Muhammad Najib is Mail Med 2005 t Ttt

7/30/2019 Muhammad Najib is Mail Med 2005 t Ttt

1/138


2/138

I hereby declared that I have read this thesis and in my opinion this thesis is sufficientin terms of scope and quality for the award of the degree of Master of Engineering

(Electrical-Electronics & Telecommunication)

Signature : __________________________

Name : Prof. Dr. Tharek bin Abd Rahman

Date : 4th

April 2005


3/138

i

PERFORMANCE STUDY ON HIGH DATA RATES

MODULATION TECHNIQUES OF W-CDMA IN MULTIPATHFADING CHANNEL

MUHAMMAD NAJIB BIN ISMAIL

A project report submitted in partial fulfillment of the

requirements for the degree of

Master of Engineering (Electrical-Electronics & Telecommunication)

Faculty of Electrical Engineering

University Technology Malaysia

APRIL 2005


4/138

ii

I declare that this project report entitled Performance Study on High Data Rate

Modulation Techniques of WCDMA in Multipath Fading Channel is the result of my

own research except as cited in the references. The project report has not been accepted

for any degree and is not concurrently submitted in candidature of any other degree.

Signature : __________________________

Name : MUHAMMAD NAJIB BIN ISMAIL

Date : 4th

April 2005


5/138

iii

DEDICATION

To my dearest wife, son and parents


6/138

iv

ACKNOWLEDGEMENT

I would like to express my gratitude and appreciation to my project supervisor

Professor Dr. Tharek Abdul Rahman for all his guidance, helps and patience during the

course of this project.

Moreover, I extend my gratitude to my friends Azlin Mohd Fahmi and Maslinda

Rasli for their opinions, advices and thorough discussions for making this project well

organized, efficient and successful.

Last but not least, I would like to say thank you to my family especially to my wife

for her patience and advices throughout my course of work on this project.


7/138

v

ABSTRACT

Downlink transmission (base station to mobile terminal) using high data rate M-

ary Quadrature Amplitude Modulation (QAM) and Quadrature Phase Shift Keying

(QPSK) modulation schemes are considered in a Wideband-Code Division Multiple

Access (W-CDMA) system. The performances of these modulation techniques are

evaluated when the system is subjected to a number of users as well as noise and

interference in the channel. Additive White Noise Gaussian (AWGN) and multipath

Rayleigh fading are considered in the channel. Computer simulation tool, MATLAB,

will be used throughout the research to evaluate Bit-Error-Rate (BER) for W-CDMA

system models. Two approaches are used in this simulation. They are simulations using

Simulink and simulations using M files. A study of different modulation techniques is

needed so that a W-CDMA system can choose suitable modulation technique to suit the

channel quality, thus delivering optimum and efficient data rate to mobile terminal. It is

discovered that the performance of 16-QAM is significantly degraded in AWGN and

multipath Rayleigh fading channel compared to that of QPSK. Error correction coding is

needed to be used in this system particularly with 16-QAM to ensure better performance

of WCDMA system.

Index Terms Multipath Rayleigh fading, AWGN, Direct Sequence Spread Spectrum

(DSSS), Code Division Multiple Access (CDMA), BER, signal-to-noise ratio (SNR),

QPSK and 16-QAM


8/138

vi

ABSTRAK

Projek ini melibatkan kajian mengenai transmisi jalinan ke bawah (downlink)

(stesen tapak ke terminal boleh gerak) yang melibatkan penggunaan teknik-teknik

modulasi QPSK dan 16-QAM di dalam Pemodulatan Pembahagian Kod Pelbagai

Capaian Lebar Jalur Luas (Wideband Code Division Multiple Access, WCDMA).

Prestasi teknik-teknik modulasi ini dinilaikan ke atas sistem WCDMA yang dikenakan

hingar dan interferen pada saluran (channel) sistem ini. Pertambahan Bunyi Hingar Putih

Gaussian (AWGN) dan kelenturan pelbagai laluan Rayleigh (multipath Rayleigh fading)

dipilih untuk digunakan pada saluran di dalam sistem ini. Di dalam kajian ini, programcomputer MATLAB telah digunakan untuk mensimulasikan sistem WCDMA untuk

menilai kadar kesilapan bit di dalam sistem WCDMA. Dua kaedah telah digunakan iaitu

simulasi menggunakan Simulink dan simulasi menggunakan file M. Penyelidikan

terhadap teknik-teknik modulasi adalah diperlukan bagi sistem WCDMA supaya teknik

modulasi yang sesuai dapat digunakan secara dinamik oleh sistem ini supaya ia dapat

disesuaikan dengan keaadaan saluran. Kaedah ini adalah untuk memastikan

penghantaran data daripada stesen tapak ke terminal boleh gerak adalah pada tahap yang

laju, efisyen dan optimum. Keputusan simulasi computer ini telah menunjukkan

penurunan prestasi bagi teknik modulasi 16-QAM di dalam saluran AWGN dan

kelenturan pelbagai laluan Rayleigh jika ia dibandingkan dengan QPSK. Kod

pembetulan kesilapan adalah diperlukan di dalam sistem ini terutamanya jika 16-QAM

digunakan untuk memastikan sistem WCDMA berada pada keadaan yang baik.


9/138

vii

TABLE OF CONTENT

ACKNOWLEDGEMENT iv

ABSTRACT v

ABSTRAK vi

TABLE OF CONTENT vii

LIST OF TABLES xi

LIST OF ABBREVIATION xv

LIST OF APPENDIX xvi

1 INTRODUCTION 1

1.1 Background of the Problem 1

1.2 Problem Statements 2

1.3 Project Objective 2

1.4 Scope of Work 4

1.5 Significant of the Project Research 5

2 MODULATION SCHEMES IN WCDMA 7


10/138

viii

2.1 Bit Rate and Symbol Rate 9

2.2 Quadrature Phase Shift Keying (QPSK) 10

2.3 M-ary Quadrature Amplitude Modulation (QAM) 11

2.4 Wideband-Code Division Multiple Access (W-CDMA) 12

2.4.1 Direct Sequence Spread Spectrum (DSSS) 12

2.4.2 Code Division Multiple Access (CDMA) 15

2.5 Noise and Interference 15

2.5.1 Additive White Noise Gaussian (AWGN) 16

2.5.2 Rayleigh Fading 17

2.6 Bit Error Rate (BER) 19

2.7 Signal-to-Noise Ratio (SNR) 20

2.8 DSSS-CDMA Bit-Error Probability Calculations 20

2.9 Theoretical DSSS-CDMA System and Channel Models 21

2.9.1 Transmitter Model 21

2.9.2 Receiver Model 232.9.3 Channel Model 23

2.9.3.1 AWGN 23

2.9.3.2 Rayleigh Fading 25

3 CONFIGURATIONS ON WCDMA SYSTEM 28

3.1 Simulation Methodology 29

3.2 Simulation Using Simulink 30

3.2.1 Simulation in Phase 1: WCDMA System in AWGN Channel 31

3.2.1.1 Assumptions in Phase 1 31

3.2.1.2 Transmitter Design 35

3.2.1.2.1 User Data Sequence Generator 35

3.2.1.2.2 Spreading Sequence Generator 36

3.2.1.2.3 Spreader 38


11/138

ix

3.2.1.3 Modulation Techniques 39

3.2.1.3.1 QPSK Modulator 39

3.2.1.3.2 16-QAM Modulator 42

3.2.1.4 Channel Design 43

3.2.1.5 Receiver Design 43

3.2.1.5.1 QPSK Demodulator 43

3.2.1.5.2 16-QAM Demodulator 45

3.2.1.6 Despreader 45

3.2.1.7 Error Rate Calculation 46

3.2.1.8 Display 473.2.1.9 Performance Analysis for Phase 1 48

3.2.2 Simulation Phase 2: WCDMA system in AWGN and Multipath Rayleigh

Fading 50

3.2.2.1 Channel 53

3.2.2.2 Performance Analysis for Phase 2 55

3.3 Simulation Using M file 56

3.3.1 Generation of Spreading Code 56

3.3.2 Code Generation by Linear Feedback Shift Register 58

3.3.3 M-Sequence 59

3.3.4 Configuration of Transmitter and Receiver 61

3.3.5 Steps Taken to Realize the Simulation in dscdma.m file 66

3.3.6 Assumption and Limitation 67

4 PERFORMANCE ANALYSIS ON WCDMA SYSTEM 68

4.1 Simulation Using Simulink 69

4.1.1 Performance Analysis of QPSK modulation technique of WCDMA in

AWGN 69


AWGN and Multipath Fading Channel 71


12/138

x

4.1.3 Performance Analysis of 16-QAM modulation technique of WCDMA in

AWGN 74



4.2 Simulation Using M files 78


AWGN 78



4.2.3 Performance Analysis Comparison of QPSK modulation technique of

WCDMA Between AWGN and Rayleigh Fading Channel 83


AWGN 89



4.3 Analysis and Discussion 90

5 CONCLUSION 92

5.1 Conclusion 92

5.2 Suggestion for Future Work 93

REFERENCES 95

APPENDIX 98


13/138

xi

LIST OF TABLES

Table no. Title Page no.

3.1 Parameters for Bernoulli Binary Generator Block 35

3.2 Parameters used in PN Sequence Generator Block 36

3.3 Parameters used in QPSK Modulator Passband Block 40

3.4 Parameters used in M-QAM modulation block 42

3.5 Parameters used in AWGN block 43

3.6 Parameters used in QPSK Demodulator Passband Block 44

3.7 Parameters used in 16-QAM Demodulator Passband Block 45

3.8 Parameters used in Error Rate Calculation Block 46

3.9 Parameters used in Display Block 48

3.10 Parameters used in multipath Rayleigh fading channel 53

4.1 Simulation result for evaluation on BER vs. SNR for 2-ray

AWGN channel for 1 user when the number of data is

200,000

78

4.2 Simulation results for evaluation on BER vs. SNR for 2-ray

Multipath Rayleigh Fading channel for 1 user when the

number of data is 200,000 at 60 kmph

80


14/138


15/138

xiii

LIST OF FIGURES

FIGURE NO. TITLE PAGE NO.

2.1 Constellation diagram of a QPSK system 12

2.2 Constellation diagram of a 16-QAM system 12

2.3 CDMA 15

2.4 Relationship among channel correlation function and power

density function

18

3.1 Simulation process for W-CDMA system models 30

3.2 WCDMA Model using QPSK modulation technique in

AWGN channel

33

3.3 W-CDMA Model using 16-QAM modulation technique in

AWGN and multipath fading channel

34

3.4 W-CDMA Model with Multipath Raleigh fading channel andAWGN channel using QPSK Modulation Technique

51

3.5 W-CDMA Model using 16-QAM in AWGN and Multipath

Raleigh fading channel

52

3.6 Three-stage M-sequence 61

3.7 WCDMA system configured using m files 62


16/138

xiv

4.1 Performance of WCMA system using QPSK in AWGN

channel

70

4.2 Performance of WCDMA system using QPSK in multipath

fading channel at 60 kmph

71


fading channel for 90 kmph

72


fading channel at 120 kmph

73

4.5 Performance of WCDMA system using 16-QAM in AWGN 74

4.6 Performance of WCDMA system using 16-QAM in

multipath fading channel at 60 kmph

75



76



77

4.9 Performance of WCDMA in 2-Rays AWGN Channels for 1user

79

4.10 Performance of WCDMA in 2-Rays Multipath Rayleigh

Fading Channels for 1 user

82

4.11 Performance Comparison of WCDMA in 2-Rays Between

AWGN and Multipath Rayleigh Fading Channels for 1 user

85

4.12 Performance Comparison of WCDMA in 2-Rays Between

AWGN and Multipath Rayleigh Fading Channels for 5 user

88

4.13 Performance Comparison of 16-QAM in WCDMA system in

AWGN channel

89


17/138

xv

LIST OF ABBREVIATION

WCDMA Wideband Code Division Multiple Access

UMTS Universal Mobile Telecommunication System

GMSK Gaussian Minimum Shift Keying

GSM Global System for Mobile Communication

AWGN Additive White Noise Gaussian Noise

QPSK Quadrature Phase Shift Keying

QAM Quadrature Amplitude Modulation

BER Bit Error Rate

SNR Signal to Noise Ratio

PN Pesudo-Noise

AMC Adaptive Modulation and Coding

HSDPA High Speed Downlink Packet Access

PDF Probability Density Function

dB Decible

ISI Inter-Symbol Interference


18/138

xvi

LIST OF APPENDIX

APPENDIX

NO.

TITLE PAGE NO.

1.1 Matlab Source Codes for Simulation Using Simulink 98

1.1.1 Generic Source Codes of Simulation for QPSK of WCDMA

system either in AWGN or Multipath Fading channel or both

98

1.1.2 Generic Source Codes of Simulation for 16-QAM of

WCDMA system either in AWGN or Multipath Fading

channel or both

99

1.2.1 Source Codes for Simulation of Sub-System of WCDMA 100

1.2.1.1 Source Codes for Simulation of Autocorrelation Function of

a Sequence

100

1.2.1.2 Source Codes for Simulation of Cross-correlation Function

of a Sequence

100

1.2.1.3 Source Codes for Simulation of Generation Function of M

Sequence

101

1.2.1.4 Source Codes for Simulation of Shifting the Contents of the

Register

102

1.2.1.5 Source Codes for Simulation of Data Spread Function 104

1.2.1.6 Source Codes for Simulation of Data Despread Function 105


19/138

xvii

1.2.1.7 Source Codes for Simulation of a Function to Sample the

Time

106

1.2.1.8 Source Codes for Simulation of a Function to Add Gaussian

Noise

107

1.2.2 Source Codes for Simulation of Main-System of WCDMA 108

1.2.2.1 Source Codes for Simulation of the Main Program of DS-

WCDMA System

108

1.2.3 Source Codes for Simulation of BER vs EbNo of WCDMA

System

113

1.2.3.1 Source Codes for Simulation of QPSK of WCDMA Systemin AWGN Channel

113

1.2.3.2 Source Codes for Simulation of QPSK of WCDMA System

in Multipath Rayleigh Fading Channel with Doppler Shift

(60kmph, 90kmph & 120kmph)

113


for AWGN vs Multipath Rayleigh Fading Channel

115


for AWGN vs Multipath Rayleigh Fading Channel for a

Single User

117


for AWGN vs Multipath Rayleigh Fading Channel for a Five

(5) Users

118

1.2.3.6 Source Codes for Simulation of 16-QAM of WCDMA

System in AWGN Channel

119


20/138

CHAPTER 1

INTRODUCTION

1.1 Background of the Problem

W-CDMA system has been identified by Universal Mobile Telecommunication

System (UMTS) as the platform of the 3

rd

generation cellular communication system.Unlike conventional narrowband signal of 2

ndgeneration (2G) communication system,

W-CDMA uses noise-like broadband frequency spectrum where it has high resistance to

multipath fading. High data rate signal transmission can be transmitted over the air by

using W-CDMA system, thus enabling large data transmission of multimedia rich

applications such as high-resolution pictures and video to end-users. Thus, suitable

modulation technique and error correction scheme have to be used in W-CDMA system.

In 2G network, modulation scheme such as GMSK is widely used in Global System of

Mobile Communication (GSM). GMSK can only deliver data rate of 1 bit per symbol.

Obviously, such modulation scheme is not suitable for the next communication system.

Thus, there is a need to study the performance of new modulation technique that could

deliver higher data rate effectively in a multipath fading channel.


21/138

2

1.2 Problem Statements

To deliver multimedia content application over cellular networks, a high data rate

modulation scheme is one of the important criteria besides good error correction coding.

However, the implementation of high data rate modulation techniques that have good

bandwidth efficiency in W-CDMA cellular communication requires perfect modulators,

demodulators, filter and transmission path that are difficult to achieve in practical radio

environment. Modulation scheme that capable to deliver more bits per symbol is

susceptible to errors caused by noise and interference in the channel. Moreover, errors

can be easily produced as the number of users is increased and the mobile terminal is

subjected to mobility.

1.3 Project Objective

The objectives and aims of this project are to look at the performance of high

data rate modulation techniques at channels that are subjected to Additive White

Gaussian Noise (AWGN) and multipath Rayleigh fading. Modulation schemes that will

be considered in this project are Quadrature Phase Shift Keying (QPSK) and 16-ary

Quadrature Amplitude Modulation (16-QAM). This performance study will be carried

out by varying the chip rate of pseudo-noise (PN) generator. Furthermore, multiple

access scheme i.e. WCDMA will be also studied by comparing certain number of users

under static and mobility environment that are subjected to AWGN and multipath

Rayleigh fading. The performances of WCDMA under these channels fading are based

on Bit Error Rate (BER) at downlink (base station to mobile terminal) transmission.


22/138

3

There will be three WCDMA wireless cellular system models that will be used in

this project. The models are

1. WCDMA system in AWGN channel

2. WCDMA system in AWGN and Multipath Rayleigh Fading.

3. Multi-user WCDMA system in AWGN and Multipath Rayleigh Fading.

Relationship for multiple rays using QPSK and QAM in W-CDMA system models

for the followings parameters will be obtained using MATLAB. They are:

1. Bit Error Rate (BER) versus Signal-to-Noise ratio (SNR) in AWGN channel for

QPSK modulation technique.

2. BER versus SNR in AWGN channel for 16-QAM modulation scheme.

3. BER versus SNR in AWGN and multipath Rayleigh fading channel with Doppler

shift (60kmph, 90kmph and 120kmph) for QPSK modulation technique.


shift (60kmph, 90kmph and 120kmph) for 16-QAM modulation scheme.

5. BER versus SNR to compare between AWGN channel and multipath Raleigh

fading channel for different number of user for QPSK modulation technique.


fading channel for different number of user for 16-QAM modulation technique.

Once data for BER and SNR under various parameters are obtained, the data are

tabulated. Graphs of BER as a function of SNR under different modulation

techniques as well as different velocities of mobile terminal subjected to noise and

interference channel will be plotted. These graphs will be studied and compared so

that a conclusion on suitable high data rate modulation scheme can be drawn.


23/138

4

1.4 Scope of Work

This project is an entirely simulation project using scientific computer simulation

software, MATLAB 6.5.2. Two approaches will be used in this project. They are

simulation using Simulink and simulation using m files. It will be simulated in multi-

user environment based on Direct Sequence Spread Spectrum (DSSS), Wideband-Code

Division Multiple Access (W-CDMA). There will be no error correction coding or

channel coding employed for this simulation models.

There are two extreme cases of channel noise and fading that will be subjected

to the W-CDMA system models. Firstly, the model is simulated with different

modulation techniques under thermal noise, represented by Additive White Noise

Gaussian (AWGN). Then, the channel is simulated with various different parameters

using Non-Line of Sight (N-LOS) multiple reflected rays represented as multipath

Rayleigh fading.

The performance of the modulation schemes are studied when the mobile

terminal is static and mobile with different speeds. The performance measurement is

based on BER. Thus, suitable modulations techniques will be determined and concluded

based on BER that will be plotted as a function of function of SNR.


24/138

5

1.5 Significant of the Project Research

The current trend to achieve high data rate cellular communication drives the

interest of this research. There are many ways one can improve data rate in a W-CDMA

system. However, two significant areas that could give significant boost to the

improvement of W-CDMA system is modulation scheme and error correction or channel

coding.

There are many modulation schemes that have the potential to deliver higher data

rate but there is a trade off between data rate and multipath environment. Modulation

techniques that can deliver more bits per symbol normally generate lots of error when

they are subjected to multipath channels. Recently, there is intensifying research about

Adaptive Modulation and Coding (AMC) [14]-[16]. The principle of AMC is to change

the modulation and coding format (transport format) in accordance with instantaneous

variations in the channel conditions, subject to system restrictions. AMC extends the

systems ability to adapt to good channel conditions. Channel conditions should be

estimated based on feedback from the receiver. For a system with AMC, users closed to

the cell site are typically assigned higher order modulation with higher code rates (e.g.

64 QAM with R=3/4 Turbo Codes). On the other hand, users closed to the cell boundary,

are assigned lower order modulation with lower code rates (e.g. QPSK with R=1/2

Turbo Codes). AMC allows different data rates to be assigned to different users

depending on their channel conditions. Since the channel conditions vary over time, the

receiver collects a set of channel statistics which are used both by the transmitter and

receiver to optimize system parameters such as modulation and coding, signal

bandwidth, signal power, training period, channel estimation filters, automatic gain

control, etc [3].


25/138

6

Thus, this project will scrutinize suitable modulation techniques that are capable

to deliver highest data rate without compromising errors in multipath fading

environment. The performance of these modulation techniques will be simulated by

using computer simulation tool, MATLAB.


26/138

CHAPTER 2

MODULATION SCHEMES IN WCDMA

The evolution of wireless cellular technology from 1G to 3G has a similar aim

that is capable to deliver high data rate signal so that it can transmit high bit rate

multimedia content in cellular mobile communication. Thus, it has driven many

researches into the application of higher order modulations [1]-[4] and [17]-[18].

The current second generation Global System for Mobile Communication (GSM)

system provides data services with 14.4 kbps for circuit-switched data and up to 22.8

kbps for packet data. High-Speed Circuit Switched Data (HSCSD) and General Packet

Radio Services (GPRS) with multi-slot operation can only slightly increase the data rate

due to the Gaussian Minimum Shift Keying (GMSK) modulation. Enhance Data Rate for

the GSM Evolution (EDGE) is proposed as a transition to 3G as a new Time Division

Multiple Access (TDMA) based radio access using the current (800, 900, 1800 and 1900

MHz) frequency bands. EDGE enables significantly higher peak rates and approximately

triples the spectral efficiency by employing 8-Phase Shift Keying (8PSK) modulation.


27/138

8

WCDMA is another 3G-system operation in 5MHz bandwidth to support both

high-rate packet data and circuit-switched data. High Speed Downlink Packet Access

(HSDPA) is currently being developed as the evolution of WCDMA systems to

considerably increase the data rate by using adaptive modulation and coding (AMC),

hybrid automatic repeat request (HARQ), fast cell selection (FCS) and multiple input

multiple output (MIMO) antenna processing [4].

In cellular system, different users have different channel qualities in terms of

signal to noise ratio (SNR) due to differences in distance to the base station, fading and

interference. Link quality control adapts the data protection according to the channelquality so that an optimal bit rate is obtained for all the channel qualities [1-4]. Thus, the

system adopts AMC to suit the link quality. WCDMA systems can employ the high-

order modulation (8PSK or M-QAM) to increase the transmission data rate with the link

quality control.

However, there is a trade off in employing bandwidth efficient M-QAM

modulation scheme. The complexity of the receiver increases linearly with M (number

of orthogonal sequences) and exponentially with the number of bits per symbol. The

achievable bandwidth efficiency of the system is limited by the maximum possible

number of orthogonal sequences and by acceptable complexity of the receiver [2].

To minimize Inter-symbol Interference (ISI), noise and channel fading, a

wireless system needs to have a robust system to minimize, if not to eliminate, these

unfavorable effects. A typical W-CDMA transmitter system consists of bit generator, TC

encoder, rate matcher, interleaver, spreader, modulator, scrambler, and pulse shaper. On

the other hand, a receiver consists of a matched filter, channel estimator, rake receiver,

despreader, demodulator, deinterleaver, and TC decoder. Maximal ratio combining of

rake results amplitude boost is very favorable for M-PSK demodulation due to its greater

separation of the received symbol constellation. However, it is not the case for the M-

QAM. For an amplitude-modulated signal (M-QAM), amplitude change could produce

incorrect symbol detection [1].


28/138

9

2.1 Bit Rate and Symbol Rate

To understand and compare different modulation format efficiencies, it is

important to understand the difference between bit rate and symbol rate. The signal

bandwidth for the communications channel depends on the symbol rate or also known as

baud rate.

symbolperdtransmittebitsofnumber

ratebitrateSymbol (1)

Bit rate is the frequency of a system bit stream. For example, a radio with an 8-

bit sampler is sampled at 10 kHz for voice. The bit rate, the basic bit stream rate in the

radio, would be 8 bits multiplied by 10k samples per second giving 80 kbps. In this

example, extra bits required for synchronization, error correction, etc are ignored for

simplicity. In GMSK, only one bit can be transmitted for each symbol. Thus, the symbol

rate for this modulation technique is 80 kbps. However, high data rate like 8-PSK, as it

will be reviewed in the next section, can transmit 3 bits per symbol. Thus, the symbol

rate, if this modulation scheme is employed, is 26.7 kbps. The symbol rate for 8-PSK is

three times smaller than that of GMSK. In other words, 8-PSK or any high order (M)

modulation scheme can transmit same information over a narrower piece of RF

spectrum.


29/138

10

2.2 Quadrature Phase Shift Keying (QPSK)

QPSK is one example of M-ary PSK modulation technique (M = 4) where it

transmits 2 bits per symbol. The phase carrier takes on one of four equally spaced

values, such as 0, S/2, S and 3S/2, where each value of phase corresponds to a unique

pair of message bits as it is shown in figure 2.2. The basis signal for QPSK can be

expressed as

> @

-

tiEtiEts ssQPSK 21 1sin2

1cos IIS

i = 1,2,3,4

(3)

(1,1)

(-1,-1) (-1,1)

(1,-1)

Q

I

Figure 2.1: Constellation diagram of a QPSK system

Special characteristics of QPSK are twice data can be sent in the same bandwidth

compared to Binary PSK (BPSK) and QPSK has identical bit error probability to that of

BPSK. When QPSK is compared to that of BPSK, QPSK provides twice the spectral


30/138

11

efficiency with the same energy efficiency. Furthermore, similar to BPSK, QPSK can be

differentially encoded to allow non-coherent detection.

Due to these advantages of QPSK, it has been employed as the modulation

technique in UMTS 3G wireless cellular networks where the following data rate can be

achieved depending on the channel quality.

i. 144 kbps for high mobility

ii. 384 kbps for low mobility

iii. 2 Mbps for indoor or static environment.

2.3 M-ary Quadrature Amplitude Modulation (QAM)

QAM is a modulation technique where its amplitude is allowed to vary with

phase. QAM signaling can be viewed as a combination of Amplitude Shift Keying

(ASK) as well as Phase Shift Keying (PSK). Also, it can be viewed as ASK in two-

dimension. Figure 2.2 shows the constellation diagram of 16-ary QAM (16-QAM). The

constellation consists of a square lattice of signal points. The general form of an M-ary

signal can be defined as

tfb

T

Etfa

T

Ets ci

s

ci

s

i SS 2sin2

2cos2 minmin (4)

Ttdd0 i = 1,2, .,M

where Emin is the energy of the signal with the lowest amplitude and ai and bi are a pair of

independent integers chosen according to the location of the particular signal point.


31/138

12

Q

I

Figure 2.2: Constellation diagram of a 16-QAM system

Theoretically, higher order of M-ary QAM enables data to be transmitted in a

much smaller spectrum. However, the symbols are easily subjected to errors due to noise

and interference because the symbols are located very closed together in the

constellation diagram. Thus such signal has to transmit extra power so that the symbol

can be spread out more and this reduces power efficiency as compared to simpler

modulation scheme. Also the radio equipment is more complex.

2.4 Wideband-Code Division Multiple Access (W-CDMA)

2.4.1 Direct Sequence Spread Spectrum (DSSS)


32/138

13

A DSSS system spreads the baseband data by directly multiplying the baseband

data pulses with a pseudo-noise sequence that is produced by a pseudo-noise (PN) code

generator [5]. A PN sequence is a binary sequence with an autocorrelation that

resembles, over a period, the autocorrelation of a binary sequence. Its autocorrelation

also roughly resembles the autocorrelation of band-limited white noise. The PN

sequence is usually generated using sequential logic circuits (i.e. feedback shift register).

A single pulse or symbol of the PN waveform is called chip. Spread spectrum signals are

demodulated at receiver through cross-correlation with locally generated version of the

pseudorandom carrier. Cross-correlation with the correct PN sequence de-spreads the

spread spectrum signal and restores the modulated message in the same narrow band asthe original data, whereas cross-correlating the signal from an undesired user results in a

very small amount of wideband noise at the receiver output.

Unlike modulation and demodulation techniques that have primary objective to

achieve power and/or bandwidth efficiency in AWGN channel, the transmission

bandwidth of DSSS has several orders of magnitude greater than the minimum required

signal bandwidth. In other words, DSSS modulation transforms an information signal

into a transmission signal with a larger bandwidth. It is achieved by encoding the

information signal with a code signal that is independent of the data and has a much

larger spectral width than that of information signal. In DSSS, many users can

simultaneously use the same bandwidth without significantly interfering one another.

DSSS is normally used in Code Division Multiple Access (CDMA) scheme.

The received DSSS signal for a single user can be represented as

TS tftptmT

Ets c

s

s

ss 2cos2

(5)

where m (t) is the data sequence, p (t) is the PN spreading sequence, fc is the carrier

frequency and is the carrier phase angle at t= 0.


33/138

14

There are numerous advantages of DSSS for cellular radio system as they are

describe as follows:

1. DSSS has interference rejection capability since each user is assigned with a

unique PN code that is approximately orthogonal to the codes of other users.

2. Capable to resist radio jamming by a narrowband interferer.

3. DSSS eliminates the need of frequency planning since all cells can use the

same channels.

4. It has high resistance to multipath fading. Since DSSS signals have uniform

energy over large bandwidth, only a small portion of the spectrum will undergofading. The delayed version of PN sequence arrived at W-CDMA receiver will

have poor correlation with the original PN sequence and the receiver will

ignore it. This situation will occur even if the delay is only one chip form the

intended signal. In other words, the multipath signal would appear invincible to

the receiver.

5. Apart from resistance to multipath fading, DSSS can exploit the delayed

multipath components to improve the performance of the system. This can be

done by using RAKE receiver where it consists of a bank of correlators. Each

correlator will correlate to a particular multipath component of the desired

signal. The correlated outputs are weighted according to their strengths and

summed to obtain the final signal estimate.

Two conditions have to be satisfied for a technique to be classified as a spread spectrum

technique.

1. The transmission bandwidth must be larger than the information bandwidth.

2. The resulting radio-frequency bandwidth must be determined by a function

other than the information being sent. This excludes such modulation

techniques such as frequency modulation (FM) and (PM).


34/138

15

2.4.2 Code Division Multiple Access (CDMA)

CDMA is a multiple access scheme employed normally with DSSS. Each user

has a unique code that is orthogonal to one another. In CDMA, the power of multiple

users at a receiver determines the noise floor after decorrelation. CDMA can be viewed

with a figure 4.4 as it is shown below.

Code

Frequency

Channel 1

Channel 2

Channel 3

Channel N

Time

Figure 2.3: CDMA

2.5 Noise and Interference


35/138

16

2.5.1 Additive White Noise Gaussian (AWGN)

The term noise refers to unwanted electrical signals that are always present in

electrical systems [6]. The term additive means the noise is superimposed or added to the

signal that tends to obscure or mask the signal where it will limit the receiver ability to

make correct symbol decisions and limit the rate of information transmission. Thus,

AWGN is the effect of thermal noise generated by thermal motion of electron in all

dissipative electrical components i.e. resistors, wires and so on [6]. Mathematically,

thermal noise is described by a zero-mean Gaussian random process where the random

signal is a sum of Gaussian noise random variable and a dc signal that is

z= a + n (6)

where pdf for Gaussian noise can be represented as follows where 2 is the variance ofn.

2

2

1exp

2

1)(

VSV

azzp (7)

A simple model for thermal noise assumes that its power spectral density Gn(f) is

a flat for all frequencies and is denoted as

2)( 0

NfGn (8)

where the factor of 2 is included to indicate that Gn(f) is a two-sided power spectral

density. When noise power has such a uniform spectral density, it is referred as white

noise. The adjective "white" is used in the same sense as it is with white light, which


36/138

17

contains equal amounts of all frequencies within the visible band of electromagnetic

(EM) radiation.

Since thermal noise is present in all communication systems and is a prominent

noise source for most system, the thermal noise characteristics that are additive, white

and Gaussian are most often used to model the noise in communication systems.

2.5.2 Rayleigh Fading

Since signal propagation takes place in the atmosphere and near the ground, apart

form insignificant effect of free path loss, Ls, the most notable effect of signal

degradation is multipath propagation. The effect can cause fluctuations in the received

signal's amplitude, phase and angle of arrival, giving rise to terminology multipath

fading[7].

Generally, there are two fading effects in mobile communications: large-scale

and small-scale fading. Large-scale fading represents the average signal power

attenuation or path loss due to motion over large areas. On the other hand, small-scale

fading refers to the dramatic changes in signal amplitude and phase that can be

experienced as a result of small changes (as small as a half-wavelength) in the spatial

separation between a receiver and transmitter. Small-scale fading is also called Rayleigh

fading because the envelope of received signal can be represented by a Rayleigh pdf [7].The received signal consists of large number of multiple reflective paths and there is no

line-of-sight signal component. When there is a dominant non-fading signal component

present, such as a line-of-sight propagation path, the small-scale fading envelope is

described by a Rician pdf [5].


37/138

18

Figure 2.4: Relationship among channel correlation function

and power density function

TheDoppler spread is a measure of the spectral broadening due to the time rate

of change (time variant) of the channel parameters. Figure 2.4 (d) shows a Doppler

power spectral density, S(v), plotted as a function of Doppler-frequency shift, vbased on

dense-scatterer channel model. For the case of the dense-scatterer model, a vertical

receive antenna with constant azimuthally gain, a uniform distribution of signals arriving

at all arrival angles throughout the range (0,2p), and an unmodulated continuous wave

(CW) signal, the signal spectrum at the antenna terminals is


38/138

19

2

1

1)(

dd f

vf

vS

S

(9)

where fd is Doppler Spread andfc carrier frequency. The largest magnitude (infinite) of

S(v) occurs when the scatterer is directly ahead of the moving antenna platform or

directly behind it. Thus, from this situation, the magnitude of the frequency shift is given

by

OVfd (10)

where V is relative velocity and O is the signal wavelength. fd is positive when the

transmitter and receiver move towards each other, and negative when moving away from

each other. Equation 10 describes the Doppler frequency shift. In a typical multipath

environment, the received signal arrives from several reflected paths with different path

distances and different angles of arrival, and the Doppler shift of each arriving path is

generally different from that of another path. The effect on the received signal is seen as

a Doppler spreading or spectral broadening of the transmitted signal frequency, rather

than a shift. The Doppler power spectral density is infinite for Doppler components that

arrive at exactly 0q and 180q. Thus the angle of arrival is continuously distributed and

the probability of components arriving at exactly these angles is zero.

2.6 Bit Error Rate (BER)

BER is a performance measurement that specifies the number of bit corrupted or

destroyed as they are transmitted from its source to its destination. Several factors that


39/138

20

affect BER include bandwidth, SNR, transmission speed, transmission medium and

transmission speed.

2.7 Signal-to-Noise Ratio (SNR)

SNR is defined as the ratio between signal power to noise power and it is

normally expressed in decibel (dB). The mathematical expression of SNR is

dBpowernoise

powersignalSNR log10 (11)

2.8 DSSS-CDMA Bit-Error Probability Calculations

There are two approaches to calculate BER for DSSS-CDMA operating under

AWGN channel [8]-[10]. The first approach uses accurate BER approximations because

it is presumed that BER evaluation is numerically cumbersome. There are many

researches on this approach and most widely used approximation is the so-called

Standard Gaussian Approximation (SGA) [8]-[10]. In the SGA, a central limit theorem

(CLT) is employed to approximate the sum of the multiple-access interference (MAI)

signals as an AWGN process additional to the background Gaussian noise process. To

detect desired user signal, the receiver design consists of a conventional single-user

matched filter (correlation receiver). The average variance of the MAI over all possible

operating conditions is used to compute the SNR at the filter (correlator) output. SGA is

widely used because it is easy to apply. However, it is known based on performance

analysis that SGA often overestimate system performance especially for small number of

users. Thus, Improved Gaussian Approximation (IGA) is created to overcome the


40/138


41/138

22

)cos()()(2 kckkkk twtabtPs T (12)

where Pk represents transmitted signal power, bk(t) is data signal, ak(t) is spreading

signal, wc is carrier frequency and k is carrier phase. The kth users data signal is a

random process that is a rectangular waveform, taking values from with service rate, and

is expressed as

)()()(

jTtPbtb Tk

j

j

k

f

f

(13)

where PT(t) = 1, for Ttdd0 , andPT= 0, otherwise. Thejth data bit ofkth

user is

denoted as bj(k)

. Data source are assumed uniform, i.e.

2/111)()( kjr

k

jr bPbP . The spreading signal ak(t) can be expressed as

)()()(

lTtatak

l

l

k

f

f

\ (14)

where (t) is an arbitary chip waveform that is time-limited to [0,Tc) and Tc is chip

duration. Chip waveform is assumed to be normalized according to cT

Tdttc

)(02\ .

The lth

chip of the kth

user is denoted al(k)

, which assumes values from {-1,+1}. All

signature sequences {ak(k)

}are assumed to be random in the following sense. Every chip

polarity is determined by flipping an unbiased coin. Further justification for the random

chip sequence assumption is provided in. There areNchips for one data symbol and the

period of the signature sequence is N. We normalize the chip duration so that Tc=1 and,

thus, T=N. Note that if the chip waveform is rectangular, i.e.,

)()()(

cT

k

jjkjTtPata

c

f

fthe transmitted signal becomes the well-known phased-

coded SS model [9].


42/138

23

For QPSK modulation scheme, the transmitted signal of kth

user in the sub-

system i is

)cos()()(2)cos()()(2)( ikcQ

ik

Q

ikiikc

I

ik

I

ikiik ttctbPttctbPts TZTZ (15)

where )(tbIik and )(tbQ

ik are the In-phase and Quadrature-phase signal.

2.9.2 Receiver Model

The received signal r(t) at the input of the matched filter receiver is given by

)()cos()()(2)()()(11

tnttatbAPtnthstr kckkkkkk

K

k

kk

K

k

u

IZWW

(16)

where * denotes convolution and kckkk WZTEI is assumed a uniform random

variable over [0, 2). The average received power of the kth

signal isE[Pr] =E[A

2k]Pk.

2.9.3 Channel Model

2.9.3.1 AWGN

The transmitted signal for BPSK modulation is subjected to AWGN process n(t),

that has two-sided power spectral densityN0/2 andAk= 1, k=1, .,K. Ak is independent,


43/138

24

Rayleigh-distributed and account for the fading channel attenuation of all signal. The

first order of probability density function (pdf) is given by

)()( ),0[2/2 aIaeaP aAk f

(17)

Due to the fact that SGA considers an average variance value for Multi Access

Interference (MAI) or in other words, the first moment of, the IGA exploits knowledge

of all moments of. It was shown in [19] that the BER for an AWGN channel obtained

from IGA is significantly more accurate than the BER obtained from the SGA especially

for small number of user, k. Thus by applying SIGA, overall BER can be represented as

[13]

|

222 31

12

1

31

12

11

3

1

N

N

N

N

N

NP

SIGA

e

]]]] VPVPP]

(18)

where and 2

are given by

)1(3

2 K

N]P (19)

and

9

1)2(181843

45

1)1( 2

2 NKNNK]V (20)

where this Holtzmans method is extended by applying first and second moment for the

received power.


44/138

25

2.9.3.2 Rayleigh Fading

The output of a low pass filter (LPF) of a synchronous system i.e. W1 = W2 = =

Wkfor user 1 can be represented as

111

011 )cos()()(

nIS

dttwtatry cT

(21)

where n1 is a zero-mean Gaussian random variable with variance 4/02

1NNn V , S1 is

the signal component NAS 11 r , and the interference termI1 is given by

dttatabAIT

kk

kk

k

k )()()cos( 10

)(

0

2

1

I (22)

Since a sum of independent Gaussian random variable has Gaussian distribution, it

follows that I1 is a Gaussian random variable with zero-mean and

variance 212

22

1 k

K

k

I N UV

. By symmetry and using the independenceI1 and n1, one has

K

k k

SSYNC

Ae

NNN

NAQP

2

2

1

20

1

4

1

U

(22)

and averaging over the pdf of A1, BER for a Rayleigh-faded user is

K

k k

SYNC

e

N

NP

2

2

1

0

41

11

2

1

U

(23)


45/138

26

From the equation above, one sees that the interferers act like additional independent

Gaussian background noise. This is because the MAI on the flat Rayleigh fading channel

has a Gaussian first-order distribution assuming synchronous transmission. This implies

that the optimum receiver that does not perform user-interference cancellation is a

correlator detector. However, this is not the case of asynchronous transmission. For

uniformity, uniform random signature sequences NE k /1][2

1 U and

NK

NN

PSYNCe

14

1

11

2

1

0

(24)

In asynchronous transmission subjected to flat Rayleigh fading, average BER is

computed by using characteristic function, ). The proof for the following characteristic

function can be found in [8]. Average characteristic function of MAI Ik, given B, is

> @),1()1,()1,(,14

2

)()(

22

)1(

,

jiJjiJjiJjiJ

dSBS

B

Bj

Bj

A

Ai

Ai

N

kkBSI kk

u

))

ZZ

(25)

Using the fact that theIk's given B are independent, the characteristic function for

total interference termI, given B, is

))K

k

BIBI k2

)()( ZZ (26)

The conditional BER for target user, after averaging over pdf ofA1, can be expressed by

symmetry as


46/138

27

> @ ZZZZSV dNN

NP nBI

n

Be

-u))

22

22 2

1exp)()(12

112

11

1

(27)

When the effect of background noise is negligible, 021|nV and 11 |) n , thus equation

27 becomes

> @

ZZZS

ZZZZ

SdN

N

dNN

P

BI

nBIBe

-)

-u))

f

f

22

0

22

0

2

1exp)(

22

1

2

1exp)()(1

21

(28)

Equation (26), (27), and (25) [or (26), (28) and (25) for noiseless case] give the

average BER experience by a target user with a signature sequence that has a given

value of B. The average BER for all users of for one target user averaged over all

signature sequences randomly assigned by a base station for each request is

Be

N

B

N

B

N

e PP

1

0

1)1(2 (29)


47/138

CHAPTER 3

CONFIGURATIONS ON WCDMA SYSTEM

The research project is started with literature review on high data rate modulation

schemes, DSSS, W-CDMA, and fading effects on channel. Then, a generic model of

DSSS W-CDMA as it is shown in figure 3.1 is simulated using QPSK and it is followed

by QAM. QPSK and QAM are chosen in this project because there are the primary

candidates to deliver higher data rate for High Speed Downlink Packet Access

(HSDPA), an extension of 3G networks [14]-[16]. The simulation is done under noise

and multipath fading channel using MATLAB 6.5.2.

As it is shown in figure 3.1, the user data is assumed to be Bernoulli distributed

and it is represented as bn(t). Each user data is then multiplied with independent ordifferent PN code produced by a PN generator using XOR logical operator. The

multiplied signal of each user is represented as sn(t) after the signal is modulated by

either QPSK or QAM. Each signal is added before it is subjected to the channel. At the

receiver, the signalsk(t) is demodulated before the user data is separated from PN code

by XOR logical operator. Finally, when the necessary simulations are done, tables and

graphs of BER as a function of SNR for various parameters are plotted. Analysis,

comments and conclusion will be drawn based on the simulation results.


48/138

29

AWGN and Rayleigh fading are chosen to represent fading effect in the channel

because we want to make a comparison of WCDMA system models in two extreme

channel conditions. There are many fading effects that can be categorized as large-scale

and small-scale fading. Rayleigh fading represents the worst case of multipath fading

where it represents small-scale fading due to small changes in position. On the other

hand, AWGN represents the thermal noise generated by electrical instruments.

3.1 Simulation Methodology

Computer simulation is the most suitable, powerful and efficient way to represent the

actual or real situation of mobile radio system. Thus, MATLAB 6.5.2 has been identified

to simulate W-CDMA model based on related theories, formulae and parameters. Two

approaches are adopted in this project. Firstly, the simulation is simulated using

Simulink and it follows with simulation using m files. Throughout this project, the bit

rate for the signal generator is 384kbps.

There will be three WCDMA wireless cellular system models that will be used in

this project. The models are

1. WCDMA system in AWGN channel

2. WCDMA system in AWGN and Multipath Rayleigh Fading.3. Multi-user WCDMA system in AWGN and Multipath Rayleigh Fading.


49/138

30

WCDMA system in

AWGN

WCDMA system in

AWGN and multipath

Rayleigh fading

Multi-user WCDMA systemin AWGN and multipath

Rayleigh fading

QPSK

16-QAM

QPSK

16-QAM

QPSK

16-QAM

Stage 1

Stage 2

Stage 3

Test

Simulation

Check

for any

errors

Performance

analysis: BER

& SNR

Continue next

subsequent

stage

Figure 3.1: Simulation flow chart for W-CDMA system models used in Simulink and M

files

3.2 Simulation Using Simulink

Two types of simulation have been chosen to study the performance of

modulation techniques of WCDMA subjected to multipath fading in the channel. The

project begins with simulation using simulink. Simulink is a software package that has

the capabilities to model, simulate, and analyze dynamic systems whose outputs and

states change with time. Simulink can be used to explore the behavior of a wide range of


50/138

31

real-world dynamic systems making it suitable computer software to study the

performance of modulation techniques under multipath fading. Simulating a dynamic

system is a two-step process with Simulink. First, a graphical model of the system is

simulated, using the Simulink model editor. The model depicts the time-dependent

mathematical relationships among the system's inputs, states, and outputs. Then,

Simulink is used to simulate the behavior of the system over a specified time span.

Simulink uses information entered into the model to perform the simulation.

3.2.1 Simulation in Phase 1: WCDMA System in AWGN Channel

In Phase 1, both transmitter and receiver part are built based on the system model as

shown in Figure 3.5. The channel is subjected to AWGN only. This phase is divided

into five parts as follows:

1. Assumptions

2. Transmitter part

3. Receiver Part

4. Channel Part

5. Performance Analysis

3.2.1.1 Assumptions in Phase 1

The assumptions made for this phase of simulation are stated as follows:

x Evaluation of the performance is made on one user in the multi-user

environment. It considers the rest of the users contribute the multi user

interference to the reference user in the system.


51/138

32

x Decision statistic of the receiver, Z0 is modeled as Gaussian Random

Variable.

x Decision statistic of the desired user, I0 is deterministic.

x Multi-User Interference (MUI), ] in the system is assumed as zero-mean

Gaussian variables and it is an AWGN. This is based on the ARIB

proposal which states that all interference from other users is modeled as

Additive White Gaussian Noise (AWGN).

x Thermal noise K is very small and negligible. This is based on the data in

ARIB proposal.

x In this simulation, only downlink (base station to mobile station)transmissiont is considered.


52/138

33

F

igure3.2:W-CDMAModelusingQPSKmodulationtechnique

inAWGNchannel


53/138

34

Figure3.3:W

-CDMAModelusing16-QAM

modulationtechniqueinAWG

Nandmultipathfadingchannel


54/138

35

3.2.1.2 Transmitter Design

3.2.1.2.1 User Data Sequence Generator

The signal is produced by Bernoulli Data Generator. The Bernoulli Binary

Generator block generates random binary numbers using a Bernoulli distribution. The

Bernoulli distribution with parameter p produces zero with probability p and one with

probability 1-p. The Bernoulli distribution has mean value 1-p and variance p(1-p). The

Probability of a zero parameter specifies p, and can be any real number between zeroand one. Table below shows the parameters used in Bernoulli Binary Generator block.

Table 3.1: Parameters for Bernoulli Binary Generator Block

Parameter Value

Probability of Zero 0.5

Initial Seeds 12345

Sample Time Tsample

Frame-based Output unchecked

Interpret Vector Parameter as 1-D unchecked

Parameters for Bernoulli Binary Generator Block

Probability of Zero: The probability with which a zero output occurs. The value

of 0.5 means the random binary numbers generated having equal amount of 0

and 1.

Initial Seeds: The initial seed value for the random number generator. 12345 has

been chosen.

Sample Time: The period of each sample-based vector or each row of a frame-

based matrix. Here, Tsample is declared in the associated m file


55/138

36

(chk_QPSK_no_noise.m) and it has a value of 1/384000 seconds or an inverse of

bit rate of 384 kbps.

Attribute of Output Signal. In this block, we declare the signal to be sample-

based where the frame-based outputs box is unchecked.

3.2.1.2.2 Spreading Sequence Generator

The PN Sequence Generator block generates a sequence of pseudorandom binary

numbers. A pseudo-noise sequence can be used in a pseudorandom scrambler and

descrambler. It can also be used in a direct-sequence spread-spectrum system. The PN

Sequence Generator block uses a shift register to generate sequences. Table 3.2 below

shows the parameter that has been used in the simulation.

Table 3.2: Parameters used in PN Sequence Generator Block

Parameter Value

Generator Polynomial [1 0 0 0 0 1 1]

Initial States [1 0 0 0 0 1]Shift (or mask) 0

Sample time, Tc Tchip

Attribute of Output Signal Sample-based output

Parameters Specific to PN Sequence Generator

a) The Generator polynomial parameter has been specifying using this format:

x A vector that lists the coefficients of the polynomial in descending order

of powers. The first and last entries must be 1. Note that the length of this

vector is one more than the degree of the generator polynomial.


56/138

37

It is known that the degree of the generator polynomial is 6 so the length of the vector is

7. So, in this simulation, the value of Generator Polynomial specifies as [1 0 0 0 0 1 1]

represent the same polynomial, p (z) = z6 + z1 + 1 to fulfill the format above.

b) The initial states parameter is a vector specifying the initial values of the

registers. The initial states parameter must satisfy these criteria:

x All elements of the Initial states vector must be binary numbers.

x The length of the Initial states vector must equal the degree of the generator

polynomial.

x At least one element of the Initial states vector must be nonzero in order for theblock to generate a nonzero sequence. That is, the initial state of at least one of

the registers must be nonzero.

So, in this simulation, the value of initial states specifies as [1 0 0 0 0 1] to satisfy the

criteria above.

c) Sample time, Tc in this case is equal to chip period. In this simulation chip

period 260.4167 ns inverse of chip rate, 3.84 M chips per second (Mcps) is used and it is

declared in the chk_QPSK_no_noise.m file. In the sample time parameter check box of

the QPSK_no_noise.mdl file, it is declared as Tchip.

d) Attribute of Output Signal.

In Simulink, each matrix signal has a frame attribute that declares the signal to be

either frame-based or sample-based, but not both. (A one-dimensional array signal is

always sample-based, by definition.) Simulink indicates the frame attribute visually by

using a double connector line in the model window instead of a single connector line. In

general, Simulink interprets frame-based and sample-based signals as follows:

x A frame-based signal in the shape of an M-by-1 (column) matrix represents M

successive samples from a single time series.


57/138

38

x A frame-based signal in the shape of a 1-by-N (row) matrix represents a sample

of N independent channels, taken at a single instant in time.

x A sample-based matrix signal might represent a set of bits that collectively

represent an integer, or a set of symbols that collectively represent a code word,

or something else other than a fragment of a single time series.

So, in this block, we declare the signal to be sample-based with unchecked the frame-

based outputs box.

3.2.1.2.3 Spreader

XOR block has been used to operate like a spreader. Spreader causes the data

symbols to be spread to a higher bandwidth, by multiplying the random binary data

symbols, bit rate equal to Tb with a high bit rate code sequence (pseudo noise chip

sequence), chip rate equal to Tc.

Parameters Specific to Logical Operator Block

a) Operator. XOR has been selected.

b) Number of input ports = 2

c) Show additional parameters box is checked.

d) Require all inputs and output to have same data type box is checked.

e) Output data type mode. Logical has been selected. To avoid any data or signal

incompatibility, the following steps are taken. Go to the menu bar of

simulation/simulation parameters/Advanced tab. Select Boolean Logic Signals to off,

then the output data type will match the input data type, which may be Boolean or

double.


58/138

39

3.2.1.3 Modulation Techniques

3.2.1.3.1 QPSK Modulator

In this simulation, Quadrature Phase Shift Keying (QPSK) Passband modulator

has been used. The M-PSK Modulator Passband block modulates using the M-ary phase

shift keying method. The output is a passband representation of the modulated signal.

The M-ary number parameter, M, is the number of points in the signal constellation.

This block uses the baseband equivalent block, M-PSK Modulator Baseband, forinternal computations and converts the resulting baseband signal to a passband

representation. The following parameters in this block are the same as those of the

baseband equivalent block:

x M-ary number

x Input type

x Constellation ordering

The input must be sample-based. If the Input type parameter is Bit, then the input

must be a vector of length log2 (M). If the Input type parameter is Integer, then the input

must be a scalar.

This block uses a baseband representation of the modulated signal as an intermediate

result during internal computations.

Table 3.3 below shows the parameter that has been used in the simulation.


59/138

40

Table 3.3: Parameters used in QPSK Modulator Passband Block

Parameter Value

M-ary number 4

Input type Integer

Symbol period (s) 1 / 3840000 Hz

Baseband samples per symbol 1

Carrier frequency (in Hz) 15000000

Carrier initial phase (in rad) pi/4

Output sample time (s) 1 / 38000000 Hz

Parameters Specific to Passband Simulation

a) M-ary number is set up to 4 means it using four points in the signal

constellation. This setting also indicates the modulator to function as a QPSK modulator.

b) Input type parameter is set up to integer means the input must be a scalar.

c) The Symbol period parameter must equal the sample time of the input signal.

The sample time of the input signal is equal to Tc = 260.4167 ns. So, the symbol period

equal to 260.4167 ns or inverse of 1/3840000 Hz.

d) Baseband samples per symbol. The Baseband samples per symbol parameter

indicates how many baseband samples correspond to each integer or binary word in the

input, before the block converts them to a passband output. In this simulation, baseband

samples per symbol specify to one baseband sample per symbol.

e) Passband simulation uses a carrier signal. Carrier frequency (fc), 15,000,000

Hertz (Hz) has been used. The actual carrier frequency that should be used is 2 GHz to

fulfill the third generation requirements. Smaller frequency is used because of the

computer ability to simulate larger value of the carrier frequency. Simulation runs very


60/138

41

slow when value 2 GHz applied. The assumption made for simulation environment, the

differences between values of carrier frequency does not affect the system.

f) Carrier initial phase in radians specify the initial phase of the carrier signal. In

this simulation, the initial phase = (pi/4) or S/4 indicates the initial phase for QPSK

modulation scheme.

g) Output sample time. The Output sample time parameter determines the sample

time of the output signal. The timing-related parameters must satisfy these

relationships:x Symbol period > (Carrier frequency)-1

x Output sample time < [2*Carrier frequency + 2/(Symbol period)]-1

Furthermore, Carrier frequency is typically much larger than the highest frequency of the

unmodulated signal.

First Condition = Symbol period > (Carrier frequency)-1

Symbol period = 260.4167 ns

Carrier frequency (fc) = 15,000,000 Hz

[Carrier frequency (fc)]-1 = 66.667 ns

So, 260.4167 ns > 66.667 ns, satisfy first condition

When first condition satisfied, the value of output sample time should satisfy the second

condition.

Output sample time < [2*(15,000,000 Hz) + 2 / (260.4167 ns)]-1

Output sample time < [30 MHz + 7.68 MHz]-1

Output sample time < 26.54 ns

So, in this simulation, the output sample time has been determined to 26.32 ns or inverse

of1 / 38,000,000 Hz.


61/138

42

3.2.1.3.2 16-QAM Modulator

Same design procedures are implemented for 16-QAM modulator. Table below

shows the parameters used in rectangular 16-QAM modulator passband block.

Table 3.4: Parameters used in M-QAM modulation block

Parameter Value

M-ary number 16

Input type Integer

Normalized method Min. Distance between symbols

Minimum distance 2

Symbol period 1/100


Carrier frequency (Hz) 15e6

Carrier Initial Phase (rad) 0

Output Sample Time 1/38e6

Normalization method determines how the block scales the signal

constellation.

Minimum distance is the distance between two nearest constellation points.

Symbol period (s) is the symbol period, which must equal the chip sample

time of the PN sequence generator

Baseband sample per symbol is the number of baseband samples that

correspond to each integer or binary word in the input, before the block

converts them to a passband output.


62/138

43

3.2.1.4 Channel Design

The following table 3.5 shows the parameters used in AWGN block.

Table 3.5: Parameters used in AWGN block

Parameter Value

Initial seeds 1237

Mode Signal to Noise ratio (Es/No)

Es/No EbNo

Input Signal Power (Watt) 1

Symbol-period (s) Tchip

In the generic m file, the EbNo will produce a sequence of 2 EbNo intervals for

12 EbNo. The symbol period of AWGN block is Tchip that is equivalent to 1/38e6.

3.2.1.5 Receiver Design

3.2.1.5.1 QPSK Demodulator

The M-PSK Demodulator Passband block demodulates a signal that was

modulated using the M-ary phase shift keying method. The input is a passband

representation of the modulated signal. The M-ary number parameter, M, is the number

of points in the signal constellation. This block converts the input to an equivalent

baseband representation and then uses the baseband equivalent block, M-PSK

Demodulator Baseband, for internal computations. The following parameters in this

block are the same as those of the baseband equivalent block:

x M-ary number

x Output type

x Constellation ordering


63/138

44

The input must be a sample-based scalar signal. Similar parameters for QPSK

Demodulator will be used as QPSK Modulator except for parameter input sample time.

In this simulation, input sample time is equal to output sample time of QPSK Modulator.

Table3.6 below shows the parameter that has been used in the simulation.

Table 3.6: Parameters used in QPSK Demodulator Passband Block

Parameter Value

M-ary number 4

Input type IntegerSymbol period (s) 1 / 3840000 Hz


Carrier frequency (in Hz) 15000000

Carrier initial phase (in rad) pi/4

Input sample time (s) 1 / 38000000 Hz

Delays from QPSK Demodulation

Digital modulation and demodulation blocks sometimes incur delays between

their inputs and outputs, depending on their configuration and on properties of their

signals. Refer to the Release Notes Communication Blockset for Use with Simulink,

all passband demodulators except OQPSK will experience delays in amount of one

output period. So, QPSK passband demodulator causes delays of one ouput period occur

in this simulation block.

To calculate the bit error rate correctly, additional delay of 1 second to the

transmitted signal to synchronize it with the received signal. This is done directly in the

mask for the Error Rate Calculation block by setting the Receive delay to 1.


64/138

45

For the same reason, the PN chip sequence and received signals need to be

synchronizing before they enter the Despreader block. In this case, Integer Delay block

has been used, which delays a signal by the number of sample periods specified by the

Delay parameter. Set the Delay to 1. This is indicated by the exponent -1 on the block.

The delay synchronizes the PN chip sequence signal with the received signal so that the

Despreader block can recovered back the original data symbols correctly.

16-QAM Demodulator

As in QPSK demodulator, similar design procedure will be employed for 16-QAM. Table below shows the parameters used in rectangular 16-QAM Demodulator

block. The parameters are the same parameters used in initializing the parameters in 16-

QAM modulator block.

Table 3.7: Parameters used in 16-QAM Demodulator Passband Block

Parameter Value

M-ary number 16

Input type Integer

Normalized method Min. Distance between symbols

Minimum distance 2

Symbol period 1/100


Carrier frequency (Hz) 15e6

Carrier Initial Phase (rad) 0

Output Sample Time 1/38e6

3.2.1.6 Despreader

In order to recover the data symbols from the spreading signal, the process of

dispreading is applied. This is done by XOR the high bit rate noise-like signal with a


65/138

46

local spreading chip code that has the same sequence with the transmitting code. When

this right code is chosen with right synchronization, in this case delay is one output

period; the output from the XOR block will be exactly the same as the source signal.

Parameters for Despreader block are same like the Spreader block.

3.2.1.7 Error Rate Calculation

The Error Rate Calculation block compares input data from a transmitter with

input data from a receiver. It calculates the error rate as a running statistic, by dividing

the total number of unequal pairs of data elements by the total number of input dataelements from one source.


Table 3.8: Parameters used in Error Rate Calculation Block

Parameter Value

Receive delay 1

Computation delay 0

Computation mode Entire frame

Output data Port

Reset port box Unchecked

Stop simulation box Checked

Target number of errors 5000

Maximum number of symbols 5000

a) Receive delay set up to 1 due to delay causes by QPSK Demodulator. Refering

to the release notes Communication Blockset for Use with Simulink, the delay should

be put as 1 to ensure the transmitted signal synchronize with the received signal.

b) Computation mode is set to entire frame. Then the block compares the entire

transmitted frame with the entire received frame.


66/138

47

c) Output Data. This block produces a vector of length three, whose entries

correspond to:

x The error rate.

x The total number of errors, that is, comparisons between unequal elements.

x The total number of comparisons that the block made.

Output data parameter is set to Port, and then an output port appears. This output port

contains the running error statistics. Output port from this block connected to the

Display.

d) The simulation stops when the maximum number of symbols is reached at 5000

data symbols even the target number of errors not reached 5000 errors.

3.2.1.8 Display

The Display block shows the value of its input, the amount of data displayed and the

time steps at which the data is displayed are determined by block parameters:

x The display format can be control by selecting a Format choice: short, which

displays a 5-digit scaled value with fixed decimal point

x The Decimation parameter enables you to display data at every nth sample,

where n is the decimation factor. The default decimation, 1, displays data at

every time step.

x The Sample time parameter enables you to specify a sampling interval at which

to display points. This parameter is useful when you are using a variable-step

solver where the interval between time steps might not be the same. The default

value of -1 causes the block to ignore the sampling interval when determining the

points to display.



67/138

48

Table 3.9: Parameters used in Display Block

Parameter Value

Format Short

Decimation parameter 1

Floating display cox unchecked

Sample time -1

The display block shows the bit error rate, the number of errors and the total number of

bits that are transmitted.

3.2.1.9 Performance Analysis for Phase 1

In this simulation, a generic m file is used together with simulink to simulate the

BER vs Eb/No graphs (refer to Appendix, section 1.1). This m file declares the

parameters defined in the simulinks block diagram check box. For example, variable

Tsample declared in the m file is the sampling time of Bernoulli Binary Generator.

Tchip, on the other hand, is the sampling time of spreading sequence generator.

EbNoVec is signal to noise ratio and it is taken at 6 evenly points starting from 0. Then,

for loop is used to calculate the BER for every EbNoVec assigned to it. The value of

EbNo will be stored in the work space. Commandsim is used to simulate simulink mdl

file. Finally, commandsemilogis used to create the graph for BER vs Eb/No.

First, the simulation is done by running the concerned mdl file. Once the output

values are stored in the workspace, the associated m file is typed under the command

window and it is run. Finally, BER graph vs EbNo graphs are obtained once the

simulation is completed.


68/138

49

The generic m file used to generate BER vs Eb/No graph

M = 4;

Tsample = 1/384000; % Bernoulli Binary Sampling time

Tchip= 1/3840000; % Chip sampling time

BERVec = [];

EbNoVec = [0:2:12];

for n=1:length(EbNoVec);

EbNodB = EbNoVec(n);

sim('WCDMA_QPSK_baseband');

BERVec(n,:) = BER;

end;

semilogy(EbNoVec,BERVec(:,1),'+');

legend('Bit error rate');

xlabel('Eb/No (dB)'); ylabel('Error Probability');

title('Bit Error Probability');

In this phase, the system is simulated based on the following conditions





69/138

50

3.2.2 Simulation Phase 2: WCDMA system in AWGN and Multipath Rayleigh

Fading

In Phase 2, both transmitter and receiver part are built based on the system model

as shown in figure 3.2 and figure 3.3. In this model, multipath Raleigh fading channel

block is added in the system. The rest of system blocks and parameters are unchanged.

In this phase of simulation, the model is simulated in the baseband simulation

environment. The input of multipath Raleigh fading block requires complex signal

which can be obtained through baseband simulation only.

Moreover, in passband simulation, the simulation models the carrier frequency.

Since the carrier frequency is usually a high frequency signal, modeling passband

communication systems involves high computational loads. To alleviate this problem,

baseband simulation techniques are used.


70/138

51

Figure3.4:W-C

DMAModelwithmultipathRa

leighfadingchannelandAWGNchannelusingQPSKModulation


71/138

52

Figure

3.5:W-CDMAModelusing16-QAMinAWGNandmultipathRaleighfadingchannel


72/138

53

3.2.2.1 Channel

Table 3.10 below is used to initialize the parameters in the multipath Rayleigh fading

block.

Table 3.10: Parameters used in multipath Rayleigh fading channel

Parameter Value

Maximum Doppler shift (Hz) 55.56 / 83.33 / 111.111

Sample time (s) 1/3840000Delay vector (s) 2-ray [0 2e-6]

3-ray [0 2e-6 3e-6]

Gain vector (s) 2-ray [0 -3]

3-ray [0 -3 1]

Normalize gain vector to 0 dB overall gain box Checked

Initial seed 40

a) Maximum Doppler shift (Hz). Relative motion between the transmitter and

receiver causes Doppler shifts in the signal frequency. The Jakes PSD (power spectral

density) determines the spectrum of the Rayleigh process. This implementation is based

on the direct form simulator described in reference [1]. Some wireless applications, such

as standard GSM (Global System for Mobile Communication) systems, prefer to specify

Doppler shifts in terms of the speed of the mobile. If the mobile moves at speed v

making an angle T of with the direction of wave motion, then the Doppler shift is

Tcos

c

vfFd (3.20)


73/138

54

where f is the transmission carrier frequency and c is the speed of light. The Doppler

frequency is the maximum Doppler shift arising from motion of the mobile.

In this project, to determine maximum Doppler shift for third generation systems,

we assume that the transmission carrier frequency, f = 2GHz. We assume mobile moves

at three different speed, v = 60 km/hr, v = 90 km/hr and v = 120 km/hr. Different speeds

modeling the system in three different situations; in the middle of town, in the main road

and in the highway. Angle T set to 60 degree.

Based on the assumptions above, the maximum Doppler shift for every speed value is:

When v = 60 km/hr;

HzFd

ms

GHzhrkmFd o

56.55

60cos103

2/6018

u

u

(3.20(a))

When v = 90 km/hr;

HzFdms

GHzhrkmFd o

33.83

60cos

103

2/9018

u

u

(3.20(b))

When v = 120 km/hr;

HzFd

ms

GHzhrkmFd o

111.111

60cos103

2/12018

u

u

(3.20(c))

b) Sample time equal to chip code rate, Tc = 3840000 Hz.

c) Delay vector is a vector that specifies the propagation delay for each path. In this

project, we assume the delays for 2 paths are 0 second and 2e-6 second and the delays

for 3 paths are 0 second, 2e-6 second and 3e-6 second.

d) Gain vector is a vector that specifies the gain for each path. The gains for 2

paths are 0 dB and -3 dB. The gains for 3 paths are 0 dB, -3 dB and 1 dB.


74/138

55

e) Normalize gain vector to 0 dB overall gain box. Checking this box causes the

block to scale the Gain vector parameter so that the channel's effective gain (considering

all paths) is 0 decibels.

3.2.2.2 Performance Analysis for Phase 2

Same procedures are used to do the performance analysis for phase 2 as they are

done in phase 1. Performance analysis for this system is based on the following

conditions:





In summary, there are six procedures to be based on for simulation in phase 1 and

phase 2 as they are shown as follows:









fading channel for different number of user for QPSK modulation technique.


fading channel for different number of user for 16-QAM modulation technique.


75/138

56

3.3 Simulation Using M file

Another method is simulation using M-files. A script can be written in MATLAB

editor or another text editor to create a file containing the same statements that can be

typed at the MATLAB command line. The file is saved under a name that ends in .m.

The MATLAB language used in m file is a high-level matrix/array language with control

flow statements, functions, data structures, input/output, and object-oriented

programming features. It allows both simple and complicated programs to simulate all

real-time situations.

3.3.1 Generation of Spreading Code

In CDMA, the choice of code sequence is very important in respect to multiuser and

multipath interference encountered by the signal in the channel. To combat these

interferences, the code has to have the following properties:

1. Each code sequence generated from a set of code-generation functions must be

periodic with a constant length.


easy to distinguish from its shifted code.


easy to distinguish from other code sequences.

The first and second requirements are important with respect to the multipath

propagation effects that occur in mobile outdoor and indoor radio environments.


76/138

57

However, the third requirement is important with respect to the multiple access

capability of communication systems. Thus, to ensure a distinction level of codes for

requirements 1 and 2, an autocorrelation function and a cross-

Date post:	04-Apr-2018
Category:	Documents
Upload:	srikandiujung
View:	231 times
Download:	3 times

Muhammad Najib is Mail Med 2005 t Ttt

Documents