3 I/O Description of Serial to Parallel . . . . . . . . . . . . . . . . . . . . . . 49

3.4 Samples for Pulse shaping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

3.5 I/O Description even pulse shape . . . . . . . . . . . . . . . . . . . . . . . . 51

3.6 I/O Description Odd pulse shape . . . . . . . . . . . . . . . . . . . . . . . . 52

3.7 I/O Description of Evenstream_out . . . . . . . . . . . . . . . . . . . . . . 54

3.8 I/O Description of Oddstream_out . . . . . . . . . . . . . . . . . . . . . . . 55

3.9 I/O Description of IQaddition . . . . . . . . . . . . . . . . . . . . . . . . . . 57

3.10 I/O Description of Decoding . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

3.11 Hardware Resources of VIRTEX 5 LX110T . . . . . . . . . . . . . . . . . . 61

3.12 Clock frequency for VIRTEX 5 LX110T . . . . . . . . . . . . . . . . . . . . 61

3.13 Synthesis options summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

3.14 Device Utilization Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

3.15 Advance HDL Synthesis Report . . . . . . . . . . . . . . . . . . . . . . . . . 65

3.16 Hierarchical power analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

x
4.1 Residue Representation in RNS . . . . . . . . . . . . . . . . . . . . . . . . . 72

4.2 Symbol to Chip Mapping using RNS code for 2.4 GHz ZigBee Transceiver . 76

4.3 Synthesis options summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

4.4 Device Utilization Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

4.5 Advance HDL Synthesis Report . . . . . . . . . . . . . . . . . . . . . . . . . 86

xi
List of Figures

1.1 Short Range Wireless network Comparison . . . . . . . . . . . . . . . . . . 2

2.1 Star and Peer to peer Topology . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.2 ZigBee Cluster Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2.3 ZigBee Protocol Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.4 PHY and MAC Reference Model . . . . . . . . . . . . . . . . . . . . . . . . 18

2.5 Data Transfer Between Devices . . . . . . . . . . . . . . . . . . . . . . . . . 22

2.6 Data Transfer From Coordinator to Network Device . . . . . . . . . . . . . 24

2.7 Data Transfer From Network Device to Coordinator . . . . . . . . . . . . . 25

2.8 Modulation and spreading . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

2.9 Symbol to chip mapping process . . . . . . . . . . . . . . . . . . . . . . . . 27

2.10 O-QPSK chip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2.11 O-QPSK Pulshaping for I and Q Channel . . . . . . . . . . . . . . . . . . . 29

3.1 Block Diagram of ZigBee transmitter . . . . . . . . . . . . . . . . . . . . . . 31

3.2 Simulink Model of ZigBee Transmitter . . . . . . . . . . . . . . . . . . . . . 31

3.3 Input data and PN sequence . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.4 DSSS output in Simulink . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.5 Serial to Parallel Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3.6 Serial to Parallel conversion in Simulink . . . . . . . . . . . . . . . . . . . . 35

3.7 I-phase and Q-phase pulse shaping in Simulink . . . . . . . . . . . . . . . . 36

3.8 Modulation output waveform . . . . . . . . . . . . . . . . . . . . . . . . . . 37

xii
3.9 Block Diagram of ZigBee Receiver . . . . . . . . . . . . . . . . . . . . . . . 38

3.10 Simulink Model of ZigBee Receiver . . . . . . . . . . . . . . . . . . . . . . . 38

3.11 Demodulation in Simulink . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

3.12 I-phase and Q-phase pulse shaping in Simulink . . . . . . . . . . . . . . . . 41

3.13 I-phase and Q-phase passed through low pass filter . . . . . . . . . . . . . . 42

3.14 I-phase and Q-phase passed through sampled and hold . . . . . . . . . . . . 43

3.15 I-phase and Q-phase passed through comparator . . . . . . . . . . . . . . . 43

3.16 Despreading at receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

3.17 ZigBee baseband Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.18 Bit to Symbol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.19 Symbol to chip and DSSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

3.20 I-phase Q-phase separation . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.21 Q-phase pulse shaping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

3.22 I-phase pulse shaping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

3.23 ZigBee Baseband Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.24 Q-phase signal at Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

3.25 I-phase signal at Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3.26 Parallel to serial conversion at Receiver . . . . . . . . . . . . . . . . . . . . 56

3.27 I-phase and Q-phase Combination at Receiver . . . . . . . . . . . . . . . . . 56

3.28 Decoding at Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

3.29 Chip mapping at Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

3.30 Received data bit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

3.31 XILINX VIRTEX 5 FPGA Board with XC5VLX110T . . . . . . . . . . . . 60

3.32 Digital Clock Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

3.33 Digital Clock Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

3.34 ChipScope Pro Internal signals . . . . . . . . . . . . . . . . . . . . . . . . . 65

xiii
3.35 Top level RTL Schematic of ZigBee Baseband Transceiver . . . . . . . . . . 66

4.1 Basic Structure of RNS Processor . . . . . . . . . . . . . . . . . . . . . . . . 73

4.2 Performance in AWGN Channel . . . . . . . . . . . . . . . . . . . . . . . . . 77

4.3 Performance in Rayleigh fadding Channel . . . . . . . . . . . . . . . . . . . 78

4.4 DSSS Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

4.5 Bit to Symbol to Chip Conversion for RNS . . . . . . . . . . . . . . . . . . 80

4.6 Pulse shaping of RNS based Transceiver . . . . . . . . . . . . . . . . . . . . 81

4.7 DSSS signal at Receiver for RNS based Transceiver . . . . . . . . . . . . . . 82

4.8 Decoding of RNS Based Transceiver . . . . . . . . . . . . . . . . . . . . . . 83

4.9 Decoder for RNS based transceiver . . . . . . . . . . . . . . . . . . . . . . . 83

4.10 Chipscope Pro Internal signal for RNS based Baseband Receiver . . . . . . 86

xiv
Abbreviations

AWGN Additive white gaussian noise

BER Bit error rate

CRT Chinese remainder theorem

CMT Clock manager tiles

DCM Digital clock manager

DSCDMA Direct-sequence code division multiple-access

DSSS Direct sequence spread spectrum

DU Data unit

ED Energy detection

FFD Full function device

FPGA Field programmable gate array

GTS Guaranteed time slot

I-phase In Phase

I/O Input/Output

MAC Medium access control layer

MC/DSCDMA Multi carrier Direct-sequence code division dynamic multiple-access

OQPSK Offset quadrature phase shift keying

OSI Open system interconnect

PHY Physical layer

PN Pseudo noise

PPDU Protocol Data Unit

Q-phase Quadrature Phase

xv
 RNS Residue number system

RFD Reduced function device

SSP Security service provider

WPAN Wireless personal area network

WLAN Wireless local area network

xvi
 1
Introduction

Current millennium has seen explosive growth in wireless communication. The short range

wireless communication has been used for accessing networks and services without cables,

which is a fast-growing technology for providing flexibility and mobility. Major benefits of

the technology includes the dynamic network formation, low cost and easy of deployment

[1, 2]. There are different protocol standards presented in Figure 1.1 used for the short range

wireless communication namely the Bluetooth [3], ZigBee [4] and Wi-Fi [5]. Among these

standards ZigBee over IEEE 802.15.4 protocol can meet a wider variety of real industrial

needs due to its long-term battery operation, wider useful range and reliability of the mesh

networking architecture [6].

ZigBee and Wireless Personal Area Networks (WPANs) are used to transfer short mes-

sages like commands of information over a short distance. Like WPAN, ZgBee needs little

infrastructure for data exchange. This feature allows the ZigBee alliance group for small,

power-efficient and inexpensive solutions to implement in a wide range of wireless devices.

So ZigBee alliance group along with IEEE 802.15.4 standard forms the ZigBee protocol

networking layers for low data rate short range wireless communication. This standard

provides Physical layer(PHY) and medium access control (MAC) specifications for low

data rate wireless connection for fixed, portable and moving devices with limited battery

1
1 CHAPTER 1. INTRODUCTION

Power Consumption
Complexity
Cost IEEE
802.11b
Standards Data Rate Typical Range Applications
Bluetooth Wireless sensor
ZigBee 20 to 250 Kbps 10 to 100 m
Network
Wireless
Bluetooth 1 to 3 Mbps 2 to 10 m Headset or
Wireless Mouse
ZigBee Wireless
Data Rate IEEE 802.11b 1 to 11 Mbps 30 to 100 m Internet
Connection

Figure 1.1: Short Range Wireless network Comparison

power consumption. Low battery power consumption is biggest advantage of the standard

because of the fact that, most of time the network devices are in sleep mode.

1.1 Motivation

ZigBee standard was specifically developed to address the need for very low cost implemen-

tation of low data rate wireless networks with ultra low power consumption. The ZigBee

Standard reduces the implementation cost by simplifying the communication protocols at

reduced data rate. The minimum requirements to meet ZigBee specifications are relatively

relaxed as compared to other standards such as IEEE 802.11 and Bluetooth. This reduces

the complexity and cost of implementing ZigBee compliant transceivers. This feature of

ZigBee enhances the application in wireless sensor network [7] for gathering information

from sensors such as temperature,humidity, pressure and other physical parameters.

One of the general trend in research is new implementation strategies for the present

technology in real time with an aim to reduce the chip size. For this, FPGA is one of

the best hardware tool which is generally used for transceiver implementation [8]. In the

literature, several researchers have provided implementation of different part of the ZigBee

standard using FPGA [9]. Since VIRTEX 5 FPGA board has wide range of capacity and

is popularly used in industry applications for implementation and verification of hardware

design. Hence, in this present work the baseband transceiver was implemented on VIRTEX

2
1 CHAPTER 1. INTRODUCTION

5 FPGA board.

In fact Direct Sequence Spread Spectrum (DSSS) is used in ZigBee transceiver which

uses a PN sequence. In general random PN sequence is used for this purpose. Recently [10]

residue number system (RNS) has been used in DSCDMA. This code provides improved

performance in terms of better bit error rate. and exhibits better cross correlation. This

gives a motivation for implementing RNS based PN sequence for ZigBee transceiver.

1.2 Background of Research

Significant work has been carried out over past few years in the field of FPGA implemen-

tation and ASIC implementation for short range wireless communications. This section

summarizes some of these.

The advantages of ZigBee in short range wireless communication brought in many

researchers trying to implement hardware for real time applications. Oh and Lee [11]

implemented a 2.4 GHz radio transceiver for WPAN in 0.18−µm CMOS technology. In this

Offset Quadrature Phase Shift Keying (OQPSK) pulse shaping has been implemented for

2.4 GHz band of IEEE 802.15.4 standard. A circuit implementation of ZigBee transceiver

for 868/915 MHz was proposed by Wang et. al [12]. The modulator and demodulator

for 868/915 MHz band of IEEE 802.15.4 standard have been implemented in [13, 14] for

ZigBee wearable devices in medical applications.

The ZigBee standard supports DSSS for baseband modulation [1]. In general Pseudo

Noise(PN) random sequence is used in DSSS. However, in ZigBee a standard spreading se-

quence is used. Recently the RNS based PN sequence has attracted attention of researchers

for FIR and IIR filter design [15]. RNS based codes system design can provide implementa-

tion with low computational complexity and can be used for efficient architecture for VLSI

implementation [16]. It has been demonstrated that, the RNS based PN sequence provides

superior performance than PN sequence in DSCDMA [10]. Recently Zhang and Yang [17]

proposed a redundant residue number system(RRNS) assisted multi carrier direct-sequence

code division dynamic multiple-access (MC/DS-CDDMA) for cognitive radios (CRs). Here

3
1 CHAPTER 1. INTRODUCTION

the author have used three types of receiver multiuser diversity aided multistage minimum

mean-square error multiuser detector (RMD/MS-MMSE MUD) for signal detection. A

2.45 GHz WPAN modulator and demodulator compliant to IEEE 802.15.4 physical layer

has been proposed in [18], where a non coherent demodulation scheme is used to overcome

the complexity and power dissipation problem by using phase shift down sampling to de-

tect the maximum phase accumulation, which is the location of correct data. A simulation

based study on transceivers using DSSS and OQPSK modulation was investigated in [19]

. The bit error rate (BER) performance of the transceiver in additive white gaussian noise

(AWGN) and rayleigh flat-fading channels have been analyzed here. A low power baseband

processor for IEEE 802.15.4 standard, which estimate and compensate carrier phase error

at baseband with carrier phase synchronization has been proposed in [20]. Here an existing

packet detection algorithm is used to estimate the large carrier offset. The monitoring of

remote controlled switching devices with power management system using ZigBee is pre-

sented in [21]. A low power 2.4 GHz radio transceiver consuming 14.7 mA current in receive

mode and 15.7 mA in transmitting mode with Packet Error Rate(PER) of 1% has been

designed by Kluge et.al in [22]. A low cost demodulator for ZigBee receiver which extracts

the symbol directly from the baseband signal rather than recovering the PN sequence was

presented [23].

1.3 Objective of Thesis

The work focuses in this thesis is to design and implement of ZigBee baseband transceiver

for IEEE 802.15.4 for short range wireless communication system. The design is preferred

to use in wireless sensor network applications as per the customer requirement, like in

industry automation, medical applications and military applications etc.

The work reported in this thesis was primarily carried out as a part of design for defence

application. It was designed to develop a ZigBee transceiver based on IEEE 802.15.4. The

RF section design of this ZigBee transceiver was completed in [24]. This thesis presents

the baseband processing of ZigBee transceiver. The objectives of this thesis involved in

4
1 CHAPTER 1. INTRODUCTION

design includes.

• Simulation of the ZigBee transceiver system using MATLAB/SIMULINK

• Hardware description language verilog HDL based design and simulation in modelsim

simulator.

• Implementation of the design in VIRTEX 5 FPGA board.

• Design a new form of baseband transceiver using new set of PN sequence based on

residue number system(RNS).

• Analyze the performance of the RNS based baseband transceiver with pseudo noise(PN)

sequence based ZigBee transceiver in AWGN channel and Rayleigh fading single path

channel.

• Verilog HDL simulation of RNS based baseband transceiver using modelsim simula-

tor.

• Implementation of RNS based design on VIRTEX 5 FPGA board.

1.4 Thesis Organization

This section presents the outline of the thesis.

The thesis has been divided into five chapters, remaining four chapters are described

here.

Chapter-2: ZigBee for IEEE 802.15.4 Standard

The basic of IEEE 802.15.4 standard for short range wireless communication is de-

scribed briefly. The ZigBee protocol’s networking layer and the network topologies used

in ZigBee are presented. Various type of data transfer techniques like beacon enabled and

non-beacon enabled along with the physical layer specifications of IEEE 802.15.4 are de-

scribed. The modulation scheme used in ZigBee baseband transceiver is also discused in

this chapter.

5
1 CHAPTER 1. INTRODUCTION

Chapter-3: FPGA Implementation of ZigBee Baseband Transceiver

The FPGA implementation scheme for ZigBee transceiver is explained in this chapter.

ZigBee baseband transmitter and receiver are first designed in hardware programming

language verilog HDL and simulated in modelsim simulator. The transmitter consists

of bit to symbol, symbol to chip, serial to parallel blocks and pulse shaping circuit are

described and each of these blocks are simulated on modelsim. Following this the deign

has been implemented on VIRTEX 5 FPGA Board. In similar manner the receiver consists

of parallel to serial conversion, correlation of received data bit and decoding blocks. All

the blocks are first simulated on modelsim simulator and next implemented on VIRTEX 5

FPGA board. The chapter covers the details of VLSI module of transceiver, clock circuit

generator and input output description of each module. Detail description of digital clock

manager(DCM) is also presented.

Chapter-4: ZigBee Transceiver Design using Residue Number System based PN se-

quence for Spreading

Design of a new ZigBee transceiver using a new set of PN sequence based on Residue

Number System(RNS) is proposed in this chapter. The transceiver uses these codes for

spreading and despreading. The chapter presents brief description of RNS which includes

the basics of RNS , moduli-set selection, Chinese Remainder Theorem (CRT) etc. The

chapter presents the performance of RNS based PN sequence in ZigBee. The bit error

rate performance of the transceiver is compared with standard ZigBee BER performance.

Following this the proposed PN sequence is used for DSSS in the transceiver.

Chapter-5: Presents conclusion and Future scope of work. The chapter provides an

analysis on contribution of the work reported in this thesis. Limitation of work and some

direction for future work is also presented here.

6
 2
ZigBee for IEEE 802.15.4 Review

2.1 Introduction

ZigBee standard provides a set of specifications for short range wireless communication with

low battery power consumption, in mesh networking topology. This enables applications to

work in remote areas without intervention. One such application is sensor network, which

can monitor patients health condition by a physician who is away from the patient and

patient can be in remote area. The patient wears a ZigBee device which in turn is interfaced

with a ZigBee sensor [25, 26]. The ZigBee sensor collects information of patients health

parameters like blood pressure, pulse rate, sugar etc and sends this information to personal

computer placed inside patient’s cabin with the help of ZigBee protocol from where the

information transferred to the physicians through broadband internet connection [2, 14].

Wireless ZigBee sensors have been used to detect the structural strength of large build-

ing by placing ZigBee enabled wireless sensors at different part of the building [27]. All the

sensors combining to form a ZigBee wireless network and are able to gather the structural

information of the building on periodic basic. It helps the authority of the building for in-

spection of structural strength of the building before opening to the public after a disaster

like earthquake or tsunami etc[28].

7
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

Table 2.1: ZigBee Terminology in contrast to IEEE 802.15.4

ZigBee Terminology IEEE 802.15.4 terminology
ZigBee Coordinator IEEE 802.15.4 PAN Coordinator
ZigBee Router IEEE 802.15.4 Coordinator(FFD)
ZigBee end device Reduced function device(RFD)

There are many standards developed for the collection of information from different

source point. This includes short range wireless networking standards like IEEE 802.11[29]

wireless local area network (WLAN), Bluetooth [30]. Each standard has its own advantage

for particular field of application. For short range, low data rate and very low power

consumption with low implementation cost the ZigBee standard is most suitable.

2.2 ZigBee Data Communication

To enhance data communication, the ZigBee network uses mesh topology, IEEE 802.15.4

network devices like Reduced Function Device (RFD), Full Function Device(FFD) and

various data transfer techniques such as beacon enabled and non-beacon enabled etc. In

ZigBee data communication IEEE 802.15.4 devices plays major role which are termed as

ZigBee devices. The ZigBe standard terminology used in the network is presented at Table

2.1.

2.2.1 ZigBee Network Device

A full function device (FFD) can perform all the tasks and responsibilities mentioned in

IEEE 802.15.4 standard and can play any role in the network topology. A FFD acts as

master to all of the RFDs associated with the network designated as coordinator and can

communicate with any device in the network [31, 32].

A FFD can act as three different types of role such as a coordinator, a PAN coordinator

and a device. A coordinator in a network must be a FFD having the capability of relaying

message from FFD to RFD. The coordinator can be referred as PAN coordinator if it is the

principal coordinator of the network. And the device in the network other than coordinator

is known as device [4].

8
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

The capability of reduced function device (RFD) is limited in the standard. The appli-

cations of RFD are for simple applications like turning on and off switches etc. The RFD

can communicate with a full function device only not with any other RFD. The reduced

function device has normally less power as compared to full function device (FFD).

2.2.2 Topologies in ZigBee

As per the requirement of application IEEE 802.15.4 operates in two types of topology

such as star topology and peer to peer topology. The Figure 2.1a shows star topology and

Figure 2.1b shows peer topology of formation of ZigBee network.

PAN
Coordinator PAN
Coordinator
Reduced function device Reduced function device
Full function device Full function device
Communication Flow Communication Flow
(a) Star Topology (b) Peer Topology
Figure 2.1: Star and Peer to peer Topology

• Star topology

In star topology the communication is established between the full function device

(FFD)and all other reduced function devices. An FFD activated first time will estab-

lish its network with the RFDs and becomes PAN coordinator. The PAN coordinator

9
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

starts a network chooses a PAN identifier which is currently not used by any other

network. This allows independent operation of star network [32].

A PAN coordinator facilitates some associated application to initiate and terminate

the network communication so it is used to initiate, terminate and routing of commu-

nication around the network, for which PAN coordinator is considered as the primary

controller of the PAN. The network devices operating on a network have unique 64-

bit addresses which is used for identification of the device. This address may be used

for direct communication within the PAN, or a short address can also be allocated

by the PAN coordinator when the device associates with the network [1].

Mostly the PAN coordinator gets supply from a main supply but all other RFD are

as usually battery powered. This topology mainly applied for home automation,

personal computer peripherals, different games and toy application.

• Peer to peer topology

Peer to peer topology is different from star topology. In this topology the devices can

communicate with each other as long as they are in the range of the network. The

device which first communicates to the network termed as PAN coordinator, con-

structs the network structure formation and also restricts the topological formation

of network. Peer to peer topology can be help full for creating larger topology or

formation of mesh topology by connecting the star topologies each other along with

peer to peer topology. Application of peer topology lies in industrial control and

monitoring, wireless sensor network, asset tracking and for security purposes.

• Mesh topology

Mesh topology is the combination of many star topology and peer to peer topology. In

ZigBee standard mesh topology of the network is helpful to create a ZigBee cluster as

presented in Figure 2.2 [2]. This topology helpful to create cluster tree network where

most of the devices use FFDs for allowing other devices to associate with the network.

10
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

In a network, RFDs are used as end device as RFD does not allow other device to

associate with network due to its limited resources. Out of all FFDs one FFD acts

as PAN coordinator, which first initiates the network by choosing an unused PAN

identifier and broadcasts beacon messages to neighbor devices. The device receiving

beacon message can request the PAN coordinator to join the network, if the PAN

coordinator permits the device to join it adds a new device as a child device in its

neighbor list. Then the new device adds the PAN coordinator as its parent in its

neighbor list and begins transmitting periodic beacon messages. Other new devices

can join the network with that device. If the original candidate device unable to join

the network at the PAN coordinator then it will search for another PAN coordinator

[4].

2.3 IEEE 802.15.4 Physical Layer Specifications

2.3.1 Introduction

IEEE 802.15.4 specifies the physical layer protocol functions and interactions with MAC

layer which defines the hardware level requirement such as jamming resistance, receiver

sensitivity and transmitter output power etc. Physical layer of ZigBee networking protocol

layer performs following tasks.

• Activation and deactivation of the radio transceiver.

• Energy detection(ED) with in the current channel.

• Link quality indicator for received packets.

• Clear channel assessment (CCA) for carrier sense multiple access with collision avoid-

ance (CSMA-CA).

• Channel frequency selection.

• Data transmission and reception.

11
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

Reduced function device

Full function device (PAN Cordinator )
First PAN Coordinator

Communication Flow
Figure 2.2: ZigBee Cluster Tree

• Energy detection

In ZigBee network if a device need to transmit signal message then it first goes to

receive mode to detect the signal energy in the desired channel. This process is

known as energy detection(ED). For 2.4 GHz the signal energy spread over eight

symbol periods. In this process the receiver estimates the signal energy level not the

type of signal. The Energy detection process unable to detect the weak signal with

lower energy level close to receiver sensitivity level. Hence the receiver sensitivity

energy level is chosen to be lowest signal energy to detect the weak signal of lower

12
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

energy with a packet error rate of less than 1%. After this process the MAC layer

requests the PHY layer to perform energy detection then the PHY layer returns an

8-bit integer representing the energy level of the desired frequency channel of interest.

• Carrier sense(CS)

Carrier sense technique is used to verify the availability of the channel for communi-

cation or method of free channel detection in network. Similar to ED here also the

device wants to transmits data first goes to receive mode to identify whether any

signal is present in the desired frequency channel. In CS the signal is demodulated to

verify whether the signal modulation and spreading are compliant with the charac-

teristics of the PHY that is currently in use by the device. If the signal in the channel

is compliant to the IEEE 802.15.4 PHY than the device considers the channel is busy

even if the energy level is high in the channel.

• Link Quality Indicator(LQI)

Link quality indicator is a parameter which indicates the quality of data packets

received by the receiver. The quality of packets can be found by the measurement

of received signal strength (RSS), which in turn is the measure of total energy of

the received signal. Quality or error in data packets can also be found out by the

measurement of signal to noise ratio (SNR). Higher the SNR value lower the chance

of error in data packet or vice versa as mentioned by (2.1).

SN R = T otal Signal power(PS )/T otal N oise power(PN ) (2.1)

LQI is performed for every data packet received by the receiver, and the LQI infor-

mation is reported to the MAC sub layer by using a response. The LQI information

is also available to the network and application layer of the architecture which is used

for data path selection. Network layer can use the LQI values of the device in the

network to decide the path to route a message. The path has higher LQI values have

better chance of reception of message by the receiver.

13
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

• Clear channel assessment(CCA)

CCA is a part of PHY layer management service which performs in first step of

CSMA-CA. MAC layer requests PHY layer to perform clear channel assessment which

ensures that the channel is free. In CCA the results of Energy Detection (ED) and

carrier sense (CS) decides whether the channel is assumed to be free or not. CCA is

performed in following three modes.

• Mode 1

In this mode of operation, the Energy detection result considered to detect the usage

of frequency channel. If energy level is above the ED threshold, channel is considered

as busy. Here the ED energy threshold is set by the device manufacturer.

• Mode 2

In this mode the carrier sense (CS) result taken in to account for checking the avail-

ability of free channel. Here the channel is busy only if information in frequency

channel is compliant of PHY layer of the device that is performing CCA.

• Mode 3

This is the hybrid of Mode I and Mode II, where the combination of mode I and

mode II are performed as logically OR and logically AND as given below.

(i) The detected energy level is above the threshold and the carrier in the channel is

sensed. (ii) The detected energy is at above the threshold or carrier in the channel

is sensed.

2.3.2 Data rate and frequency of operation

There are three bands of operating frequency for ZigBee. They are 868 MHz, 915 MHz,

and 2.4 GHz band. Since 2.4 GHz frequency band is available world wide, the 2.4 GHz

band is widely used for short range wireless communication. 868 and 915 MHz band are

14
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

Table 2.2: Data rate and frequency of operation

Frequency(MHz) Number of Modulation Chip Bit Rate Symbol Spreading
Channels Rate Rate Method
(Kchips/s) (Ksym-
bol/s)
868-868.6 1 BPSK 300 20 20 Binary
DSSS
902-928 10 BPSK 600 40 40 Binary
DSSS
2400-2483.5 16 O-QPSK 2000 250 62.5 16-array
Orthogo-
nal

Table 2.3: IEEE 802.15.4 Channel assignment

Channel Page Channel Number Description
0 868 MHz band (BPSK)
0 1-10 915 MHz band (BPSK)
11-26 2.4 GHz band (O-QPSK)
0 868 MHz band (ASK)
1 1-10 915 MHz band (ASK)
11-26 Reserved
0 868 MHz band (O- QPSK )
2 1-10 915 MHz band (O-QPSK )
11-26 Reserved
3-31 Reserved Reserved

respectively used in Europe and North America. The specifications for different band of

operation is given in Table 2.2.

2.3.3 Channel Assignment

The initial release of IEEE 802.15.4 2003 version did not include channel pages. In initial

release there were provision of more than 27 channels, so it can exceeds up to 32channels.

To support the growing number of channels channel assignment were defined through

combination of channel number and channel pages in IEEE 802.15.4 2006 version. In

IEEE 802.15.4 standard the channel page 0-2 are used for 868/915 MHz and 2.4 GHz

bands and channel page 3-31 are reserved for future extension of channel number[1]. To

distinguish between the supported PHYs channel pages are introduced in the current release

as presented in Table 2.3.

15
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

By using the above table the center frequency of different bands are calculated

• Center frequency for 868 MHz band

fc = 906 + 2 × (Channel N umber − 1) (2.2)

by using (2.2) the center frequency for channel number 4 can be calculated to be 912

MHz.

• Center frequency for 2400 MHz band

fc = 2405 + 5 × (Channel N umber − 11) (2.3)

by using (2.3) the center frequency for channel number 12 calculated to be 2410 MHz.

2.3.4 ZigBee protocol Architecture

Figure 2.3 shows ZigBee Networking protocol Layers [2]. Out of seven layers of OSI ar-

chitecture ZigBee implements the layers which are essential for low power, low data rate

applications. In ZigBee architecture the lower two layers ie Physical layer (PHY) and

medium access control (MAC) are defined by IEEE 802.15.4 and all other layers are de-

fined by ZigBee. The security features are defined by both ZigBee and IEEE 802.15.4

standard. The network architecture that implements all layers as shown in Figure 2.3 is

known as ZigBee protocol networking layers. The layer to layer communication takes place

with the help of service access points (SAP), which are the mutual points in the archi-

tecture at which one layer can requests the services of another adjacent layer and after

getting the request the SAP forwards the request to next layer. The role of service access

points(SAPs) in ZigBee protocol architecture are as follows.

• NLDE-SAP(Network Layer Data Entity Service Access Point) Application layer unit

or APS Security Management requests data service to network layer.

• NLME-SAP(Network Layer Management Entity Service Access Point) Application

layer device Management unit can request any management request to network layer

by this SAP.

16
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

Application Layer

APS Security Management ZDO Management Plane

Security
NLDE-SAP NLME-SAP
Service
Provider Network Layer
(SSP)
Network Security Network Routing Network
Management Message Broker Management Management

PLME-SAP PLME-SAP

Medium Access control Layer

PD-SAP PLME-SAP

Physical Layer

Figure 2.3: ZigBee Protocol Layers

• MLDE-SAP(MAC Layer Data Entity Service Access Point) Network Layer requests

any data service from MAC.

• MLME-SAP(MAC Layer Management Entity Service Access Point) Network man-

agement entity requests to this SAP for any control service from MAC layer.

• PD-SAP(PHY Data Service Access Point) The MAC layer requests any data service

from PHY layer

• PLME-SAP(Physical Layer Management Entity Service Access Point) The MAC

layer requests for any control services from PHY layer.

2.3.5 PHY Layer Services

Physical layer is an intermediate stage between MAC layer and the ZigBee transceiver.

This layer includes two type of services such as PHY data service and PHY management

service. PHY data service is responsible for transmission and reception of PHY protocol

17
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

data unit (PPDU) across the radio channel which is accessed through service access point

PD-SAP. PHY management entity is known as physical layer management entity (PLME)

which provides the layer management services to MAC and is accessed through PLME-

SAP. PHY layer also maintains a database of objects associated with PHY layer which is

called as PHY information data base ((PIB).

Data Management
MLME
MAC NWK
MAC PIB NLDE NLME

MCPS -SAP MLME -SAP

PD-SAP MLME-SAP
MAC
MLME
MAC Common Part
PLME sublayer MAC PIB
PHY
PHY PIB

PD-SAP MLME -SAP

Transceiver PHY PLME

PHY PIB

(a) PHY Reference Model Interferences MAC (b) MAC sub layer reference model
Figure 2.4: PHY and MAC Reference Model

PHY Data Service

The local MAC layer generates a requests for transmission and provides MAC Protocol

Data Unit(MPDU) which is transmitted to PHY layer. After getting request from MAC,

PHY process the request and acknowledges to MAC layer with a report of successful or

unsuccessful transmission containing following information. The PHY reference model for

interfacing MAC is shown in Figure 2.4a. Once the data is received by radio transceiver

then the PHY layer notifies the MAC layer for arrival of MPDU along with the LQI

information.

18
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

The reason of unsuccessful transmission:

• The radio transceiver may be in receiving mode, it should be in transmitting mode

because the radio transceiver unable to transmit and receive simultaneously.

• The radio transmitter may be busy for transmitting. (It is assigned with one task of

transmission)

• The Radio transceiver may not be in enable state, it is in disable state.

PHY Management Services

PHY management services are performed by Physical Layer Management Entity Service

Access point (PLME-SAP). The PLME-SAP is used to transport management commands

between PHY and MAC. PLME-SAP performs following tasks [1].

• Clear channel assessment: When CSMA requires a channel access the MLME requests

PLME for CCA and the results of CCA can be (i) transceiver disabled (ii) channel is

available for communication (iii) channel is busy because it is used by another device

for transmission.

• Signal Energy detection: Energy detection request is created by MLME and issued

to PLME. After detection of energy level it is reported back to MLME. The energy

detection may fail due to unavailability of channel or due to a disabled radio.

• Enabling and disabling the radio transceiver: MLME may request the PLME to

put the transceiver in one of the three states such as (i) transceiver disabled (ii)

transmitter enabled (iii) receiver enabled.

• Obtaining information from PHY-PIB: MLME transmits requests to PLME for read-

ing PHY attributes stored in PHY-PIB. After reading the values PLME provides it

to MLME.

19
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

• Setting the value of PHY layer information base attribute: PHY layer can change

the read-only PHY attributes and also all other attributes can be changed or set to

a given value by receiving a request from MLME.

The work reported in this thesis is design of physical layer only.

2.3.6 MAC Services

The MAC layer acts as a interface between PHY and next higher layer to MAC such as

network layer. The higher layer here is ZigBee networking protocol layer. Similar to PHY

the MAC has also a sub layer reference model for its operation. Figure 2.4b describes this

reference model clearly.

MAC sub layer can access the physical radio transceiver by a process of performing

following tasks.

• Creates network beacon when the ZigBee device is a coordinator.

• Synchronization with network beacons

• Supports PAN association and disassociation

• Supporting the security of device

• Employs channel access mechanism such as CSMA-CA

• Maintains the Guaranteed Time Slot (GTS) mechanism

• Provides reliable link between two peer MAC layers.

MAC layer has two type of services namely Data service and management service. The

MAC layer management entity (MLME) which performs management services accessed

by MLME-SAP. It interacts with the counterpart of ZigBee network layer namely network

layer management entity (NLME).The data service is accessed through MAC common part

sub layer access points (MCPS-SAP). Like PHY, MAC has also its own data base known as

20
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

MAC PAN Information base (MAC-PIB), which includes all primitives and packet formats

[1].

CSMA-CA

Carrier Sense Multiple Access with Collision Avoidance (CSMA-CA) is a channel access

mechanism protocol without centralized control, which operates as follows [1]. Multiple

devices can use the same frequency channel for their communication medium with the help

of CSMA-CA.

The node that wants to transmit a data packet first performs the clear channel assess-

ment procedure, i.e., it listens to the medium, for a prescribed time. If the medium is found

to be free or clear (or idle) during that time, the node can transmit its packet. Otherwise

if the channel is busy for transmitting another stations data or packet than the concerned

station waits for a certain time for channel to be free. Different MAC algorithms used in

different ways to calculate the time they need to listen to the channel during clear channel

assessment procedure and to calculate the time to wait (i.e., the duration of the back off

period) before the next transmission attempt. It is possible that the transmissions from two

or more nodes overlap in time, which results in a collision and loss of all packets involved.

If lossless communication is desired, collisions must be detected so that the lost packets

can be retransmitted. Since a collision can be detected only at the receiver side, some form

of acknowledgment from the receiver may be needed; some MAC protocols provide this

facility, while others leave it to some of the upper layers like transport layer. In terms of

reaction time, the MAC protocol facility is more efficient than the transport layer facility.

In the basic CSMA protocol, carrier sensing is performed only at the sending node.

GTS

GTS stands for guaranteed time slot, which is helpful for network devices to transmit

without using CSMA-CA. The PAN coordinator dedicates a specific time slot to a particular

device. This is called a guaranteed time slot (GTS). Therefore, a device with an allocated

21
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

GTS will start transmitting during that GTS without using the CSMA-CA mechanism.

2.3.7 Data transfer between devices

When data is received by radio transceiver, the PHY layer notifies MAC layer for reception

of MAC PDU (MPDU). PHY layer not only provides the MPDU to the MAC layer, it

also delivers the Link Quality Indicator(LQI) information. Figure 2.5a presents the data

transfer algorithm from application layer of one ZigBee device to another. The data always

does not come from the application layer it may come from or generated by the MAC layer

without involvement of upper layer. Data is provided by the ZigBee device object(ZDO)

or an application object to application support sub layer (APS). During transmission data

packet travels through all the layers and each layer adds its own header (preamble) with

the data unit (DU) and then passes to the adjoining lower layer. Data unit in each layer

is recognized by name of the layer and are defined as follows. [1].

ZDO or ZDO or
Application Application
object object

Application Data Service

Layer A Layer B
APDU APDU

Network Data Service

Request

NPDU NPDU

MAC Data Service Indication

MPDU MPDU

PHY Data Service

Response
PPDU PPDU

Confirm
Radio Radio

Source Destination

(a) Data Transfer Between Layers (b) Two way Handshaking

Figure 2.5: Data Transfer Between Devices

22
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

• The data unit in APS layer is called application protocol data unit (APDU) after the

addition of header and footer in APS layer.

• The data unit APDU received from APS is received by Network (NWK) layer and

performs the header addition to produce the network protocol data unit(NPDU)

• Similarly the data unit in MAC layer is known as MAC protocol data unit(MPDU)

• The data unit (MPDU) from MAC layer received by PHY layer and reformed to

PHY protocol data unit (PPDU) by adding the header which is later transmitted by

radio.

on receiver side, the data unit is passed upward from lower layer to higher layer. The

headers added in transmitter side are removed in different layers till data unit reaches the

originated application layer.

2.3.8 Data Transfer Methods

In the network the data packet travels through various nodes like network coordinator, end

device, network router in the mesh networking topology[33][1]. Hence, to facilitate data

communication the standard follows different types of data Transfer methods like.

• Data transfer from a device to a coordinator

• Data transfer from a coordinator to a device

• Data transfer between two peer devices

The data transfer in the network from device to coordinator or coordinator to a device can

be classified as Beacon Enabled and Non-beacon Enabled, which are explained below.

• Beacon Enabled Data Transfer: In beacon enabled data transfer the PAN coordinator

periodically transmits a beacon message. The PAN coordinator transmits beacon

message to the devices associated with the network to ensure that all the devices are

synchronized. The disadvantage of using beacon enabled data transfer is that all the

23
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

devices in the network wake up at regular intervals and listens to the beacon message

for synchronizing their clocks and go back to sleep mode. In this case many devices

in the network do not perform any other than clock synchronization [34]. This results

higher power consumption comparison to non-beacon data transfer there by reducing

battery life.

Coordinator Device Coordinator Device

Beacon Message
Data Request
Data Request

Ack
Ack

Data
Data

Ack (Optional)
Ack (Optional)

(a) Beacon Enabled (b) Non-beacon Enabled

Figure 2.6: Data Transfer From Coordinator to Network Device

• Non-Beacon Enabled Data Transfer: The ZigBee network where the PAN coordinator

does not transmit beacon message is known as non-beacon network. In this case of

data transfer, the devices in the network do not synchronize to each other which in

turn consumes less power. Therefore the battery life is high in comparison to beacon

enabled data transfer [35].

Data Transfer Primitive

Primitives convey the required information by providing a particular service. These prim-

itives are an abstraction, because they specify only the provided service rather than the

means by which it is provided. A service is specified by describing the service primitives

and parameters that characterize it. A service may have one or more related primitives

24
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

Coordinator Device Coordinator Device

Beacon Message
Data

Data

Ack if requested
Ack if requested

(a) Beacon Enabled (b) Non-beacon Enabled

Figure 2.7: Data Transfer From Network Device to Coordinator

that constitute the activity which is related to the particular service. Each service primi-

tive may have zero or more parameters used to convey the required information to provide

the service.

The primitives can be of as follows

• Request: The request primitive is passed from layer A to layer B to request, that a

service is initiated.

• Indication: The indication primitive is passed from layer B to layer A to indicate

internal layer B event, that is significant to layer A. This event may be logically

related to a remote service request or it may be caused by layer B internal event.

• Response: The response primitive is passed from layer A to layer B to complete the

procedure invoked by indication primitive.

• Confirm: The confirm primitive is passed from layer B to layer A to convey the

results of previous service requests

PHY layer management entity provides a set of service primitives for request and re-

sponse which are described by 2.4 to 2.7. Communication between layer to layer takes

25
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

place through these service primitives.All primitives, PHY packet format (PPDU format)

are processed as per the two way handshaking method as presented in Figure 2.5b.

< T heP rimitive > .request (2.4)

< T heP rimitive > .indication (2.5)

< T heP rimitive > .response (2.6)

< T heP rimitive > .conf irm (2.7)

2.4 ZigBee 2.4 GHz PHY Specifications

The physical layer specification of ZigBee for 2.4 GHz frequency band supports 250 Kbps

data rate with Modulation and spreading functions as described below.

2.4.1 Modulation and Spreading

2.4 GHz physical layer facilitates 16-ary quasi-orthogonal modulation technique. Each sym-

bol transmitted consists of four information data bits which are used to select sixteen nearly

orthogonal pseudo-random noise(PN) sequences. The PN sequences for all successive data

symbols are concatenated during transmission. The aggregated chip sequence developed

from DSSS are modulated on to carrier using offset quadrature phase-shift keying.

2.4.2 Reference Modulator

Figure 2.8 specifies the block diagram for modulation and spreading of physical layer for

2.4 GHz frequency band. The modulator and spreading functions including Bit to symbol

mapping, Symbol to chip mapping and O-QPSK modulator are described as follows.

26
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

Bit to Symbol Symbol to Chip O-QPSK

Mapping Mapping Modulator

Figure 2.8: Modulation and spreading

32 Bit Chip Sequence

C 0 C 1 C 2 • • • • • C 29 C 30 C 31 • • • • • C 0 C 1 C 2 • • • • • C 29 C 30 C 31

B0 B1 B2 B3 • • • • • • • B13 B14 B15 B16

4 Bit Symbol

Figure 2.9: Symbol to chip mapping process

Bit to symbol mapping

Digital signal of data bit ‘1’ and ‘0’ inside the PHY Protocol Data Unit(PPDU) is encoded

as per the spreading function. Here binary inputs are mapped to data symbols. For four

bit symbol there are sixteen possible symbol starting from 0000 to 1111 are mapped to

data symbol. Thus there are sixteen symbol to chip mapping as presented in following

section.

Symbol-to-chip mapping

In this block each data symbol of 4 bits are mapped to 32 bit chip sequence as prescribed

in the ZigBee standard. The process of mapping is presented in Figure 2.9 and the DSSS

sequence for mapping is presented at Table 2.4 [36, 1].

27
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

Table 2.4: Symbol to Chip Mapping for 2.4 GHz Band

Data Symbol(Binary) Chip values (C0 C1 . . . C31 )
0000 11011001110000110101001000101110
1000 11101101100111000011010100100010
0100 00101110110110011100001101010010
1100 00100010111011011001110000110101
0010 01010010001011101101100111000011
1010 00110101001000101110110110011100
0110 11000011010100100010111011011001
1110 10011100001101010010001011101101
0001 10001100100101100000011101111011
1001 10111000110010010110000001110111
0101 01111011100011001001011000000111
1101 01110111101110001100100101100000
0011 00000111011110111000110010010110
1011 01100000011101111011100011001001
0111 10010110000001110111101110001100
1111 11001001011000000111011110111000

O-QPSK Modulation

Each data symbol is represented by 32 bit chip sequence which increases the chip rate by

32 times of symbol rate, and produces a 2 Mbps DSSS signal. Even bits and odd bits of

DSSS signal is separated into I-phase data stream and Q-phase data stream by assigning

odd bits and even bits as presented in Figure 2.10. To perform offset between I-phase and

Q-phase data stream, the Q-phase chip is delayed by chip period(Tc ) 2.10.

2T c

I-Phase C0 C2 C4 ••• C 28 C 30

Q-Phase C1 C3 C5 ••• C 29 C 31

Tc
Figure 2.10: O-QPSK chip

28
2 CHAPTER 2. ZIGBEE FOR IEEE 802.15.4 REVIEW

Pulse shaping

Half sine pulse shaping of I-phase and Q-phase data stream is achieved by processing the

signal with 2.8. This helps to reduce side lobes keeping inter symbol interference (ISI) low.

The Pulse shaping process presented in Figure 2.11

{
sin(Π 2Tt c ) 0 ≤ t ≤ 2Tc
p(t) = (2.8)
0 otherwise

Figure 2.11: O-QPSK Pulshaping for I and Q Channel

2.5 Summary and Discussion

The analysis of ZigBee standard along with IEEE 802.15.4 standard for short range wireless

communication explained briefly in this chapter. The analysis includes ZigBee networking

topologies, various data transfer methods, ZigBee networking protocol layers and physical

layer specifications for 2.4 GHz frequency band of IEEE 802.15.4 standard. The process

of data transfer from between ZigBee devices with resource capability are presented in

this chapter. During analysis it was found that ZigBee is one of the better technology

of short range wireless applications like in industry automation, medical applications etc.

Physical layer features such as input data rate, bit to symbol and spread spectrum for

implementation of ZigBee baseband transceiver is analyzed here which is validated and

implemented in next chapter.

29
 3
FPGA Implementation of 2.4 GHz ZigBee Transceiver

In previous chapter, short range wireless communication for ZigBee was discussed. In order

to develop the ZigBee transceiver the physical layer properties for 2.4 GHz described in

previous chapter are taken as reference. This chapter presents the implementation process

of ZigBee baseband transceiver. Which is simulated using hardware description language

verilog HDL and finally implementation on VIRTEX 5 FPGA board.

3.1 ZigBee Transmitter in Simulink

With an aim for realization of ZigBee transceiver, the system was validated through MAT-

LAB/Simulink environment in this section. The block diagram of ZigBee transmitter is

presented in Figure 3.1. This transmitter design in Simulink is shown in Figure 3.2 which

is explained in following subsections.

IEEE 802.15.4 standard uses the spreading methods to improve the effect of receiver

sensitivity, jamming resistance, and also to reduce the effect of multi path interference

[37, 38]. The spreading technique used here is DSSS. In this, every four incoming data bit

is combined together to form a four bit symbol and this four bit symbol is mapped onto

an unique 32 chip sequence available in a lookup table. Hence there are sixteen different

32 bit chip sequences. The set of chip sequences is shown in Table 2.4. The chip sequences

30
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Input ,250 Inphase data

Modulation with

Serial to parallel conversion

Kbps Half sine pulse
quadrature
shaping
carrier
Bit to Symbol Symbol to chip
Mapping Mapping +

Offset Half sine pulse Modulation with

delay shaping inphase carrier
2 Mbps Transmitter
Quadrature output
data

Figure 3.1: Block Diagram of ZigBee transmitter

Scope6

Scope9
Sine Wave1
Sine Wave3

Unipolar to
Unipolar to
Random Bipolar D
Q Bipolar
Integer Converter
Converter
CLK Scope7 Product
Random Integer Unipolar to Unipolar to
Generator Bipolar Clock !Q Bipolar
!CLR
Converter2 J Converter1
Q
D Flip-Flop
CLK Product2
Product4 Scope11
!Q
K Scope2

Bipolar to J-K
Flip-Flop Unipolar to
Unipolar to Unipolar D
PN Sequence Q Bipolar
Bipolar Converter
Generator Converter Scope1
Converter CLK
Bipolar to Unipolar to
PN Sequence Unipolar to Unipolar !Q Bipolar Product1
!CLR
Generator Bipolar Converter Converter
Converter3 1 D Flip-Flop1

Constant
Product3

Sine Wave2

Sine Wave4

Scope3

Figure 3.2: Simulink Model of ZigBee Transmitter

are also known as Pseudo random noise sequence (PN sequence). The chip sequence is the

collection of zeros and ones. To avoid the similarity between any two chip sequences, the

standard follows a procedure or algorithm of doing the cross-correlation between the chip

sequences. The cross-correlation is calculated by multiplying the sequences together and

then calculating the summation of the result. The 32 bit chip sequence contains a sequence

of zeros and ones in unipolar form. Before calculation of the cross correlation the unipolar

sequence is converted to bipolar sequence ie a ‘0’ is represented by ‘-1’ and ‘1’ is as it is. If

x(n) and y(n) are two sequences then the cross correlation of x(n) and y(n) is calculated

by 3.1.

n=∞
∑
rxy (0) = x(n)y(n) (3.1)
n=−∞

31
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

The relation 3.1 and 3.2 calculates the cross correlation of x(n) and y(n) where neither

of these sequences shifted. Hence it defines that the two sequences are dissimilar to each

other. In this case the sequences are known as orthogonal to each other and termed as

orthogonal sequences. The sixteen pseudo noise sequences used in the standard IEEE

802.15.4 are not completely orthogonal to each other and are called as quasi orthogonal

or nearly orthogonal [2]. The cross correlation of other Pseudo Noise(PN) sequences are

calculated by shifting one sequence with a margin of time period ‘k’ hence the relationship

is presented as 3.2

n=∞
∑
rxy (k) = x(n)y(n − k) (3.2)
n=−∞

3.1.1 Symbol to Chip Mapping

ZigBee can transmit data at 250 Kbps.

• A input data stream of 250 Kbps was generated by using random integer generator

block in communication tool box. The parameters like M- ary number, initial seed,

sample time and output data type are configured to get fixed binary stream. In

a real time scenario, this data stream is supplied by application that will generate

information to be transmitted [11].

• Using PN sequence generator block in communication tool box, the 32 bit PN se-

quence is generated from 4 bit symbol. This sequence is of 2 Mbps. The parameters

for generator polynomial, initial states, sample time and output data type were ad-

justed to generate the 32 bit Pseudo Noise code.

• The input data stream and PN code are converted to Non Return to Zero (NRZ)

format and multiplied to each other to generate DSSS signal.

An input data stream, a PN sequence and DSSS output for a ZigBee transmitter

generated using Simulink is presented in Figure 3.3a, 3.3b and 3.4.

32
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

(a) Input data in Simulink

(b) pnsequence
Figure 3.3: Input data and PN sequence

Figure 3.4: DSSS output in Simulink

33
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

In phase signal (1 Mbps )

Serial Data (2 Mbps )
I4 I3 I2 I1
Q4 I4 Q3 I3 Q2 I2 Q1 I1
Q4 Q3 Q2 Q1

Q-phase signal (1 Mbps )

Figure 3.5: Serial to Parallel Conversion

3.1.2 Serial to Parallel Conversion

ZigBee standard uses OQPSK modulation. In order to achieve this. the input chip sequence

of 2 Mbps serial data stream is converted to even bit stream and odd bit stream by splitting

the sequence to even bit and odd bit of serial data stream separately. This is presented

Figure 3.5. From the figure it is clear that the odd bits are assigned to odd bit stream

and even bits are assigned to even bit stream by naming the signal as I-phase signal and

Q-phase signal respectively.

with the help of Simulink tool box, the I-phase and Q-phase signal extracted from the

DSSS serial data, each of these is 1 Mbps. Splitting the signal into I-phase and Q-phase

requires a clock, a JK flip flop and two D flipflops. Aligning the odd and even streams

in time requires half bit offset which is achieved with D flip flop. The I-phase signal and

Q-phase signal after serial to parallel conversion taken from the simulink scope is given in

Figure 3.6a and 3.6b respectively.

3.1.3 Half sine pulse shaping

The NRZ form of I-phase and Q-phase data of shape sharp edges. Transmission of these

would generate large number of side bands. In order to remove side bands the pulses to be

transmitted and passed through a pulse shaping circuit. This is achieved by multiplying

the signal with 90◦ out of phase sine waves to get the half sine pulse shaping. This is shown

in Figure 3.12a and 3.12b.

34
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

3.1.4 RF Modulation
√
After pulse shaping the even bit stream be (t) is superimposed on a carrier 2ps cos ω0 t
√
and the odd bit stream b0 (t) is superimposed on a carrier 2ps sin ω0 t with the use of two

multipliers. This generates two signals, these are so (t) and se (t) respectively. These signals

are then added to generate transmitted output signal vm (t) which is given by (3.3).

√ √
vm (t) = 2ps so (t) cos ω0 t + 2ps se sin ω0 t (3.3)

The two terms in (3.3) represents the I-phase stream so (t) and Q-phase stream se (t).

Both se (t) and s0 (t) occupy the same spectral range but they are individually identifiable

because of the phase quadrature of the carriers. These four possible output signals have
√
equal amplitude 2ps and are in phase quadrature. They have been identified by their

corresponding values of b0 and be .

This modulation is carried out with 2.4 GHz high frequency sinusoidal carrier. By

using sine wave block in Signal Processing tool box, 2.4 GHz high frequency signal is

(a) I-phase data in Simulink

(b) Q-phase data in Simulink

Figure 3.6: Serial to Parallel conversion in Simulink

35
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

generated by adjusting signal processing tool box parameters. Then I-phase half sine pulse

shaping signal is multiplied by a sine wave and Quadrature half sine pulse shaping signal

is multiplied by its orthogonal carrier i.e., cosine signal which is 90◦ phase shift of original

sinusoidal carrier. After modulation the I-phase and Q-phase signal outputs taken from

Simulink scope is presented in Figure 3.8a and 3.8b. Then signals are combined together

to get the transmitter output in ‘sum’ block from commonly used blocks as obtained in

Figure 3.8c

3.2 ZigBee Receiver in Simulink

This section describes ZigBee receiver architecture and its validation in Simulink. The block

diagram of ZigBee receiver and simulink design is presented in Figure 3.9 and 3.10 , here

coherent detection has been used. The receiver implements reverse process of transmitter

blocks and consists of demodulation, halfsine pulse shaping, low pass filter, sample and

hold, comparator and despreading. following sub sections describes each of the blocks.

(a) I-phase pulse shaping in Simulink

(b) Q-phase pulse shaping in Simulink

Figure 3.7: I-phase and Q-phase pulse shaping in Simulink

36
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

3.2.1 Receiver Operation

The incoming received signal is applied to two synchronous demodulators, each of which

consists of amultiplier followed by a low pass filter. The received signal signal from two

modulators are 90◦ out of phase. They can be represented as 3.4 and 3.5.

t
c1 = cos(ω0 t) cos 2Π + n(t) (3.4)
4Tb

t
c2 = sin(ω0 t) sin(2Π ) + n(t) (3.5)
4Tb

(a) I-phase modulation signal through Simulink

(b) Q-phase modulation signal through Simulink

(c) transmitter output at Simulink

Figure 3.8: Modulation output waveform

37
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Received
Input Sample
Low Pass
and
Filter Baseband
Hold Circuit Comparator
output
t

Parallel to Serial Converter

cos 2πf c t sin 2π Delay
4Tb
De-Spreading

Sample
Low Pass
and
Filter
Hold Circuit Comparator

t
sin 2πf c t cos 2π
4Tb
Figure 3.9: Block Diagram of ZigBee Receiver

Sine Wave5

B-FFT
Spectrum
Scope
1
Lowpass Scope15
z 1
10 Product5 Sign
Lowpass Filter Unit Delay Bipolar to
Gain Integrate Unipolar
and Dump Converter
Product8 Sign3
Scope Scope13 Scope14 Integrate
Sum1
and Dump1

Lowpass
Product6
Sign1
Lowpass Filter1
Sine Wave6
Unipolar to
PN SequencePN Sequence
Bipolar
B-FFT Generator Generator
Converter
Spectrum Unipolar to
Scope1 Bipolar
Converter3

Figure 3.10: Simulink Model of ZigBee Receiver

38
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Where, c1 and c2 are transmitted signal and n(t) is AWGN. Here it has been assumed

that the channel is single path AWGN channel.

The demodulator at receiver is a coherent receiver its output contains baseband data

and higher frequency harmonics represented as (3.4) and (3.5). Higher frequency harmonics

are removed from the signal by passing through a low pass filter. Generally, a third

order Butterworth filter having a cutoff frequency of 2/Tb Hz is used for the extraction of

baseband data. The resultant signal after low pass filter is passed through a zero order

sample and hold circuit which holds the data for one bit period, later is passed through a

decision device comparator for thresholding. For signal data levels s1 and s2 the threshold

is provided by (3.6), if input to comparator is greater than threshold value, it decodes the

bit as ‘1’ otherwise it decodes as ‘0’.

s1 + s2
s∗ (t) = (3.6)
2

3.2.2 Demodulator

In the process of FPGA design of ZigBee receiver, the performance of ZigBee receiver was

first analyzed in simulink environment. The performance of simulation using Simulink is

presented below.

The performance ZigBee receiver was corrupted with AWGN and the noisy signal is

considered as received signal. This signal was at 2.4 GHz carrier frequency. This noisy

received signal is presented in Figure 3.11a is passed through I-phase and Q-phase demod-

ulator for 2.4 GHz carrier frequency and 90◦ phase shift between them. The demodulated

I-phase Q-phase signal are displayed in Figure 3.11b and 3.11c respectively.

3.2.3 Half sine pulse shape

In OQPSK, abrupt phase changes occurs due to the multiplication of abrupt rectangular

baseband waveform with the Quadrature carrier causing loss of phase continuity, so there

is no phase continuity. These abrupt changes give rise to spectral components at high

39
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

(a) Received signal with noise in Simulink

(b) I-phase demodulation in Simulink

(c) Q-phase demodulation in Simulink

Figure 3.11: Demodulation in Simulink

40
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

frequencies. The baseband spectral range is very large and multiplication by a carrier

translates the spectral pattern without changing its form. It is required to alleviate this

difficulty by passing the baseband signal through a low pass filter to suppress the side

lobes. Such filtering cause inter symbol interference. To overcome this problem half sine

pulse shaping can be implemented. In this design, half sine was used for pulse shaping

to exhibit phase continuity and to smoothen the baseband waveform. This is achieved by

multiplying sine wave with a the bit period of 4Tb in both I and& Q channel.

(a) I-phase pulse shaping in Simulink

(b) Q-phase pulse shaping in Simulink

Figure 3.12: I-phase and Q-phase pulse shaping in Simulink

3.2.4 Low Pass Filter

A low pass filter of cutoff frequency 500 KHz is designed by using analog filter design

block, where the parameters like design method, filter type, filter order and pass band edge

frequency are configured. The output of low pass filter in both I-phase and Q-phase is

presented in Figure 3.13a and 3.13b. The two outputs are next fed to sample and hold

circuit.

41
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

3.2.5 Sample and Hold

The sample and hold circuit samples the signal for every T time period, by using sample

time and time period T in zero order hold circuit. Here a relative time period of 2µs is

used for sampling to get the sample and hold output as shown in Figure 3.14a, 3.14b. The

output taken from this stage forms the even and odd bit stream from receiver.

3.2.6 Parallel to serial conversion

The I-phase signal and Q-phase signal at receiver combined together in a switch block to

form the 2 Mbps serial data, which is the DSSS signal transmitted by the transmitter.

To make accurate alignment between I-phase and Q-phase during combination, a half bit

delay is introduced to I-phase signal at receiver. Here the parameters like threshold value

and criterion for parallel to serial conversion is optimized to generate 2 Mbps serial data

as shown in Figure 3.16a.

(a) I-phase through low pass filter

(b) Q-phase through low pass filter

Figure 3.13: I-phase and Q-phase passed through low pass filter

42
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

(a) I-phase after sampling

(b) Q-phase after sampling

Figure 3.14: I-phase and Q-phase passed through sampled and hold

(a) I-phase data stream at receiver

(b) Q-phase data stream at receiver

Figure 3.15: I-phase and Q-phase passed through comparator

43
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

3.2.7 Despreading

The received 2 Mbps DSSS signal is despreaded by using same PN sequence which is used

for serial to parallel conversion at transmitter. The original data stream is received by

multiplying the PN sequence with DSSS signal which is presented at Figure 3.16a. Here

the PN sequence is the replica of PN sequence used in transmitter.

(a) 2 Mbps serial data stream at receiver

(b) PN sequence at receiver

(c) Baseband output data

Figure 3.16: Despreading at receiver

44
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Baseband Transmitter
shaping_even Inphase signal

PNsequence

Parallel
Datain
input Data

shaping_odd Q-phase signal

Figure 3.17: ZigBee baseband Transmitter

3.3 Baseband Transmitter Design for FPGA

Following the simulation of ZigBee transceiver in simulink, the transceiver was described in

verilog HDL and simulated on modelsim before downloading the design to FPGA. After the

transmitter was thoroughly simulated in MATLAB/Simulink, the transmitter was designed

using hardware description language verilog HDL and implemented in FPGA as per the

reference modulator diagram mentioned in Chapter 2 [1]. The basic baseband module of

ZigBee transmitter for implementing in FPGA presented in Figure 3.17. The transmitter

consists of following stages.

3.3.1 Input data stream

Input data stream is a sequence of binary bits which is the input signal to transmitter.

The input data is stored in a memory and read from the memory. The input data bit

is transmitted at a rate of 250kbps. So in the design at every 4µs one binary input

bit is transmitted. Clock signal of 4µs time period is generated by the Digital Clock

Manager(DCM). The design of clock circuit is presented in FPGA implementation section,

it provides clock signal for 10 MHz, 1 MHz, 250 KHz.

45
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.1: I/O Description Datain Module

SL No PIN Name Direction Size Description
1 Clock(clk) Input 1 Synchronous Clock of 250
KHz
2 Reset Input 1 Asynchronous Reset
3 Input Input 1 Data input
4 Enable Input 1 Valid Signal for input Data
5 Symbol Output 4 4 bit output termed as symbol
which is of data rate 62.5 kilo
symbol per second
6 Enout output 1 Valid Signal for output Data

Input Output Description of Datain Module

The first block of transmitter is the input data to symbol conversion. In verilog HDl

design the incoming data bits 1 and 0 are stored in a 4 bit register to form a symbol,

which is having 62.5 kilo symbols per second. Input output diagram of input module is

depicted in Figure 3.18a and Table 3.1 presents its description. The signals shown in Figure

3.18b represents clock signal of frequency 250 KHz named as ‘clk’, the incoming data bit

consisting of four bit symbol named as ‘symbol’ and one bit input data of 250 Kbps named

as ‘input data’ for 270000 ns simulation time.

Clock
Symbol
Reset
DATAIN
Input
Enout
Enable

(a) I/O Diagram of Datain Module

(b) Bit to symbol Mapping(Modelsim Simulation)

Figure 3.18: Bit to Symbol

46
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

3.3.2 Symbol to chip conversion

As per ZigBee baseband design at 2.4 GHz, a combination of 4 bit input data symbol

provide the respective 32 bit chip sequence for IEEE 802.15.4, this was presented in Table

2.4. Symbol to chip mapping produces a 2 Mbps serial unipolar signal which is performed

OQPSK modulation in parallel to serial block. In HDL design, 32 bit chip sequence are

transmitted by clock signal ‘clk’ generated by DCM.

Input Output Description

Input to this module is 4 bit symbol, which are mapped to 32 bit chip sequence resulting a

output of 2 Mbps serial data. The I/O description of this module is presented in Table 3.2

. The signal output for this subsystem is presented at Figure 3.19b. Figure 3.19b displays

clock signal of frequency 250 KHz named as ‘clk’, the incoming input data of 250 Kbps

named as ‘input data’, data symbol consists four input data bit named as ‘symbol’, clock

signal of frequency 1 MHz named as ‘clk32’, 32 bit chip sequence in 32 bit register inside

the PNSEQUENCE module named as ‘chip sequence’ and 2 Mbps direct sequence spread

spectrum signal named as ‘DSSS 2 mbps data’ for 270000 ns simulation time.

Clock
PNSeq
Reset
PNSEQUENCE Enparallel
Enout
Serial Data
Symbol

(a) I/O Diagram of PNSEQUENCE Module

(b) Symbol to chip Generator(Modelsim Simulation)

Figure 3.19: Symbol to chip and DSSS

47
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.2: I/O Description pnsequence module

SL No PIN Name Direction Size Description
1 Clock(clk32) Input 1 Synchronous Clock of 1 MHz
2 Reset Input 1 Asynchronous Reset
3 Enout Input 1 Valid Signal for input Data
4 Symbol Input 4 4 bit input termed as symbol
which is of data rate 62.5 kilo
symbol per second
5 PNSeq Output 1 2 Mbps output after symbol to
chip mapping
6 Enparallel Output 1 Valid signal for output data
7 Serial Data Output 1 2 Mbps DSSS output after
spread spectrum

3.3.3 Serial to Parallel Conversion

After Symbol to chip mapping serial data constitute a 2 Mbps digital signal which is the

combination of all chip sequences as per the symbol at input. ZigBee standard at 2.4 GHz

uses OQPSK modulation. For this reason the DSSS serial data is converted to a 2 channel

parallel data for I and Q processing channels. Serial to parallel conversion is implemented

with a set of flip flops and flag registers in verilog HDL design.

In verilog HDL design the serial to parallel conversion is performed with the help of

Digital Clock Manager(DCM) clock signal clk32 of 1 MHz. During the active edge of clock

period along with the the flag value 2 Mbps serial data converted to two 1 Mbps parallel

data. during conversion the odd bits of serial data assigned to oddbit stream and even

bits are assigned to even bit stream. since the first bit of the serial data is an odd bit it

is assigned first and even bit assigned next. the process continues as long as input serial

data is available.

Input Output Description

After chip mapping the output is 2 Mbps serial data in pnsequence module, which is

separated in to I-phase signal and Q-phase signal in this module. I-phase signal contains

the odd bits and Q-phase signal contains even bits of serial 2 Mbps signal. I/O description

of this module is presented in Table 3.3. The modelsim result from simulation is presented

48
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.3: I/O Description of Serial to Parallel

SL No PIN Name Direction Size Description
1 Clock(clk32) input 1 Synchronous Clock of 1 MHz
2 Reset input 1 Asynchronous Reset
3 Enparallel input 1 Valid Signal for input Data
4 Serial Data input 1 After DSSS the 2Mbps data
stream
5 PNSeq input 1 chip sequence without spread
spectrum
6 Enshaping output 1 Valid Signal for pulse shaping
7 Oddbit output 1 Every oddbit of serial data
stream assigned to this pin
8 Evenbit output 1 Every Evenbit of serial data
stream assigned to this pin

at Figure 3.20b for 270000 ns simulation time. This figure includes the clock signal ‘clk32’

of frequency 1 MHz, 32 bit register named as ‘flag’ consists of 32 bit chip sequence, spread

spectrum signal of 2 Mbps data rate named as ‘DSSS 2 Mbps data’, I-phase signal and

Q-phase signal named as ‘Inphase signal’ and ‘Qphase signal’ respectively. Each of these

signals becomes 1 Mbps data rate after serial to parallel conversion.

Clock Evenbit
Reset SERIAL
Serial Data TO Enshaping
PARALLEL
PNSeq
Oddbit
Enparallel

(a) I/O Diagram of Serial to parallel Converter

(b) Serial to parallel(Modelsim Simulation)

Figure 3.20: I-phase Q-phase separation

49
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.4: Samples for Pulse shaping

Sampling time (ns) Sample value for bit 1 Sample value for bit 0
in twos complement
100 00000000 00000000
200 00010000 11110000
300 00100000 11100000
400 00110000 11010000
500 00111000 11001000
600 01000000 11000000
700 00111000 11001000
800 00110000 11010000
900 00100000 11100000
1000 00010000 11110000

3.3.4 ZigBee transmitter Pulse Shaping

In our design unipolar signal has been used. Here during symbol to chip conversion the

verilog logic replaces ‘-1’ with ‘0’ and ‘1’ is left as it is. During Pulse shaping the bits one

and zero in even stream and odd stream are shaped by half sine wave, hence the name is

half sine pulse shaping. Here pulse shaping has been implemented by taking continuous

values from lookup table in memory. Here stored sample values are multiplied with respect

to bit one and zero of even bitstream and odd bitstream to generate continuous value

signals.

The samples are multiplied with the help of flag value and DCM. The even bitstream

and odd bit stream are of time period 1000 ns. So for every 100 ns one sample is multiplied

as per the flag value to get the pulse shape as shown in simulation result presented in

Figure 3.21b and 3.22b. The lookup table uses 10 continuous values for pulse shaping.

More samples can be multiplied to get the sharpness in the pulse shaping.

Input Output Description

The I-phase signal and Q-phase signals are pulse shaped by continuous values. In this

design the sampling values are fixed integers as mentioned in earlier section. The I/O

diagram and description of this module are as follows and the modelsim result is presented

50
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.5: I/O Description even pulse shape

SL No PIN Name Direction Size Description
1 Clock(clk50) Input 1 Synchronous Clock 10 MHz
2 Reset Input 1 Asynchronous Reset
3 Enshaping Input 1 Valid Signal for pulse shape
4 Evenbit Input 1 Evenbit PIN output from se-
rial to parallel module is as-
signed to this pin
5 Even pulseshape Output 8 8 bit sample value assigned to
this output pin

in 3.21b. This result interprets clock signal of frequency 10 MHz named as ‘clk_sampling’,

Q-phase signal termed as ‘Qphase signal’. This signal produces Q-phase pulse shaping

signal termed as ‘Qphase pulse shaping’ after multiplication of sampling values with Q-

phase signal. This result shown for simulation time of 270000 ns from top to bottom

respectively. Similarly for I-phase signal, pulse shaping description is mentioned in 3.5

with the modelsim results as presented in Figure 3.22b. Figure 3.22b interprets clock

signal of frequency 10 MHz named as ‘clk_sampling’, I-phase signal named as ‘Inphase

signal’ and I-phase pulse shaping signal named as ‘Inphase pulse shaping’ for simulation

time of 270000 ns.

Clock
Reset EVEN Even
PULSE SHAPE pulseshape
Evenbit
Enshaping

(a) I/O Diagram of Even pulse shaping

(b) Q-phase pulse shaping(Modelsim Simulation)

Figure 3.21: Q-phase pulse shaping

51
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Clock
Enaddition
Reset ODD
PULSE SHAPE Odd
Oddbit
pulseshape
Enshaping

(a) I/O Diagram of Odd pulse shaping

(b) I-phase pulse shaping(Modelsim Simulation)

Figure 3.22: I-phase pulse shaping

Table 3.6: I/O Description Odd pulse shape

SL No PIN Name Direction Size Description
1 Clock(clk50) Input 1 Synchronous Clock
2 Reset Input 1 Asynchronous Reset
3 Enshaping Input 1 Valid Signal for pulse shape
4 Oddbit Input 1 Oddbit PIN output from se-
rial to parallel module is as-
signed to this pin
5 Odd pulseshape Output 8 8 bit sample value assigned to
this output pin
6 Enaddition Output 1 Valid Signal for IQaddition
assigned to this output pin

3.4 Baseband Receiver Design for FPGA

This section concentrates on baseband design for synchronous detection of ZigBee trans-

mitter. Input to receiver is the pulse shape signal from the transmitter. The pulse shaping

function performed at the transmitter reduces Inter Symbol Interference (ISI). So it has

been assumed that the pulse shaping signal is the baseband signal after demodulation at

the receiver. The baseband receiver design covers Sample extraction of even bit and odd

bit streams, parallel to serial conversion, chip to symbol conversion. The baseband receiver

is designed in verilog HDL as per the block diagram presented at Figure 3.23.

52
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Baseband Receiver

Evenstream _out
Q-phase signal

IQ-addition

Decoder
Received
Input

Oddstream _out
Inphase signal

Figure 3.23: ZigBee Baseband Receiver

3.4.1 Pulse shape detection

Here, synchronized communication system is assumed. from the received signal the I and

Q channels are separated and the evenbit stream and oddbit stream are identified. The

pulse output of the transmitter is provided to the input of the baseband receiver for further

processing.

3.4.2 Parallel data(even and odd stream)

In verilog HDL design the DCM clock signal clk50 of frequency 10 MHz is used to detect

the samples in odd pulse shape and in even pulse shape. So according to the samples in the

pulse shape waves the even bit and odd bit streams are detected at receiver. The half bit

delay in even bit stream is provided with clock signal clk50, which is same as transmitter

clock signal clk50 created by DCM.

Input Output Description

Input output description of I-phase signal and Q-phase signals at receiver are presented in

Table 3.7 and 3.8. The I/O diagram and signals are presented at Figure 3.24a to Figure

3.25b. The simulation result shown in Figure 3.24b represents clock signal of frequency

10 MHz named as ‘clk50’ and the signals ‘sampling value’ and ‘register’, which consists of

53
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.7: I/O Description of Evenstream_out

SL No PIN Name Direction Size Description
1 Clock(CLK50) Input 1 Synchronous Clock of 10 MHz
2 Reset Input 1 Asynchronous Reset
3 Even pulseshape Input 8 8 bit even pulse shape from
evenpulseshape module as-
signed to this pin
4 Enaddition Input 1 Valid Signal for pulse shape
assigned to this pin
5 Evenbit_out Output 1 Even bit detected from Even
pulse shape assigned to this
output pin

8 bit sampling values used for pulse shaping and the Q-phase signal at receiver named as

‘evenbit_out’ for 270000 ns simulation time.

The signals shown in Figure 3.25b represents clock signal of frequency 10 MHz named

as ‘clk50’ and the signals ‘Sampling value’ and ‘QphaseRegister’ which shows 8 bit sam-

pling values used for pulse shaping along with the I-phase signal at receiver named as

‘oddbit_out’ for 270000 ns simulation time.

Clock

Reset EVENSTREAM
Evenbit_out
Even OUT
pulseshape
Enaddition

(a) I/O Diagram of Evenstream_out

(b) Even data stream at Receiver(Modelsim Simulation)

Figure 3.24: Q-phase signal at Receiver

3.4.3 Parallel to serial conversion

In this module the two channel ‘I’ and ‘Q’ parallel data of 1 Mbps each combined together

to form a 2 Mbps serial signal. In HDL design the data bits of I-phase and Q-phase signal

54
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Clock

Reset ODDSTREAM
Oddbit_out
Odd OUT
pulseshape
Enaddition

(a) I/O Diagram of Oddstream_out

(b) Odd data stream at Receiver(Modelsim Simulation)

Figure 3.25: I-phase signal at Receiver

Table 3.8: I/O Description of Oddstream_out

SL No PIN Name Direction Size Description
1 Clock(CLK50) Input 1 Synchronous Clock of 10 MHz
2 Reset Input 1 Asynchronous Reset
3 Odd pulseshape Input 8 8 bit odd pulse shape from
oddpulseshape module as-
signed to this pin
4 Enaddition Input 1 Valid Signal for pulse shape
assigned to this pin
5 Oddbit_out Output 1 Even bit detected from Even
pulse shape assigned to this
output pin

at receiver are assigned to a 4 bit register alternately. The Figure 3.26 represents the

formation of serial data from I-phase and Q-phase signal. The clock signal clk32 is used for

parallel to serial conversion, which is used for serial to parallel conversion at transmitter.

As per the active edge of clk32 and flag value we are assigning the I-phase and Q-phase

data bit to form the 2 Mbps serial data.

Input Output Description

The input and output description of this module with modelsim simulation of serial data

is presented in Table 3.9 and Figure 3.27b respectively. The Figure 3.27b presents the

clock signal ‘clk32’ of frequency 1 MHz, I-phase and Q-phase signal respectively termed as

‘oddbit_out’ and ‘evenbit_out’ and the received DSSS serial signal at receiver named as

55
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

In phase signal

IQ combination
I4 I3 I2 I1

I1 Q1 I2 Q2

Q4 Q3 Q2 Q1

Q-phase signal

Figure 3.26: Parallel to serial conversion at Receiver

‘Received 2 Mbps serial data’ for simulation time of 270000 ns.

Clock
Flag_Enable
Reset
Evenbit_out IQaddition Serialdata_out
Oddbit_out
ENIQaddition

(a) I/O Diagram of IQaddition

(b) IQ Combination at Receiver(Modelsim Simulation)

Figure 3.27: I-phase and Q-phase Combination at Receiver

3.4.4 Chip to symbol mapping at receiver

In this module the received 2 Mbps serial data is stored in a 32 bit register with the

help of clock signal clk32 generated by DCM, which is used as symbol to chip mapping

in transmitter. Here the lookup table used in transmitter for symbol to chip mapping is

used in reverse order ie chip to symbol mapping for decoding the received symbol. The

received chip sequence is compared with the 16 different chip sequence transmitted by the

transmitter. Figure 3.28 demonstrates this process matching. Here, the XNOR operation

is carried out on received chip sequence with each of 16 different chip sequences used at

transmitter. In XNOR operation, the counter with maximum value provides the symbol

56
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.9: I/O Description of IQaddition

SL No PIN Name Direction Size Description
1 Clock(clk32) Input 1 Synchronous Clock of 1 MHz
2 Reset Input 1 Asynchronous Reset
3 Evenbit_out Input 1 one bit data from Evenstream
out module assigned to this
pin
4 Oddbit_out Input 1 one bit data from Oddstream
out module assigned to this
pin
5 ENIQaddition Input 1 Valid input signal for IQad-
dition assigned to this output
pin
6 Flag_Enable Output 1 Valid input signal for
Chipmapping assigned to
this output pin
7 Serialdata_out Output 1 Combined 2 Mbps serial data
after IQ addition, assigned to
this output pin

transmitted by the help of lookup table of chip to symbol mapping. This 4 bit symbol is

stored in a 4 bit register named as ‘dataout’, then MSB of ‘dataout’ is sent to the output

as received data bit. The Figure 3.30 shows modelsim simulation result of recovered data

bit. This figure displays clock signal of frequency 250 KHz named as ‘clk’, the received

symbol at receiver after chip to symbol mapping is described by a signal of 4 bit register

named as ‘Symbol_received’ and the received signal at receiver is termed as ‘Input databit

received’ which is shown in simulation result over 270000 ns simulation time.

Input Output Description

The 2 Mbps serial data after addition in IQaddition module, is stored in 32 bit flag and the

decoding performs in this module. The I/O description of chip to symbol decoding module

is presented in Table 3.10. The modelsim simulation result in Figure 3.29b displays chip

mapping at receiver. The figure has 16 different chip mapping sequences of 32 bit each,

which indicates the correlation of chip sequences at receiver. Out of all chip sequence, the

chip sequence has highest matching with received chip sequence is considered as transmitted

chip sequence.

The signals in Figure 3.29b represents the clock signal ‘clk32’ of 1 MHz and 32 bit

57
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Chip Sequence 1 Counter1

Chip Sequence 2 Counter2

Chip Sequence 3 Counter3

•••••••
•••••••
Chip Sequence 16 Counter16

C0 C 1 C 2 C 3 • • • • • C 15 C 16

Out stream after I & Q Combination

Figure 3.28: Decoding at Receiver

register named as ‘flag_2Mbps’ are shown for 270000 ns simulation time. The signals

termed as ‘chipmapping1’, ‘chipmaping2’, ‘chipmaping3’, ‘chipmaping4’, ‘chipmaping5’,

‘chipmaping6’, ‘chipmaping7’, ‘chipmaping8’, ‘chipmaping9’, ‘chipmaping10’, ‘chipmap-

ing11’, ‘chipmaping12’, ‘chipmaping13’, ‘chipmaping14’, ‘chipmaping15’, ‘chipmapping16’

represents chip mapping output of ‘flag_2Mbps’ signal with all possible chip sequences of

corresponding transmitted symbol and received serial signal of 2 Mbps data rate named as

‘Received 2 Mbps serial data’ are shown for 270000 ns simulation time.

3.5 FPGA Implementation

After the system level simulation in Simulink, then the module was programmed using

hardware description language verilog HDL for simulation. The baseband transceiver mod-

ule simulated, synthesized and finally implemented on Virtex 5 FPGA. In the process of

implementation on FPGA, clock signal generation is very vital [39]. In order to get the

required clock signal the digital clock manager available in FPGA was used.

58
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Clock Symbol_out
Reset
Decoding Databit _Received
Serialdata _out
Flag_Enable

(a) I/O Diagram of Decoding module at Receiver

(b) Chip mapping at Receiver(Modelsim Simulation)

Figure 3.29: Chip mapping at Receiver

Table 3.10: I/O Description of Decoding

SL No PIN Name Direction Size Description
1 Clock(clk) Input 1 Synchronous Clock of 250
KHz
2 Reset Input 1 Asynchronous Reset
3 Flag_Enable Input 1 Valid input signal for decod-
ing assigned to this pin
4 Serialdata_out Input 1 Combined 2 Mbps serial data
from IQaddition module, as-
signed to this pin
5 Symbol_out Output 4 4 bit Symbol extracted from
decoding module assigned to
this output pin
6 Databit_Received Output 1 MSB of the Symbol is as-
signed to this output pin
which is data bit transmitted

59
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Figure 3.30: Received data bit

Figure 3.31: XILINX VIRTEX 5 FPGA Board with XC5VLX110T

Following verilog simulation of the transceiver, this section presents VIRTEX 5 FPGA

board specification, clock circuit generation, synthesis report of implemented design and

representation of board internal signals for receiver.

3.5.1 VIRTEX 5 FPGA

The general purpose development board from Xilinx which embeds Virtex 5 FPGA is

ML505. This is useful for prototyping on the new architectural features of Virtex-5 FPGA.

ML505 contains Virtex-5 XC5VLX110T device [40]. This device contains block RAM, DSP

blocks, source synchronous interface block, enhanced clock manager etc. Among all devices

of of VIRTEX 5 FPGA family the XC5VLX110T FPGA has higher logic capacity than

others. The Table 3.11 shows available resources for LX110T. 100 MHz clock frequency

chosen from AH15 pin of FPGA board for generation of clock circuitry [41]. Clock pin

description of VIRTEX 5 board is presented in Table 3.12 [41].

60
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.11: Hardware Resources of VIRTEX 5 LX110T

Slices Flipflops LUTs Multipliers Block RAM(Kb) CMTs
17280 69120 69120 64 4608 6

Table 3.12: Clock frequency for VIRTEX 5 LX110T

Clock Name FPGA PIN Frequency in MHz
USER_CLK AH15 100
CLK_33MHz_FPGA AH17 100
CLK_27_FPGA AH18 27
CLK_FPGA_P L19 200
CLK_FPGA_N K19 200

3.5.2 Digital Clock Manager(DCM)

A properly synthesized clock circuit is essential to achieve implementation task so that the

clock signals should be able to reach all the points of the design at the same time. In order

to achieve this, a perfectly synchronized clock circuitry was designed by the help of Digital

Clock Manager(DCM).

The global clock MUX buffer and clock manager tiles(CMT) provides the solution for

designing desired clock network. There are 6 CMTs available on the board, each contains

two DCMs and one phase locked loop(PLL) [42].The DCM can generate 90◦ , 180◦ and

270◦ phase-shifted version of external clock. Three different clock frequency signals are

generated by using VIRTEX 5 board clock frequency of 100 MHz. The generation of

internal clocks of 10 MHz, 1 MHz, 250 KHz is done by using Figure 3.32, where the clock

divider factors 10 and 4 are used. This digital clock manager develops clock signal as

presented in Figure3.33.

3.5.3 Implementation on VIRTEX 5 FPGA

Xilinx Synthesis Tool (XST) performs the synthesis and optimization in FPGA. During

this process following steps are carried out for design, synthesis and optimization:

• Mapping and optimization on an entity by entity or module by module basis

• Global optimization on the complete design

61
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

IBUF
BUFG
BOARD RESET DCM 1
CLKFX_DIVIDE CLK50 (10MHz)
IBUFG
=10
BOARD CLOCK( 100MHz )

LOCKED

BUFG
DCM 2
CLKFX_DIVIDE CLK32 (1MHz)
=10

LOCKED

DCM 3 CLK (250 KHz)

CLKFX_DIVIDE
=4 RESET

Figure 3.32: Digital Clock Manager

Figure 3.33: Digital Clock Manager

The FPGA Device Macro Generator module provides the XST HDL Flow with a cata-

log of functions. These functions are identified by the inference engine from the Hardware

Description Language (HDL) description. Their characteristics are handed to the Macro

Generator for optimal implementation. The set of inferred functions ranges in complexity

from simple arithmetic operators (such as adders, accumulators, counters and multiplex-

ers), to more complex building blocks (such as multipliers, shift registers and memories).

Inferred functions are optimized to deliver the highest levels of performance and efficiency

for the selected VIRTEX 5 architecture, and then integrated into the rest of the design.

For arithmetic functions, XST provides the following elements:

• Adders, Subtractors and Adder/Subtractors

• Cascadable Binary Counters

62
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

• Accumulators

• Incrementers, Decrementers and Incrementer/Decrementers

• Signed and Unsigned Multipliers

Table 3.13 shows the options to fine-tune FPGA synthesis to meet design constraints.

Table 3.14 shows The Device Utilization Summary. Which is the output of HDL syn-

thesis step to estimate the number of slices and gives the number of flip-flops, IOBs.

Table 3.15 shows the advanced synthesis report, which includes XST additional macro

processing by improving the macros recognized at the HDL synthesis step, or by creating

the new, more complex ones, such as dynamic shift registers. The implemented design

in VIRTEX 5 LX110T FPGA board has maximum clock frequency of 11.792 MHz and

maximum output required time is 3.264 ns. The total logic power consumed by the design

is 0.45 mW.

3.5.4 Synthesis Report

The internal signals along with received baseband data was analyzed through ChipScope

Pro analysis. The received I-phase, Q-phase and baseband data bit received signal are

analyzed as depicted in Figure 3.34. The top level RTL schematic of this transceiver

obtained from the ISE design environment is presented in Figure 3.35.

63
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.13: Synthesis options summary

Synthesis options summary

Input File Name "datain_tb.prj"
Input format mixed
Ignore Synthesis Constraint File No
Output File name "datain_tb"
Automatic FSM Extraction YES
FSM Encoding Algorithm No
FSM Style LUT
RAM Extraction Yes
RAM Style Auto
Rom Extraction Yes
Mux style Auto
Decoder Extraction YES
Priority Encoder Extraction Yes
Shift Register Extraction Yes
Logical Shifter Extraction Yes
Mux Extraction Yes
Resource Sharing YES
Use DSP Block Auto

Table 3.14: Device Utilization Summary

Slice Logic Distribution

Number of Slice Registers 973 out of 69120 1%
Number of Slice LUTs 1179 out of 69120 1%
Number used as logic 1178 out of 69120 1%
Number used as Memory 1 out of 17920 0%

I/O Utilization
Number of bonded IOBs 8 out of 640 1%
Number of BUFG/BUFGCTRLs 10 out of 32 31%
Number of DCM_ADVs 3 out of 12 33%

64
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.15: Advance HDL Synthesis Report

Advance HDL Synthesis Report

ROMs 1
128x1-bit ROM 1
Adders/Substracters 32
6-bit adder 32
Counters 6
2-bit up counter 1
4-bit up counter 2
5-bit up counter 2
7-bit up counter 1
Registers 711
Flip-Flops 711
Latches 545
Comparators 48
Multiplexers 1

Figure 3.34: ChipScope Pro Internal signals

65
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Figure 3.35: Top level RTL Schematic of ZigBee Baseband Transceiver

66
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

3.6 Summary and Discussion

ZigBee baseband transceiver for IEEE 802.15.4 was implemented through various stages

which include.

• Initially the transceiver system simulated through MATLAB/Simulink, where Mini-

mum shift keying(MSK) modulation technique is used. The internal signal of different

stages of simulink model are analyzed.

• The baseband module is described in hardware description language verilog HDL

and simulated in mentor graphics modelsim simulator and finally realized in Virtex

5 FPGA using ISE design environment. The system provided desirable performance.

The simulation result presented above are labeled as X-axis in time(ns) and Y-axis

presents the respective signals.

• Following this, in FPGA implementation, the clock signals of 250 KHz, 1 MHz, and

10 MHz were derived by taking 100 MHz clock signal available on Virtex 5 FPGA

board. Further the design synthesis report along with board internal signals were

also analyzed for verification.

• In XILINX power analysis, the total power consumption for the ZigBee baseband

transceiver is reported to be 0.45 mW. The detail power analysis of logic power and

signal power at each stage of the design reported by the ISE synthesizer is as follows.

67
3 CHAPTER 3. FPGA IMPLEMENTATION OF 2.4 GHZ ZIGBEE TRANSCEIVER

Table 3.16: Hierarchical power analysis

Name Power(w) Logic Power(w) Signal Power(w)
Hierarchical Total Power 0.21672 0.21350 0.000322
Datain_tb 0.21401/0.21672 0.21317/0.21350 0.00084/0.00322
Decoder Module 0.00163 0.00007 0.00156
Evenstream_out 0.00009 0.00004 0.00006
Parallel1 0.00009 0.00002 0.00007
IQADDITION2 0.00011 0.00002 0.00009
SHAPINGE V EN 0.00012 0.00004 0.00008
ODDSTREAM_OUT 0.00012 0.00003 0.00009
PULSHAPING 0.00015 0.00005 0.00010
DATAIN 0.00037/0.00039 0.00004/0.00006 0.00033

68
 4
ZigBee Transceiver Design with RNS Based PN sequence

In this chapter the design of a new form of ZigBee transceiver is presented. This transceiver

uses residual number system based PN sequence instead of the chip sequence used in 2.4

GHz ZigBee for IEEE 802.15.4. This chapter introduces Residue Number System(RNS) in

brief as presented in following section.

4.1 Residue Number System

Residue Number System (RNS) has become very popular in recent time due to their su-

perior performance in terms of large dynamic range, good cross correlation and auto cor-

relation properties. The chapter discusses advantages of RNS which in turn motivated us

to take up the work.

In Residue number system, a number X is represented by a set of residues like x1 , x2 , ..xn ,

modulo given by co-prime moduli. The moduli set is known as the system base and residues

are known as residue digits. The dynamic range of the system ie the number of possible

number representation is defined by the product of all RNS moduli. In this numbers sys-

tem, addition, multiplication, substraction can be used without carry propagation between

residue digits, which enables arithmetic circuits to perform faster than the circuits for bi-

nary representation. This property permits to build arithmetic units of large numbers as

69
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

a set of fast and small circuits [43].

The important merit of RNS over conventional number number system is the absence

of carry propagation in addition and multiplication. In other number systems after an

operation if carry is generated than it is forwarded to the next bit of the operand. In RNS,

low precision is required for ranging to individual prime and co-prime number of moduli-

set, subsequently it enables lookup table implementation for different operations. Due to

lookup table the operation speed of RNS is faster with lower computational complexity

and hence low power consumption. These two features are expected to be advantageous

for VLSI implementation.

4.1.1 Chinese Remainder Theorem

Chinese Remainder Theorem(CRT) is a mathematical formulation used to find independent

unique numbers in a given dynamic range. In RNS CRT has an important role to play, so

as to find RNS representation of a number. CRT implies that if moduli of RNS is chosen

accurately then each number can be uniquely represented in the given dynamic range. In

other words CRT states that if all the divisors are co-prime to each other then the residue

representation of each number is different to each other or unique in the dynamic range
′ R′ . Here ′ R′ is the multiplication of all the divisors or modulis of the moduli-set. Since

the set of divisors or modulis are not fixed, if the set of divisors increases then the bit

representation in RNS will be increased proportionately.

Assume a1 and a2 are two integers which are relatively prime to each other and b1 and

b2 are any two integers, then there is an integer N can be determined by 4.1 and 4.2.

N ≡ b1 (mod a1 ) (4.1)

N ≡ b2 (mod a2 ) (4.2)

N is an uniquely determined modulo (a1 .a2 ). Equivalently if gcd (a1 , a2 ) = 1, then

the pair of residue classes modulo a1 and a2 corresponds to a simple residue class modulo

70
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

(a1 , a2 ) which gives a simultaneous congruences b ≡ bi (mod ai ), for i = 1, 2, 3....b for which

ai are relatively prime. Then, the solution to set of congruences is given by (4.3).

A A A
z = x1 ∗ y 1 , x2 ∗ y2 , .... xp ∗ yb (4.3)
a1 a2 an

where A = a1 , a2 ....an and yi is determined from y1 aA1 = 1( mod a1 )

4.1.2 Arithmetic in RNS

Residue number system completely depend on the congruence relation of the number.

Two numbers are said to be congruent modulo ‘m’ if the number ‘m’ divides exactly the

difference of a and b ie (a − b), then mathematically congruent modulo of a and b is

represented as a ≡ b(mod m). for example 15 ≡ 12(mod 3), 10 ≡ 6(mod 2), 11 ≡ 2(mod 3)

etc.

The RNS gives separate unique representation for all numbers in the range between 0

and M − 1. If the given integer number is greater than M − 1, then RNS representation of

the integer is the integer itself. Therefore, more than one integer may have the same residue

representation. Before this, it should be confirmed that the moduli have to be relatively

prime to be able to exploit the dynamic range. The example of residue representation

shown in Table 4.1 provides the principle as per Chinese Remainder Theorem (CRT)[44].

Choice of Moduli

By the mathematical perspective the best moduli are the prime numbers. If ‘m’ is prime,

then there must be one generator, p ≤ m − 1. then the set of all non-zero residue with

respect to ‘m’ is given by relation (4.4). There are two approach of moduli selection,

one is consecutive method and the other is arbitrary method [16, 43]. The proposed PN

sequence is generated on the basis of consecutive method of moduli-set selection [15], here

the moduli-set [255 254 253 251] is selected for generation of RNS based PN sequence for

transceiver design.

71
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Table 4.1: Residue Representation in RNS

N Relative Prime Moduli Relative Non-Prime Moduli
m1 = 2 m2 = 5 m3 = 7 m1 = 2 m2 = 4 m3 = 6
0 0 0 0 0 0 0
1 1 1 1 1 1 1
2 0 2 2 0 2 2
3 1 3 3 1 3 3
4 0 4 4 0 0 4
5 1 0 5 1 1 5
6 0 1 6 0 2 0
7 1 2 0 1 3 1
8 0 3 1 0 0 2
9 1 4 2 1 1 3
10 0 0 3 0 2 4
11 1 1 4 1 3 5
12 0 2 5 0 0 0
13 1 3 6 1 1 1
14 0 4 0 0 2 2
15 1 0 1 1 3 3
16 0 1 2 0 0 4
17 1 2 3 1 1 5
18 0 3 4 0 2 0
19 1 4 5 1 3 1
20 0 0 6 0 0 2
21 1 1 0 1 1 3
22 0 2 1 0 2 4
23 1 3 2 1 3 5
24 0 4 3 0 0 0
25 1 0 4 1 1 1
26 0 1 5 0 2 2
27 1 2 6 1 3 3
28 0 3 0 0 0 4

72
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Modulo m 1

Modulo m 2
Input Data Forward Reverse Output Data
Analog/Binary Conversion • Conversion Analog Binary
•
•
•
Modulo m n

Figure 4.1: Basic Structure of RNS Processor

Residue = |pi |m : i = 0, 1, 2, ...., m − 2 (4.4)

if m = 7 and p = 5 then residues are calculated as |50 |7 = 1, |51 |7 = 5, |52 |7 = 4, |53 |7 = 6

4.1.3 Structure of RNS Conversion

The basic structure of a typical RNS processor is shown in Figure 4.1. The RNS data

processed in parallel according to the concept of residue number system independently

without carry propagation. In RNS processor encoding of input data to RNS is called

forward conversion and decoding of RNS to conventional number representation is known

as reverse conversion.

For full RNS based system the analog input number first converted to binary form then

the residue of the number calculated during forward conversion and residue number to

binary representation during reverse conversion. During processing, conversion to binary

is an intermediate stage which is required in both forward and reverse conversion. This

makes the conversion inefficient due to increased latency and complexity.

73
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Arbitrary Moduli-set forward conversion

The special moduli-set such as 2n − 1, 2n , 2n + 1 is help full for making forward conversion

process fast and simple. The special moduli-set forward converters are more efficient con-

verters. Other than this large dynamic range requirement in some applications can not be

provide with efficient special moduli-set, which in turn demands an arbitrary moduli-set

forward conversion.

Arbitrary moduli-set forward conversion can be implemented by using lookup tables or

combinational logic. Implantation of converter with combinational logic makes the process-

ing unit complex, so the lookup table implementation is preferred over combinational logic.

Implementation of large dynamic range increases the ROM size and simultaneously the con-

version process becomes inefficient. Therefore a tradeoff between the two implementations

can be utilized using a combination of lookup table and combinational logic[45, 16].

For ZigBee application standard 16 different PN codes of length 32 bit are required.

The process of generation of the code is discussed in succeeding section. Table4.2.

4.2 RNS Based PN Sequence for ZigBee Spreading and Despreading

In this section the residue number system(RNS)[16] based PN sequence generation is ana-

lyzed. A set of new PN sequence based on RNS is proposed for ZigBee baseband transceiver,

which is used in ZigBee baseband transceiver design for spreading and despreading. The

algorithm to generate PN sequence is as follows and the generated PN sequence is presented

in Table 4.2.

Algorithm 1 RNS Based PN sequence Generation

1: Take a dynamic Range for primal generation
2: Selection of moduli set
3: For each primal generated above calculate the residue for each of the moduli
4: Convert the set of residues obtained from step 3 to their equivalent 8 bit binary repre-
sentation
5: Concanate all the binary representation of residues called as PN sequence
6: After obtaining set of PN sequences check for the cross correlation for a specified
threshold
7: The sequences satisfying above condition will be selected as RNS Based PN sequence.

74
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

4.2.1 RNS Generated PN sequence

Residue Arithmetic based PN sequence generation for multiple scenario are mentioned

here. The inputs to the block of RNS generation consists of spread factor β and the cross

correlation threshold, T. A moduli set is selected according to the requirement of no of bits

in PN sequence. The moduli set is selected either by consecutive method or by exponential

method [15]. Here the consecutive method was adopted.

As mention above the Primal is randomly selected from the range R given by R =
∏n
i=1 pi , where p is the moduli set and is represented as p = {p1 , p2 , p3 , . . . , pn }. Let A be

a set of primals chosen from range R and is expressed as A = x1 , x2 , x3 , . . . , xn . Defining

J to be set of residues given by J(i,j) = |xj , pj | for i = 1, 2, . . . , n and j = 1, 2, . . . , k. J can

be represented as.

 
mod(x1 , p1 ) mod(x1 , p2 ) . . . mod(x1 , pn )
J = . . . . . . (4.5)
 

mod(xn , p1 ) mod(xn , p1 ) . . . mod(xn , p1 )

Each row of J is then represented by 8 bit binary representation. The entire row is

termed as desired length of the PN sequence. Hence the generated sequences are tested for

their cross correlation as following.

• Correlation between ci and cj for i = j will be one.

• Correlation between ci and cj for i ̸= j must be less than a threshold value T .

The value of T can vary depending on the application, here for 32 bit PN sequence

generation threshold T = 1.5 is considered to be desirable. This process is repeated

till the required number of PN codes are generated.

4.3 Performance of RNS based PN sequence in ZigBee Transceiver

The modern transceivers have to meet the various customer requirements like high capacity,

high bandwidth efficiency at variable bit rate. On the other hand wireless environment

75
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Table 4.2: Symbol to Chip Mapping using RNS code for 2.4 GHz ZigBee Transceiver
Data Symbol(Binary) Chip values (C0 C1 . . . C31 )
0000 11101011111011101111000111110111
1000 11001101110101001101101111101001
0100 11010110110111011110010011110010
1100 10111111110010101101010111101011
0010 11000000110010111101011011101100
1010 10011100101010111011101011011000
0110 10101010101110011100100011100110
1110 11111011000011110010001001001000
0001 10011001101011001011111111100101
1001 10100001101101111100110111111001
0101 11011001111011110000100000110110
1101 00101000001111110101011010000100
0011 01101110100001011001110011001010
1011 10101111110010011110001100011100
0111 10111001110100111110110100100110
1111 11100111000000110001111001010100

signals are usually impaired by fading and multi path delay spread phenomenon, due to

which the mobile communication system do not perform well. In channels with extreme

fading causes Inter Symbol Interference (ISI), results in high probability of errors and

subsequently system’s overall performance becomes very poor.

The next generation short range wireless communication need to be of higher stan-

dard in order to provide the customer different short communication like little message

transfer and some command transfer. In this section the ZigBee baseband transceiver per-

formance is analyzed in Additive White Gaussian Noise (AWGN) channel and Rayleigh

fading channel. Here the performance of the transceiver is tested by replacing the given

PN code in the standard IEEE 802.15.4 with the newly generated RNS based PN code.

Performance evaluation was carried out through BER simulation using MATLAB software.

The BER performance of two types of transceivers using standard code and RNS based

code for AWGN channel and Rayleigh fading channel is presented in Figure 4.2 and 4.3

respectively.

BER simulation suggests that RNS based ZigBee performs better than standard code

76
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

AWGN Channel
0
10
IEEE PN Sequence
RNS Based PN Sequence

−1
10

−2
10
BER

−3
10

−4
10

−5
10
−10 −5 0 5 10
SNR in dB
Figure 4.2: Performance in AWGN Channel

based ZigBee transceiver. Following this the FPGA based implementation is considered

next.

4.4 RNS Based Baseband Transceiver in FPGA

The design principle adopted in 2.4 GHz ZigBee as discussed in chapter 3 is also used here.

Here, RNS based PN sequence is used in transceiver design for spreading. Hence sixteen

different PN sequences are mapped into sixteen different symbols in a lookup table as

presented in Table 4.2. In line with description given in chapter 3, the baseband module is

described in HDL. The HDL structure implemented has been presented in Figure 3.17 and

comprises of blocks input data stream, bit to symbol mapping, PNsequence,serial to parallel

conversion, pulse shaping, parallel to serial conversion etc. Here a minor modification

77
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Rayleigh Channel
0
10
IEEE PN Sequence
RNS Based PN Sequence

−1
10

−2
10
BER

−3
10

−4
10

−5
10
−10 −5 0 5 10
SNR in dB
Figure 4.3: Performance in Rayleigh fadding Channel

is adopted during symbol to chip conversion, where we perform direct sequence spread

spectrum of RNS based PN sequence with respective symbol and rest of all simulation

model is same.

Figure 4.4a shows that the DSSS signal obtained by doing digitally XNOR operation

of symbol with respective PN sequence. The DSSS signal is of 2 Mbps serial data, and is

converted to two 1 Mbps parallel data, In verilog HDL design the PNSEQUENCE module

of chapter 3 consists of output pin ‘Serial Data’ , which is the DSSS output for RNS based

PN sequence. This signal is processed as serial to parallel conversion in this section.

RNS based transceiver is designed in a similar way as done in previous chapter. The

clock signals CLK, CL32 and CLK50 are derived from the crystal by using digital clock

manager. The verilog HDL module of the baseband transceiver is simulated in modelsim

78
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Data input XNOR DSSS output

PN sequence

(a) DSSS Operation

(b) pnsequence
Figure 4.4: DSSS Operation

simulator as per the description and input output diagram described in previous chapter.

4.4.1 Simulation Result

Bit to Chip Conversion

The incoming input data bit is converted to symbol using the datain module and the

pnsequence module performs symbol to chip mapping. thereafter DSS signal of 2 Mbps is

obtained as output for serial to parallel conversion. These signals are presented in Figure

4.5a and 4.5b.

The Figure 4.5a represents the clock signal of frequency 250 KHz named as ‘clk’, the

incoming data bit ‘input’ of 250 Kbps data rate and four bit symbol named as ‘symbol’

for 270000 ns simulation time. The Figure 4.5b displays signals such as clock signal of

frequency 250 KHz named as ‘clk’, data symbol consists four input data bit named as

‘symbol’, clock signal of frequency 1 MHz named as ‘clk32’, 32 bit RNS based PN sequence

in 32 bit register inside the PNSEQUENCE module named as ‘PN sequence’, the incoming

input data of 250 Kbps named as ‘input’ , serial PN sequence signal named as ‘PNSeq’ and

2 Mbps direct sequence spread spectrum signal named as ‘DSSS 2 mbps data’ for 270000

ns simulation time.

79
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

(a) Bit to symbol in modelsim for RNS

(b) Symbol to Chip for RNS

Figure 4.5: Bit to Symbol to Chip Conversion for RNS

Serial to Parallel conversion

The DSSS output of 2 Mbps signal is separated into two 1 Mbps signal of even bit and odd

bits named as I-phase signal and Q-phase signal, which form the input to I-phase pulse

shaping and Q-phase pulse shaping networks.

The Figure 4.6a shows the signals such as clock signal ‘clk32’ of 1 MHz, 4 bit register

internal signal containing four bits of DSSS signal named as ‘FLAG’, the spread spectrum

signal of 2 Mbps data rate named as ‘DSSS 2 Mbps data’, I-phase signal termed as ‘Inphase

signal’ and Q-phase signal termed as ‘Qphase signal’ for 270000 ns simulation time.

The signals shown in Figure 4.6b represents the clock signal of frequency 10 MHz named

as ‘clock sampling’ used for pulse shaping, I-phase signal named as ‘Inphase signal’ and I-

phase pulse shaping signal termed as ‘Inphase pulse shaping’ for simulation time of 270000

ns. Similarly the signals in Figure 4.6c describes 10 MHz clock signal named as ‘clock

sampling’ used for pulse shaping, Q-phase signal named as ‘Qphase signal’ and Q-phase

pulse shaping signal termed as ‘Qphase pulse shaping’ for simulation time of 270000 ns.

Parallel to Serial conversion

From digital clock manager clock signal ‘clk32’ is used for addition of I-phase and Q-phase

signal at the receiver. Combined I-phase and Q-phase signal forms 2 Mbps serial data,

which is the DSSS signal transmitted by transmitter and is presented in Figure 4.7c. This

figure displays 1 MHz clock signal named as ‘clk32’, I-phase signal ‘oddbit_out’, Q-phase

80
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

(a) Serial to Parallel in modelsim for RNS

(b) I-phase pulse shaping for RNS

(c) Q-phase pulse shaping for RNS

Figure 4.6: Pulse shaping of RNS based Transceiver

signal ‘evenbit_out’, two 4 bit register to store I-phase and Q-phase signal at receiver

namely ‘FLAG’ and ‘IQbus’. The out put of this subsystem is received spread spectrum

signal of 2 Mbps named as ‘Received 2 Mbps DSSS signal’. all these signals described for

270000 ns simulation time.

The signals in Figure 4.7a represents 10 MHz clock signal named as ‘clock sampling’

and I-phase signal received at receiver termed as ‘oddbit_out’ with the help of two 8 bit

registers named as ‘samples’ and ‘Register’ used for sample detection at receiver. This

result shown for 270000 ns simulation time. Similarly the signals in Figure 4.7b shows

the clock signal ‘clock sampling’ of 10 MHz and Q-phase signal at receiver termed as

‘evenbit_out’ for 270000 ns simulation time.

Chip to symbol mapping

The received chip sequence is first stored in 32 bit register for decoding. The clock signal

‘clk32’ is used for this purpose. XNOR operation performed between the received sequence

and 16 numbers of 32 bit transmitted sequences. From output of this the sequence with

highest matching with original received chip sequence is selected and is again mapped to

four bit symbol from a lookup table producing desired symbol. The result of chip to symbol

81
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

(a) Q-phase signal at Receiver in modelsim for RNS

(b) I-phase signal at Receiver for RNS

(c) DSSS serial signal at Receiver for RNS

Figure 4.7: DSSS signal at Receiver for RNS based Transceiver

mapping in modelsim is presented in Figure 4.9a.

Figure 4.9a represents the signals such as clock signal ‘clk32’ of 1 MHz and 32 bit

register named as ‘flag_2Mbps’ are shown for 270000 ns simulation time. The signals

termed as ‘chipmapping1’, ‘chipmaping2’, ‘chipmaping3’, ‘chipmaping4’, ‘chipmaping5’,

‘chipmaping6’, ‘chipmaping7’, ‘chipmaping8’, ‘chipmaping9’, ‘chipmaping10’, ‘chipmap-

ing11’, ‘chipmaping12’, ‘chipmaping13’, ‘chipmaping14’, ‘chipmaping15’, ‘chipmapping16’

represents chip mapping output of ‘flag_2Mbps’ signal with all possible chip sequences of

corresponding transmitted symbol and received spread spectrum signal of 2 Mbps data rate

named as ‘Received 2 Mbps DSS signal’ are also shown for 270000 ns simulation time. The

signals in Figure 4.9b displays clock signal of frequency 250 KHz named as ‘clk’, the re-

ceived symbol after chip to symbol mapping is designated by signal of 4 bit register named

as ‘Symbol_RNS’ and the received signal at receiver is termed as ‘Input data received’ for

270000 ns simulation time.

4.5 Implementation on VIRTEX 5 FPGA

The transceiver for RNS based PN sequence was next implemented on VIRTEX 5 FPGA

board. The system used the digital clock manager circuit presented in Figure 3.32 as used

82
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Symbol 0000 Chip Sequence 1 Counter1

Symbol 0001 Chip Sequence 2 Counter2

Symbol 0010 Chip Sequence 3 Counter3

•••••••
•••••••

•••••••
•••••••
Symbol 1111 Chip Sequence 16 Counter16

C 0 C 1 C 2 C 3 • • • • • C 15 C 16

Out stream after I & Q Combination

Figure 4.8: Decoding of RNS Based Transceiver

(a) Chip Mapping in modelsim for RNS

(b) Data bit received for RNS

Figure 4.9: Decoder for RNS based transceiver

83
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

in chapter 3. In this design a little extra logic is implemented in comparison to chapter

3 because here the PN sequence is used for Direct sequence spread spectrum baseband

modulation for both transmitter and receiver, It is found that most of synthesis report

result is similar to results reported in chapter 3 with little variation. Similar to chapter 3

different parameters of synthesis results for this design on FPGA presented here.

The synthesis options to fine-tune FPGA synthesis to meet design constraints of RNS

based baseband transceiver is listed in Table 4.3. The device utilization summary analyzed

in Table 4.4 which estimates the number of slices, buffers, DCM etc of VIRTEX 5 FPGA

board and the Table 4.5 represents advance synthesis report.

The implemented design in VIRTEX 5 LX110T FPGA board has maximum clock

frequency of 11.792 MHz and maximum output required time is 3.264 ns. The total logic

power consumed by the design is 0.45 mW.

4.5.1 Synthesis Report

The RNS based transceiver design is synthesized in ISE design environment which is re-

ported as follows. It is observed that, in this design there is some differences in slice

logic distribution and advance synthesis report as presented below. After synthesis the

programming bit file is generated which is downloaded to Virtex 5 LX110T FPGA. The

FPGA board internal signals along with received baseband data was analyzed through

ChipScope Pro analyzer software of XILINX synthesizer. The internal I-phase, Q-phase,

2 Mbps serial signal at receiver and baseband data bit received signal are observed as

depicted in Figure 4.10.

4.6 Summary and Discussion

ZigBee baseband transceiver for RNS based PN sequence was implemented through various

stages. The contribution from the chapter can be listed as under.

• The new RNS based PN sequence was used in DSSS transmitter and receiver for the

design, this was further simulated in modelsim simulator.

84
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Table 4.3: Synthesis options summary

Synthesis options summary

Input File Name "datain_tb_RNS.prj"
Input format mixed
Ignore Synthesis Constraint File No
Output File name "dataint b”
Automatic FSM Extraction YES
FSM Encoding Algorithm No
FSM Style LUT
RAM Extraction Yes
RAM Style Auto
Rom Extraction Yes
Mux style Auto
Decoder Extraction YES
Priority Encoder Extraction Yes
Shift Register Extraction Yes
Logical Shifter Extraction Yes
Mux Extraction Yes
Resource Sharing YES
Use DSP Block Auto

Table 4.4: Device Utilization Summary

Slice Logic Distribution

Number of Slice Registers 988 out of 69120 1%
Number of Slice LUTs 1186 out of 69120 1%
Number used as logic 1185 out of 69120 1%
Number used s Memory 1 out of 17920 0%

I/O Utilization
Number of bonded IOBs 8 out of 640 1%
Number of BUFG/BUFGCTRLs 10 out of 32 31%
Number of DCM_ADVs 3 out of 12 33%

85
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

Table 4.5: Advance HDL Synthesis Report

Advance HDL Synthesis Report

ROMs 1
128x1-bit ROM 1
Adders/Substracters 32
6-bit adder 32
Counters 6
2-bit up counter 1
4-bit up counter 2
5-bit up counter 2
7-bit up counter 1
Registers 717
Flip-Flops 717
Latches 551
Comparators 58
Multiplexers 1

Figure 4.10: Chipscope Pro Internal signal for RNS based Baseband Receiver

86
 CHAPTER 4. ZIGBEE TRANSCEIVER DESIGN WITH RNS BASED PN
4 SEQUENCE

• The baseband module is designed in hardware description language verilog HDL and

simulated in mentor graphics modelsim simulator and the signal at each stage is

analyzed.

• In FPGA implementation, similar to chapter 3 the synchronous clock circuitry is

designed for 250 KHz, 10 MHz, and 1 MHz by taking 100 MHz board frequency from

AH 15 PIN. Further, the implemented designs synthesis report along with some of

internal signals of the design are also analyzed for verification.

• In XILINX power analysis, the total logic power consumption for RNS based base-

band transceiver is found to be same as 0.45mW.

87
 5
Conclusions and Future Scope of Work

This thesis presents the design and implementation of ZigBee transceiver for IEEE 802.15.4.

It has been aimed to achieve a complete ZigBeee transceiver for short range wireless com-

munication keeping in mind for military application. Each standard supporting short range

wireless communication has different specifications and different applications. These days

mobile communication has achieved wide applicability, but for application of short range

communication in isolated area mobile communication is very difficult to implement. Zig-

Bee transceiver provides a resilient technology to communicate in shorter range.

5.1 ZigBee Transceiver Design using FPGA

A FPGA design for ZigBee transceiver has been completed in this thesis. The development

was completed through following steps

• Simulation of transceiver performance was carried out using Simulink. The desired

signals of Simulink model of transceiver was analyzed during performance analysis.

• The baseband transceiver is described in hardware description language Verilog HDL

for simulation in mentor graphics modelsim simulator.

88
5 CHAPTER 5. CONCLUSIONS AND FUTURE SCOPE OF WORK

• Following this the simulated design of transceiver was synthesized by using ISE design

environment to implement on Virtex 5 FPGA. Further the board synthesis report

and FPGA internal signals are analyzed. From analysis it is reported that the imple-

mented design consumes 0.45 mW logic power at maximum clock frequency of 11.792

MHz and transportation delay from input to output is 3.264 ns.

In second phase of the design, a new proposed set of PN sequence is generated based on

RNS which is used for baseband transmission and reception. Further the BER performance

analysis of proposed RNS based PN sequence is compared with Pseudo Noise sequence for

different noise conditions. From analysis it is seen that the RNS based receiver performs

very similar to PN based ZigBee, however at low SNR the proposed RNS based coding

provides better performance. The RNS based PN sequence is used as spreading code in

transceiver implemented on Virtex 5 FPGA board in ISE design environment.

5.1.1 Design Discussion

Physical layer in ZigBee inter operates with the other upper layers of networking protocol

such as MAC layer, Network layer and application layer etc. ZigBee devices used in upper

layer are based on IEEE 802.15.4 standard operate in mesh networking topology with line

of sight (LOS) communication. The underlying IEEE 802.15.4 standard provides strong

reliability through different mechanisms at multiple layers. In line of this the design of the

physical layer developed in this thesis is expected to work with other layers.

5.2 Limitation of work

The work reported in this thesis has certain limitations. some of the limitations are listed

below.

• The thesis analyzed a baseband transceiver design suitable for 2.4 GHz RF band. It

does not take care of other frequency bands which include 868 MHz and 915 MHz.

89
5 CHAPTER 5. CONCLUSIONS AND FUTURE SCOPE OF WORK

• The receiver assumes AWGN model for the channel. Multipath effect and fading

have not been considered in this analysis.

• The clock design in this system has been developed on the FPGA hardware. There

may be difficulties associated when the transceiver is interface with other subsystem

having their independent clocks.

5.3 Future Scope

Wireless communication is a fast evolving field. Technology changes have been very rapid.

In view of this and limitations of work in thesis following are some of the directions for

future research.

• Designing a multi band ZigBee transceiver covering 868 MHz, 915 MHz and 2.4 GHz

can be useful from system application point of view.

• The system RF design possess an important challenge in any hardware development.

Integrating the proposed baseband transceiver with a RF stage would provide a

complete hardware design.

• The last step in future work can be consideration of chip fabrication based on a

complete design of a ZigBee system.

90
Bibliography

[1] IEEE Standard for Local and Metropolitan Area Networks-Part 15.4: Low-Rate Wire-
less Personal Area Networks (LR-WPANs), IEEE Computer Society Std.

[2] S. Farahani, ZigBee Wireless Network and Transceivers. Newnes, USA, 2008.

[3] S. Bluetooth, “Bluetooth specification,” 2007.

[4] Z. Alliance, “ZigBee Alliance,” WPAN industry group, The industry group responsible
for the ZigBee standard and certification, 2010.

[5] IEEE Standard for Local Area Networks-Part 11: Wireless Local Area Networks, IEEE
Computer Society Std.

[6] J.-S. Lee, Y.-W. Su, and C.-C. Shen, “A Comparative Study of Wireless Protocols:
Bluetooth, UWB, ZigBee, and Wi-Fi,” Taipei, Taiwan, 2007, pp. 46 – 51.

[7] S. Chessa, P. Baronti, P. Pillai, V. Chook, A. Gotta, and Y. Hu, “Wireless Sensor
Networks: A Survey on the State of the Art and the 802.15.4 and ZigBee Standards,”
Computer Communications, vol. 30, no. 7, pp. 1655 – 1695, 2007.

[8] K. Coffman, Real World FPGA Design with Verilog, ser. Prentice Hall Modern Semi-
conductor Design Series. Prentice Hall PTR, 2000.

[9] R. Ahmad, O. Sidek, W. Hassin, and S. Mohd, “Implementation of IEEE 802.15.4-

Based OQPSK-Pulse-shaping Block on FPGA,” Piscataway, NJ, USA, 2011, pp. 459
– 64.

[10] R. Chithra, P. Maji, S. Patra, and G. Rath, “A PN Sequence Generator Based on

Residue Arithmetic for Multi-User DS-CDMA Applications,” pp. 26–31, June 2012.

[11] N.-J. Oh and S.-G. Lee, “Building a 2.4-GHz Radio Transceiver using IEEE 802.15.4,”
IEEE Circuits and Devices Magazine, vol. 21, no. 6, pp. 43 – 51, Jan.-Feb. 2005.

[12] C.-C. Wang, J.-M. Huang, C.-Y. Chang, K.-T. Cheng, and C.-P. Li, “A 6.57 mW
ZigBee Transceiver for 868/915 MHz Band,” in Circuits and Systems, 2006. ISCAS
2006. Proceedings. 2006 IEEE International Symposium on, may 2006, pp. 4 pp. –
5198.

91
 BIBLIOGRAPHY

[13] C. Wang, C. Huang, J. Huang, C. Chang, and C. Li, “ZigBee 868/915-MHz Modula-
tor/Demodulator for Wireless Personal Area Network,” Very Large Scale Integration
(VLSI) Systems, IEEE Transactions on, vol. 16, no. 7, pp. 936–939, 2008.

[14] C. Wang, J. Huang, L. Lee, S. Wang, and C. Li, “A Low-power 2.45 GHz ZigBee
Transceiver for Wearable Personal Medical Devices in WPAN,” in International Con-
ference on Consumer Electronics (ICCE). IEEE, 2007, pp. 1–2.

[15] P. Maji and G. Rath, “A Novel Design Approach for Low pass Finite Impulse Re-
sponse Filter Based on Residue Number System,” in Electronics Computer Technology
(ICECT), 2011 3rd International Conference on, vol. 3. IEEE, 2011, pp. 74–78.

[16] A. Omondi and A. Benjamin Premkumar, Residue Number Systems: Theory and
Implementation, ser. Advances in Computer Science and Engineering Texts. Imperial
College Press, 2007.

[17] S. Zhang, L.-L. Yang, and Y. Zhang, “Redundant Residue Number System Assisted
Multicarrier Direct-Sequence Code-Division Dynamic Multiple Access For Cognitive
Radios,” IEEE Transactions on Vehicular Technology, vol. 61, no. 3, pp. 1234 –1250,
2012.

[18] C. Wang, G. Sung, J. Huang, L. Lee, and C. Li, “A Low-power 2.45 GHz WPAN
Modulator/Demodulator,” Microelectronics Journal, vol. 41, no. 2, pp. 150–154, 2010.

[19] S. Fang, S. Berber, A. Swain, and S. Ur Rehman, “A Study on DSSS Transceivers

using OQPSK Modulation by IEEE 802.15.4 in AWGN and Flat Rayleigh Fading
Channels,” Fukuoka, Japan, 2010, pp. 1347 – 1351.

[20] K. Chen and H. Ma, “A Low Power ZigBee Baseband Processor,” in International
SoC Design Conference, (ISOCC’08)., vol. 1. IEEE, 2008, pp. I–40–I–43.

[21] A. Raut and L. Malik, “ZigBee Based Industrial Automation Profile for Power Moni-
toring Systems,” International Journal on Computer Science and Engineering, vol. 3,
no. 5, pp. 2028–2033, 2011.

[22] W. Kluge, F. Poegel, H. Roller, M. Lange, T. Ferchland, L. Dathe, and D. Eggert,

“A Fully Integrated 2.4-GHz IEEE 802.15.4-Compliant Transceiver for ZigBee Ap-
plications,” IEEE Journal of Solid-State Circuits, vol. 41, no. 12, pp. 2767 – 2775,
2006.

[23] Y. Yu, X. Meng, S. Xiao, C. Ma, and T. Ye, “A New Low-cost Demodulator for 2.4
GHz ZigBee Receivers,” Journal of Electronics, vol. 26, pp. 252 – 258, 2009.

[24] J. Palepu, “Design of ZigBee Transmitter for IEEE 802.15.4 Standard,” Master’s the-
sis, NIT Rourkela, Rourkela, Odisha-769008, INDIA, May 2011.

92
 BIBLIOGRAPHY

[25] S. Ahamed, “The Role of ZigBee Technology in Future Data Communication System,”
Journal of Applied Information Technology, vol. 5, no. 2, pp. 129 – 135, 2009.

[26] J. Adams, “An Introduction to IEEE STD 802.15. 4,” in Aerospace Conference, 2006
IEEE. IEEE, 2006, pp. 8–pp.

[27] C. Suh, Z. Mir, and Y. Ko, “Design and Implementation of Enhanced IEEE 802.15.4
for Supporting Multimedia Service in Wireless Sensor Networks,” computer networks,
vol. 52, no. 13, pp. 2568–2581, 2008.

[28] D. Zito, D. Pepe, and B. Neri, “Low-power RF Transceiver for IEEE 802.15. 4 (Zig-
Bee) Standard Applications,” in 13th IEEE International Conference on Electronics,
Circuits and Systems (ICECS’06). IEEE, 2006, pp. 1312–1315.

[29] B. Crow, I. Widjaja, L. Kim, and P. Sakai, “IEEE 802.11 Wireless Local Area Net-
works,” Communications Magazine, IEEE, vol. 35, no. 9, pp. 116–126, 1997.

[30] J.-S. Lee, Y.-W. Su, and C.-C. Shen, “A Comparative Study of Wireless Protocols:
Bluetooth, UWB, ZigBee, and Wi-Fi,” Piscataway, NJ, USA, 2007, pp. 46 – 51.

[31] S. Ondrej, B. Zdenek, F. Petr, and H. Ondrej, “ZigBee Technology and Device Design,”
in Networking, International Conference on Systems and International Conference on
Mobile Communications and Learning Technologies, 2006. ICN/ICONS/MCL 2006.
International Conference on. IEEE, 2006, pp. 129–129.

[32] S. Safaric and K. Malaric, “ZigBee Wireless Standard,” in 48th International Sym-
posium ELMAR-2006 focused on Multimedia Signal Processing and Communications.
IEEE, 2006, pp. 259–262.

[33] C. Wang, K. Sohraby, R. Jana, L. Ji, and M. Daneshmand, “Voice Communications

over ZigBee Networks,” Communications Magazine, IEEE, vol. 46, no. 1, pp. 121–127,
2008.

[34] W. Guo, W. M. Healy, and M. Zhou, “Battery Discharge Characteristics of Wireless

Sensors in Building Applications,” Beijing, China, 2012, pp. 133 – 138.

[35] M. Goyal, D. Rohm, W. Xie, S. Hosseini, K. Trivedi, Y. Bashir, and A. Divjak, “A

Stochastic Model for Beaconless IEEE 802.15. 4 MAC Operation,” Computer Com-
munications, vol. 34, no. 12, pp. 1460–1474, 2011.

[36] R. Ahmad, O. Sidek, H. Hassin, S. Korakkottil, K. Mohd, and A. Husain, “Verilog-

Based Design and Implementation of Digital Transmitter for ZigBee Applications,”
Int. J. Emerg. Sci, vol. 1, no. 4, pp. 723–734, 2011.

[37] D. Torrieri, Principles of Spread-Spectrum Communication Systems. Springer New

York, 2011.

93
 BIBLIOGRAPHY

[38] H. Taub and D. Schilling, Principles of Communication Systems. McGraw-Hill, 1999.

[39] R. Woods, J. Allister, Y. Yi, and G. Lightbody, FPGA-Based Implementation of Signal

Processing Systems. Wiley, 2008.

[40] Virtex-5 FPGA User Guide, XILINX Std.

[41] Virtex-5 FPGA Packaging and Pinout Specification, XILINX Std.

[42] VIRTEX 5 FPGA Board Data Sheet, XILINX Std.

[43] P. Mohan, Residue Number Systems: Algorithms and Architectures. Springer, 2002.

[44] C. Ding, D. Pei, and A. Salomaa, Chinese Remainder Theorem: Applications in Com-
puting, Coding, Cryptography. World Scientific, 1996.

[45] M. Lu, Arithmetic and Logic in Computer Systems. Wiley, 2004.

94
Publication

1. Bijaya Kumar Muni, Sarat Kumar Patra, “FPGA Implementation of ZigBee Base-
band Transceiver system for IEEE 802.15.4 ," ICAC3-2013, MUMBAI.
2. Bijaya Kumar Muni, Sarat Kumar Patra, “Low Bit Error Rate ZigBee Baseband
Transceiver Design for IEEE 802.15.4 ," MECON-2013, Amity university, NOIDA.

95
 BIO-DATA
Bijaya Kumar Muni
Date of Birth: 12th July, 1980

Correspondence:
M.Tech (Research), Department of Electronics and Communication Engineering,
National Institute of Technology Rourkela, India – 769 008.

Ph: +91 96924 37567 (M)

e-mail: [email protected], [email protected]

Qualification

• M.Tech (Continuing)
National Institute of Technology Rourkela, Odisha, India
• B.Tech (Electronics and Communication Engineering) [First division]
Biju Patnaik University of Technology, Odisha, India
• Diploma [First class]
U.G.M.I.T, Rayagada, Odisha, India
• 10th [First division]
Board of Secondary Education, Odisha, India

Publications

• 1 Journal Articles
• 1 Conference Articles

96