Atlanta University Center

DigitalCommons@Robert W. Woodruff Library, Atlanta

University Center
ETD Collection for AUC Robert W. Woodruff Library

7-1-2012

The performance of high-order quadrature

amplitude modulation schemes for broadband
wireless communication systems
Larry Riche
Clark Atlanta University

Follow this and additional works at: http://digitalcommons.auctr.edu/dissertations

Part of the Computer Engineering Commons

Recommended Citation
Riche, Larry, "The performance of high-order quadrature amplitude modulation schemes for broadband wireless communication
systems" (2012). ETD Collection for AUC Robert W. Woodruff Library. Paper 300.

This Thesis is brought to you for free and open access by DigitalCommons@Robert W. Woodruff Library, Atlanta University Center. It has been
accepted for inclusion in ETD Collection for AUC Robert W. Woodruff Library by an authorized administrator of DigitalCommons@Robert W.
Woodruff Library, Atlanta University Center. For more information, please contact [email protected].
 ABSTRACT

COMPUTER INFORMATION SCIENCE

RICHE II. LARRY DONNELL B.S. CLARK ALTANTA UNIVERSITY, 2009

THE PERFORMANCE OF HIGH-ORDER QAUDRATURE AMPLITUDE

MODULATION SCHEMES FOR BROADBAND

WIRELESS COMMUNICATIONS

SYSTEMS

Committee Chair: Khalil A. Shujaee. Ph. D.

Thesis dated July 2012

The limited amount frequency spectrum available to wireless comnmnication systems

makes it difficult to satisfy the rapidly growing demand for wireless service. Spectral

efficiency can be increased by using higher order modulation schemes. However this

come at the cost of increased probability of error. In this paper we investigate through

MATLAB simulation, the implementation of orders of Quadrature Amplitude Modulation

(QAM) more commonly used in wired networks. The BER performance of 64, 128, 256,

512, 1024, 2048, 4096, and 8192 QAM signals in the presence of Rayleigh and Rician

multipath channels with additive white Gaussian noise are simulated.

THE PERFORMANCE OF HIGH ORDER QAM MODULATION SCHEMES FOR

BROADBAND WIRELESS COMMUNICATIONS SYSTEMS

A THESIS

SUBMITTED TO THE FACULTY OF CLARK ATLANTA UNIVERSITY

IN PARTIAL FULFILLMENT OF THE REQUIRMENTS FOR

THE DEGREE OF MASTER OF SCIENCE

BY

LARRY RICHE

DEPARTMENT OF COMPUTER INFORMATION SCIENCE

ATLANTA, GEORGIA

JULY 2012
 © 2012

LARRY DONNELL RICHE II

All Rights Reserved

 ACKNOWLEDGMENTS

This thesis is dedicated in memory of my grandmother, Alice Kelly Ford. I thank

my parents who have provided me with immeasurable amount of guidance, support, and

unconditional love throughout my life. A true acknowledgment to them is a thesis in

itseIf one whose research is forever incomplete. Mother, there are no words to express

how much I love you. To me, you are the definition of great, and I could only dream to be

half as accomplished as you. Your strength amazes me every day that I live. Dad, I thank

you for the path you showed me, and for always pushing me to confront my fears. I once

mentioned to you my childhood goal to become “smarter” than you, and well, I think I’m

halfway there. I also thank my grandmother, sister, brother and my cousin Tiffany for

always being there when I need them. Much appreciation is given to my thesis advisor,

Dr. Khalil Shujaee for guidance and always challenging me to go further in academia and

life in general. I thank Dr. Peter Molnar for the countless number of reference letters, and

exposing me to many opportunities in computer science. I thank Dr. Roy George for

supporting me and opening doors of opportunity. Thanks to Ms. Shanna Hicks and Ms.

Mia Moore. Also, thanks to Dr. Xiao-Qian Wang and Dr. Ishard Kahn for the opportunity

to work on their research projects. Thanks to my graduate peers at Clark Atlanta

University: Olayinka Ogunro Ii, Kelvin Suggs, Chemo Kah, Duminda Samarakoon,

Chantel Nicolas, Rosi Gunasinghe, Kregg Quarles, Robert Bailey, Saeed Motevali, Mihiri

Shashikala, Brandon Larkin, and Darkeyah Reuven.

11
 TABLE OF CONTENTS

ACKNOWLEDGEMENTS . ii

TABLE OF CONTENTS iii

LIST OF FIGURES v

LiST OF TABLES vi

LIST OF ABBREVIATIONS vii

Chapter

1. INTRODUCTION 1

1.1 Motivation 2

1.2 Digital Modulation Techniques 3

1.2.1 Linear Modulation 5

1.2.2 Non-Linear Modulation 6

1.3 Why Quadrature Amplitude Modulation 7

1.4 Thesis Scope and Organization 8

2. QUADRATURE AMPLITUDE MODULATION 9

2.1 Background 9

2.2 Applications 12

2.3 Mathematical Model 13

2.4 Calculating the Probability of Error 13

3. IMPLEMENTATION OF HIGH ORDER QAM SCHEMES 15

3.1 Bandwidth Efficiency 15

4. SYSTEM MODEL 17

111
 4.1 Input .19

4.2 Modulation .20

4.3 Channel 22

4.4 Basic Structure 24

5. SIMULATION 25

5.1 Simulation Models 25

5.2 Results 29

6. DISCUSSION 35

6.1 Conclusion 35

6.2 Future Work 25

REFERENCES 37

iv
 LIST OF FIGURES

1.1 Evolution of Wireless Communication Standards 1990-2011 2

1.2 Generation of BPSK modulated signal 6

2.1 Examples of types I, II, and III QAM constellations 10

2.2 Figure 2.2 “Optimum” 32-level constellation according to Smith ii

4.1 General communications block diagram 18

4.2 Example of Gray coding fbr 16 QAM 19

4.3 Distribution of input tuple for 64 QAM Simulation Model 20

4.4 Even constellation of 256 QAM 21

4.5 Odd constellation for 128 QAM 22

4.6 Basic QAM modem schematic 24

5.1 Simulink simulation model of 64 QAM transmission over AWGN channel 26

5.2 Simulink simulation model of 64 QAM transmission over Rayleigh flat fading

channel with AWGN 27

5.3 Simulink simulation model of 64 QAM transmission over Rician flat fading

channel with AWGN 28

5.4 MATLAB Communications Toolbox BERTooI 30

5.5 Simulated BER performance of even bit constellations in the presence of Additive

White Gaussian Noise 31

V
5.6 Simulated BER performance of odd bit symbol constellations in the presence of

Additive White Gaussian Noise 31

5.7 Simulated BER performance of even bit constellations in the presence of

Rayleigh flat fading channel with AWGN 32

5.8 Simulated BER performance of odd bit constellations in the presence of Rayleigh

flat fading channel with AWGN 33

5.9 Simulated BER performance of even bit constellations in the presence of Rician

flat fading channel with AWGN 34

5.10 Simulated BER performance of odd bit constellations in the presence of Rician

flat fading channel with additive white Gaussian noise 34

vi
 LIST OF TABLES

1.1 Modulation Formats and Applications 4

1.2 Classification of Modulation Schemes 5

3.3 Quadrature Amplitude Modulation Bandwidth Efficiency 11

vii
 LIST OF ABBREVIATIONS

AM Amplitude Modulation

EbNQ Energy per Bit to Noise Power Spectral Density Ratio

FM Frequency Modulation

FSK Frequency Shift Keying

LTE Long Term Evolution

OFDM Orthogonal Frequency Division Multiplexing

OFDMA Orthogonal Frequency Division Multiple Access

PM Phase Modulation

PSK Phase Shift Keying

QAM Quadrature Amplitude Modulation

QPSK Quadrature Phase Shift Keying

WiMAX Wireless Interoperability for Microwave Access

VIII
 CHAPTER 1

Introduction

How we communicate plays an important role in our everyday lives. Since the

dawn of Society mankind has sought to create more efficient and effective ways to share

information. In 1897 Guglielmo Marconi invented the first radio frequency

communications system to establish continuous communication with ships across the

English Channel [1]. This was the beginnings of a new paradigm in communications in

which information could be shared over large distances at the speed of light. Since then

there has tremendous technological advances in wireless communications overcoming

greater barriers. In global positioning systems (GPS), terrestrial devices communicate

with satellites in orbit around the Earth to determine the location of the device within an

accuracy of three meters [2]. Wireless communications has experienced explosive growth

in the past 25 years, due to demand and advances in Very Large Integrated Circuit

(VLSI) and Discrete Signal Processing (DSP) technologies. It is useful to refer to mobile

cellular technology when describing the recent evolution in wireless communications,

since its market has experienced the most growth, and gives a good representation of the

capabilities of wireless technology. Figure 1.1 shows a depiction of the advancement of

wireless standards, beginning with the second generation of technology.

Cellular communications systems are perhaps the most common type of wireless

communications systems. The first generation mobile wireless communications systems

used analog modulation to carry voice over air between transmitter and receiver. In 1978

Bell Laboratories developed the Advanced Mobile Phone System (AMPS), it was the

first cellular system to be implemented in the United States. It used Frequency

Modulation (FM) over a 30 kHz channel, within the 824-894 MHz frequency band and

Frequency Division Multiple Access (FDMA) for user multiplexing.

The second generation of cellular technology began with the launch of the Global

System for Mobile (GSM) Communications [3]. The system was introduced to Europe in

1988.

2G

D 2.C

3C

3.5G

39G [UMB]

4G

Figure 1.1 Evolution of Wireless Communication Standards 1990-2011

1.1 Motivation

The demand of high data rates in wireless applications continues to grow at an

exponential rate. Additionally the number of connected mobile devices is also increasing

According to Cisco, global mobile data traffic is expected to double yearly though 2016

[4]. The limited availability of frequency spectrum hinders this growth. The limited
 2

availability of spectrum makes achieving high data rates in wireless communications

challenging objective. High data rates are essential for demanding applications such a

real-time video streaming. High throughput is realized through tradeoffs between

bandwidth, power, and system complexity. Modulation is the process of transforming an

incoming stream of information into a format that is suitable for transmission over a

medium. There are many different types of modulation techniques choose form.

Quadrature Amplitude Modulation (QAM) has been a popular choice for implementation.

1.2 Digital Modulation Techniques

In telecommunications modulation is the process of modifying a data source into

a form more suitable for transmission over a medium [5]. The properties of a signal can

alternated to represent the differentiations in the data being transmitted. A general sine

wave can be represented as a function of time, t as follows

s(t) = A cos(a.t + )

where A is the amplitude, a 2nJ is the frequency, and is the instantaneous phase.

The choice of digital modulation scheme is crucial for mobile communications

systems. It is a key factor in determining the performance of a communications system.

There are two basic forms modulation, analog and digital. In analog modulation an

incoming data source is modulated by a continuous signal, referred to as a carrier. In

digital modulation the source signal is modulated by a discrete carrier signal. Today

mobile wireless systems are using digital modulation techniques. Table 1 .1 shows a

variety of digital modulation techniques and some of their applications. Digital signals

tend to be more robust than analog signals in terms of signal integrity. Frequency, phase,

and amplitude are the three fundamental properties of a signal which can be modulated.
 4

Typical first generation wireless standards used simple analog modulation techniques,

such as amplitude, phase, and frequency modulation. Modem 4G mobile standards use

adaptive modulation schemes which adjust the modulation scheme depending on the

signal to noise ratio [6. 7]. The implementation of digital modulation techniques began

with the second generation of wireless communication standards. The shift toward digital

modulation provided more information capacity, higher security, and better quality

service as compared to analog formats. As seen in Table 1 .2 there are two major

categories of digital modulations. The first category uses a constant amplitude carrier to

carry the information in phase or frequency variations, such as frequency shift keying

(FSK) and phase shift keying (PSK). The second category conveys the information in

carrier amplitude variations, such as amplitude shift keying (ASK) and quadrature

amplitude modulation (QAM).

Table 1.1 Modulation Formats and Applications

Modulation Format Application

MSK, GMSK GSM, CDPD
BPSK Deep telemetry, cable modems
QPSK, DQPSK Satellite, CDMA, NADC, TETRA, PHS, PDC, LMDS, DVB-S,
cable (return path), cable modems, TFTS
OQPSK CDMA, satellite
FSK, GFSK DECT, paging, RAM mobile data, AMPS, CT2, ERMES, land
mobile, public safety
8, 16 VSB North American digital TV (ATV), broadcast, cable
8PSK Satellite, aircraft, telemetry pilots for monitoring broadband video
systems
16 QAM Microwave digital radio, modems, DVB-C, DVB-T
32 QAM Terrestrial microwave, DVB-T
64 QAM DVB-C. modems, broadband set top boxes, MMDS
256 QAM Modems, DVB-C (Europe), Digital Video (US)
Table 1.2 Classification of Modulation Schemes

Modulation
Linear (Modulate A. q)
5 Exponential (Modulate dçS / di)
Binary Multilevel (>1 bit/symbol) FSK, MSK, GMSK, etc.
ASK, BPSK M-PSK, QAM

1.2.1 Non-linear Modulation

Binary phase-shift keying (BPSK) is one of the simplest modulation schemes, in

which the phase of the carrier signal is modulated with two distinct states representing

data being transmitted:

‘0’, q5=7z
‘1’, ØrrO

This is shown in Figure 1.2, where the modulated signal is a product of equivalent

baseband signal and carrier frequency. BPSK has been implemented in mobile standards

since the second generation of wireless technology.

 6

1 0 1
1101110
Equivalent baseband signal

‘VV\N\AAJV\I\N\A/VVVV\MA
Carrier

VV\fWVVv\JVWW
Modulated Sia1

Figure 1.2 Generation of BPSK modulated signal

Quadrature phase-shift keying (QPSK) is an extension of BPSK in which four different

phases or states can be transmitted opposed to two.

M-ary phase-shift keying (MPSK) is a multi-level modulation scheme. It consist

of QAM 2G standards such as IS-95 and GSM used BPSK. Gaussain Miniumal Shift

Keying (GMSK), and Different DQPSK. Binary Phase-shift Keying (BPSK), Quadrature

Phase-shift Keying (QPSK), and Quadrature Amplitude Modulation (QAM) modulation

techniques are used in 4th Generation Wireless System, LTE and WiMAX [3] [4]. For

BSPK and QPSK only the phase of a transmitted signal is allowed to vary.

1.2.2 Non-linear Modulation

The simplest FSK is binary FSK (BFSK). BFSK uses a pair of discrete

frequencies to transmit binary (Os and Is) information. With this scheme, the “1” is called
 7

the mark frequency and the “0’ is called the space frequency. The time domain of an FSK

modulated carrier is illustrated in the figures to the right.

Minimum shift keying (MSK) Minimum frequency-shift keying or minimum-shift

keying MSK is a particular spectrally efficient form of coherent FSK. In MSK the

difference between the higher and lower frequency is identical to half the bit rate.

Consequently, the waveforms used to represent a 0 and a I bit differs by exactly half a

carrier period.

Gaussian minimum shift keying (GMSK) modulation is based on MSK, which is

itself a form of continuous-phase frequency-shift keying. One of the problems with

standard forms of PSK is that sidebands extend out from the carrier. To overcome this,

MSK and its derivative GMSK can be used.

1.3 Why Quadrature Amplitude Modulation

Much research has done on the comparison of QAM with other modulation

techniques. In most cases the QAM tends to have better performance overall. For

instance, multiple papers show that QAM has better noise resistance, and is more

adaptive for channel change, and more efficient in bandwidth utilization [8-1 0]. While

PSK modulation schemes often demonstrate better BER performance than their QAM

counterparts, it may be possible to mitigate this issue with more complex hardware [9]. It

is concluded that BER performance of QPSK is better than that of I 6-QAM at the

expense of large spectral width. Alternatively, 16-QAM can carry more traffic than

QPSK at the expense of higher BER[9].

 8

1.4 Thesis Scope and Organization

The objective of this thesis is to investigate the performance of higher order

modulation schemes in the presence of noise and fading channel environments. The

thesis is organized as follows.

Chapter 2 provides a brief background of advancements in research on

Quadrature Amplitude Modulation and touches on some of its applications, then proceeds

to go into a detailed description of how it is implemented including mathematical

descriptions of a QAM signal and calculating the probability that in error occurs during

transmission. Chapter 3 examines the potential gains in the implementation of higher

order QAM. We calculate the system capacities and bandwidth efficiencies of the various

QAM schemes. In Chapter 4. we detail the design of a wireless communications system

that implements QAM. Chapter 5 covers the components of the simulation, the

mythology discussed in chapter 4 are implemented in to a MATLAB SIMULINK model.

In Chapter 6, the results of the simulations are discussed and suggestions for future

research are given.

 CHAPTER 2

Quadrature Amplitude Modulation

2.1 Background

Digital phase modulation techniques were popular in the 1950s. prior to the

development of QAM[ 11]. It was sought as an alternative to digital amplitude

modulation[12]. Later, a method unifying both schemes was explored. QAM is

essentially the combination of digital amplitude and phase modulation. It was first

suggested in 1960 by C.R. Cahn, who published a paper describing such a system[l3]. He

did this by His research suggested that amplitude and phase modulation (AM-PM)

systems would result in an increase in data throughput compared to an equivalent PSK

system of 16 or more states.

Shortly after its publication, Calm’s work was expanded by Hancock and Lucky

[14]. They determined that errors were induced by the introduction of noise into a signal.

This caused by a shift in the phasor at the receiver to represent a different constellation

point than the intended one. This type of error has a direct relation to the distance

between constellations points. Their solution was to place more points on the outer ring

of the constellation, being that points on the inner ring were closer in proximity to each

other and thus more likely to experience errors. They coined Cahn’s constellation as Type

I and there’s as Type II.

In 1962. Campopiano and Glazer introduced the Type III constellation [15].

Points in this new type of constellation were arranged to form a square. Although the

9
 I0

acronym QAM had not yet been suggested, they described the system as “the amplitude

modulation and demodulation of two carriers that have the same frequency but are in

quadrature with each other” — the first time the combination of amplitude and phase

modulation had been refer to as amplitude modulation on quadrature carriers [1.1]. One

issue they found with their type III constellation is that non-coherent detection was not

possible, therefore it had to be used in a phase coherent mode. They concluded that while

their Type III constellation offered modest performance over a Type II system, it had the

advantage of simplicity in implementation compared to Types I and II. Examples of Type

I, II, and III constellations are shown in Figure 2.1.

Type IQAM Constellation Type UQAM Constellation

• • .

• • .

• I •

• I I
Q
Type ifi QAM Constellation

Figure 2.1 Examples of types I. II and III QAM constellations from [16]
 II

Although much research had gone into the design of more optimal constellations,

by 1975 most interest had been centered on Type III or square QAM constellations [17-

19]. This was due to the minimal gains realized by the optimal constellations verses the

complexity of implementation. and advancements if coherent phase detection methods

[20, 21]. Up until this time the constellation shape of square QAM had only been

considered for systems with an even number of bits per symbol. In 1975 J.G Smith

suggested an idea design for odd bit per symbol constellations [22]. His work proposed

that odd bit symbol constellations be symmetric and that such systems were of equal

implementation complexity as even bit symbol constellations. An example of his idea

constellation shape for odd bit symbols is shown in Figure 2.2, where there are five bits

per symbol.

•• . .
• • • • . .
• • • • • •
I
••. • • •
• • • • • •
• • • •
Q

Figure 2.2 Optimum” 32-level constellation according to Smith [22] from [16]

The first major research investigating the implementation of QAM for mobile

radio came in 1987. Sundberg, Wong. and Steele published two papers proposing QAM

for the transmission of voice over Rayleigh fading channels [23, 24]. In this paper they

used gray code mapping as a method to assign symbols to constellation points. They
 12

found that certain assortments of bits constituting a symbol had different rates of error

than others. By 1990 research on QAM had become a very popular. There are many

papers that explore its implementation for mobile radio applications.

Currently 1024 and 4094 QAM are implemented in wired communication

standards. The ability of these dense forms of QAM to achieve exponential increases in

spectral efficiency makes them a prime candidate as a method to increase capacity. High

order constellations of QAM also have the advantage of increasing data throughput.

Advances in VLSI and Digital Signal Processing DSP techniques have enabled the

implementation of relatively complex forms of modulation such as QAM. It can be

expected that this trend will continue into the future.

2.2 Applications

Quadrature amplitude modulation is implemented in a variety of modern wireless

communications standards. QAM is in many radio communications and data delivery

applications. However some specific variants of QAM are used in some specific

applications and standards.

For domestic broadcast applications for example, 64 QAM and 256 QAM are

often used in digital cable television and cable modem applications. In the UK, 16 QAM

and 64 QAM are currently used for digital terrestrial television using DVB Digital
-

Video Broadcasting. In the US, 64 QAM and 256 QAM are the mandated modulation

schemes for digital cable as standardized by the SCTE in the standard ANSI/SCTE 07

2000.

ln addition to this, variants of QAM are also used for many wireless and cellular

technology applications.
 13

2.3 Mathematical Model

Digitally modulated signals can be represented in a multitude of forms using

variety of techniques derived from the three basic forms of frequency, amplitude, and

phase keying. In general a modulated signal can be represented by

bE bE
s, (1) =
— ““
1 cos(2irft)
a
+ q — “
i, sin(2,rft) 0 t T; I = 1,2... (2.1)

where Emin is the energy of the signal with the lowest amplitude, and a and b
1 are a pair

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

ary QAM does not have constant energy per symbol, nor constant distance between

possible symbol states, which is why some values of s(t) will be detected with higher

probability than others.

2.4 Calculating the Probability of Error

Here we discuss how to obtain the theoretical bit error rate (BER) and symbol

error rate (SER) for a given QAM constellation using a general equation. The BER and

SER are important factors in determining the usefulness of modulation schemes. Analysis

of theoretical and experimental results of the error rates.

For square constellations of QAM, where k = log

2 M is even, the SER, F and

BER, F can be described as follows [25-27]:

P
4_1Q 3 kE 7
M
4
( _1Q I3 I (72)
J
0
M-1N ...Jf 0
M-1N J -.

From [28]:
 14

log (I—2 )-I I kI

2
[l
I

Hj
i2’ /61og
M
2 E
J
7
J1og
1 L2 _[+jQ2i+1) 2(M—1) N
0

(2.3)

k—I
M is odd, M =IxJ, 1=22 ,and
For cross shaped QAM where k=log
2

k+l
2
J=2

4LJ—21—2J ( I (IJ) Eb ___(1+IJ_I_J)Q2

2
61og (IJ) E,,
2
61og
M (J2+J22N

(2.4)

From [281:

IogI log.J
(k)+ (l)J (2.5)
= 1og,(IJ)

where,

(1-2 4)J [ I I k-l

2
(k) 21 —I +—1jQ(2i1q 6Iog,(IJ) Eb
(2.6)
=1 L I 2 + -2 N
0

and

21
2
H
- t
P(k)=- j(_1) ‘
-L’’ + jQ[(2i+1) 2
6log,(IJ) Eb
Jj
0
_2N
(2.7)
 CHAPTER 3

Implementation of High Order QAM Schemes

3.1 Bandwidth Efficiency

= 2
k
In communications the simple expression M is used to relate the symbols to

bits. Where M is the number of unique symbols and k is the number of bits transmitter

per symbol. Using Shannon’s channel capacity formula we can calculate the potential

capacity of a system using each proposed density of QAM. Shannon’s formula is

described as follows

C = 2
Blog [i+i_J= Blog
1+
7 [bit/si (3.1)

where C is the channel capacity (bits per seconds), [29-31]. 20MHz is a commonly

allocated bandwidth for channels in 4G cellular communications systems such as LTE-A

and WiMAX [32, 33]. Table 4.1 shows the increase in capacity with respect to each

QAM scheme.

The Shannon bound can be expressed as bandwidth efficiency, i = C / B, by

using (3.1)

(1+S/N) [bit/s/Hz]
7
17=log (3.2)

15
 16

The maximum achievable bandwidth efficiency is calculated using Equation 3.2.

The results are shown in table 3.1.

Table 3.1 Quadrature Amplitude Modulation Bandwidth Efficiency

QAM Order Bits per Symbol, k Bandwidth Efficiency(per 20MHz

channel)
64 QAM 6 120 Mb/s
128 QAM 7 140 Mb/s
256QAM 8 160 Mb/s
512 QAM 9 180 Mb/s
1024 QAM 10 200 Mb/s
2048 QAM I1 220 Mb/s
4096 QAM 12 240 Mb/s
8192 QAM 13 260 Mb/s
 CHAPTER 4

System Model

In this chapter we discuss the components of a wireless communications system,

the design and methodology of the system parameter that will be incorporated into the

simulations discussed in chapter 5. A typical communications system is shown in Figure

4.1. The components are as follows:

The data source is the origin of the message to be communicated. The message

could come in a variety of forms such as audio, video, or other data. The data source is

transformed into an electrical signal by a transducer. The resulting electrical waveform is

referred to as the baseband signal. The baseband signal is modified by the transmitter,

which prepares the signal for efficient transmission. In digital communication systems the

baseband signal is then coded into binary format using coding techniques. The modulator

is the focus of this research. Baseband signals are often not suitable for transmission over

a given channel. The modulator modifies the baseband signal to facilitate transmission.

The channel is the medium over which communication takes place, such as coaxial cable,

wire, optical fiber, or over-the-air radio link. For this study the channel is a radio link,

which we discuss further in section 4. The receiver translates the signal into its original

form. At the receiver the signal is demodulated, decoded, and processed by transducer

which converts the electrical signal back into its original form,

17
 18

Data source

Received dat.

Figure 4.1 General communications block diagram.Input

The simulation model incorporates the Gray coding the map the symbols to points

on the constellation. The bits sequence to signal mapping could be arbitrary provided that

the mapping is one-to-one. However, a method called Gray coding is commonly used in

signal assignment in QAM [34]. Gray coding assigns n-tuples with only one-bit

difference to two adjacent signals in the constellation. When an M-ary symbol error

occurs, it is more likely that the signal is detected as the adjacent signal on the

constellation, thus only one of the n input bits is in error. An example of gray coding is

shown in figure 4.2 for the case of 16 QAM.

 19

Scatter plot
5

ooo moo iio ioo

• . •

.c•

ooI1 o11 ilOi 1oi

1
0 .

- 0011 m liii 1o1

oo•o oio iiio ioo

-5
-5 -3 -1 1
In-Phase

Figure 4.2 Example of Gray coding for 16 QAM

4.1 Input

The input is modeled using with a Bernoulli distribution to represent a quasi-

uniform distribution of, which can be seen in Figure 2.2. The Bernoulli distribution can

be described as follows

Pr(X=l)=1—Pr(XrrO)=1—q=p (4.1)

where probability p has a 50% chance of occurrence, meaning all possible outcomes have

the same probability [35]. This method of input distribution is chosen because it is known

that some values of input produce higher rates of error than others. Having a quasi-

uniform distribution of all inputs ensures that the simulations will represent the overall

performance of the system.

 20

5
x 10

0
0 10 20 30 40 50 60 70

Figure 4.2 Distribution of input tuple for 64 QAM Simulation Model

4.2 Modulation

Here we give an overview of the types of QAM scheme that will be implemented

in the simulation. As discussed in Chapter 2, there are two constellation types of square

QAM schemes, those with even bit symbols and those with odd bit symbols. In Figure

4.3 the constellation of an even bit symbol system for the case of 256 QAM is shown.

Figure 4.4 represents an odd bit symbol system for the case of 128 QAM. In both cases

the constellation can be extended to represent a higher order system of the same type by

expanding the number of points to M while maintaining the integrity of the constellations
 21

k
2
shape. Remembering that M =
is used to relate the number of bits per symbol to the

number of unique symbols.

Scatter Plot

20

15 •--“ .---.---.--.---.--4---.---.---.--.---.---.--.---.---,
+ • •:. • + • + + 4 + * + 4 + +
+ + •:. • + + + • + • * + . +
C I -I C
10
* S 4 * * * + + • • * + +
t
t + •:• • t + + + + t + + + + +
5 +---.---+--.--÷-4--+---.--.--.--4---.---.--.---.---+
* S + * 55 * * * * * *

t * •: • t * •:. + + * • + + t
+ • * + • •:. + + * * + + +
I
-o I

• * * + + * + + + + +
t I

a ---*--s---*--+--4--t--.-;-+--9---I.
7
I.---*---+--*---t--

• * •:. • * •:. * * * •:+ •

-10 F
+ + •* • • •: * + • •: +

-15 +---*---++---+----+--+-f-*--+--+---*---*+---+---

-20

-20 -16 -10 -6 0 6 10 16 20

In-phase (AU)

Figure 4.3 Even constellation of 256 QAM

 Scatter Plot

15

• • • * • • •
10
• • • • • •

• + 4 • • • • • • •

5 .-----.----,----4----.----.-- --.----.---4-----,---+----.
• • •• • • *

• : • • • .1. • 4 • • •
. 0
• • •
-U

• 4 •• • 4 ••

-5 .--:-------.----t---*----÷-t--._---.---i---*---.--:--*
* I • * 4 • .1. • * * •

• • •:• • • I
-10
* • .1* * 4 •

-15 --

I I I I I I
-15 -10 -5 0 5 10 15
In-phase (AU)

Figure 4.4 Odd Constellation of 128 QAM

4.3 Channel

There are several conditions present in real world environments that can distort

the original signal at the receiver. These conditions include noise added by the

environment and the multipath effects of reflection, diffraction, and scattering. It is

important to include these environmental factors into the model, to have a representation

as close to reality as possible.

Additive white Gaussian noise (AWGN) is a channel model in which the only

impairment to communication is a linear addition of wideband or white noise with a

constant spectral density (expressed as watts per hertz of bandwidth) and a Gaussian
 23

distribution of amplitude. The model does not account for fading. frequency selectivity,

interference, nonlinearity or dispersion. However, it produces simple and tractable

mathematical models which are useful for gaining insight into the underlying behavior of

a system before these other phenomena are considered. The channel capacity for the

AWGN channel is given by Equation 3.1.

The Rayleigh probability distribution function is commonly used to incorporate

the multipath effects of a typical system. The Rayleigh PDF is useful in fading channel

environment of wireless communication systems, where multiple signals are scattered,

diffracted, and reflected at the receiver.

for (r 0)
(4.2)
0 for (r<0)

•‘here u is the rms value of the received voltage signal before the envelope, o is the

time-average power of the signal before the envelope.

Rician

for (A0,r0)
Pr( )
2
a (4.3)
0 for (r<0)

Where A is the peak amplitude of the dominate signal and 1

(E) is the modified Bessel
0

function of the first kind and zero-order.

The Rician PDF is another The Rician PDF if often expressed in terms of K. It is

given by 2
/(2o
K=A
) .
 24

__I(r2÷)/2a2
A K+I’
( K(K+l)”
‘o 2r for (AO,rO)
Pr(ñ A ) (4.4)
0 for(r<0)

where A is the peak amplitude of the dominate signal, and K is the defined as the ratio

between the deterministic signal power and the variance of the multipath.

4.4 Basic Structure

Figure 4.3 shows the basic structure of an QAM transmitter and receiver.

natogLAe

sri’.rce

tf rQcery

OL.J tput

Figure 4.5 Basic QAM modem schematic from [16]

 CHAPTER 5

Simulation

5.1 Simulation Models

Here we discuss the method used to create the simulation models, and discuss key

parameters. The simulation models were created using Simulink, an extension of the

MATLAB 2012 software suite [27]. The simulation model block functions are as

follows:

• The Bernoulli Binary block is used to generate a random sequence of numbers

with a Bernoulli probability distribution.

• The Rectangular QAM Modulator Baseband block, to the right of the Random

Integer Generator block, modulates the signal using baseband QAM.

• The AWGN Channel block models a noisy channel by adding white Gaussian

noise to the modulated signal.

• The Rayleigh block introduces noise in the angle of its complex input signal

• The Rician block introduces noise in the angle of its complex input signal.

• The Rectangular QAM Demodulator Baseband block, demodulates the signal.

• The Error Rate Calculation block counts symbols that differ between the

received signal and the transmitted signal

25
 26

Figures 5.1-5.3 show the SIMULfNK simulation models for 64 QAM transmitted

over AWGN, Rayleigh fading + AWGN, and Rician fading + AWGN channels

respectively.

BnuIli Binary
GEnaratcr

Redan9ular
M cd LA!
Baeband1

Bit Efrc Rate

r4urr Bit EITc5

Mum Bitxmitted

RetanuIac o.ArA
Drrdu! atc.f
Baehan d

Figure 5.1 Simulink simulation model of 64 QAM transmission over AWGN channel
 27

64-QAM Transmission over Rayleigh I

Flat Fadin Channel No Diversity I -

Perfect Channel State Infcrrnatiorr (CSl) is assumed at the receiver.

Bemnulli Binary
Generator

Basebandi
Rayleigh
Ch an ii ci
t6Dx1 35942e-O ii Bit Error Rate
txmn> Tx
Rate
215ee--a4I NLrmnBitErrors

I Mum Bits xmitted

To Wcrspace

BER

Rectangular CAM
Demodulator
Baebarr

Figure 5.2 Simulink simulation model of 64 QAM transmission over Rayleigh flat
fading channel with AWGN
 28

64-QAM Transmission over Rician

Flat F Channel No Diversity -

Perfe Channel State Inftrmatjnn jCSI i auned at the recehier.

Bemnnulli Binary
Generatcr

Bauebana 1
lultipath Rician
Channel

Bit Esrnr Rate

liurn Bit Errcs

Nurn Bits xrnitted

RecianBulaI QAM
Derncciulaft
Eaahand

Figure 5.3 Simulink simulation model of 64 QAM transmission over Rician flat
fading channel with AWGN

A channel equalizer is implemented to estimate the effects of a given medium

over which a received signal is transmitted. In the case of the model the receiver has

perfect knowledge of distortions induced by the channel. This allows us to fundamental

behavior of QAM over a given channel. The equalizer in the simulated models divides

incoming signal by the gain of the channel to retrieve the original message. The equalizer

can be seen in Figure 5.1 accepting it inputs from the channel output and channel gain.
 29

5.2 Results

The BER performance plot where generated using the BERToo1 provided in

MATLABs Communications Toolbox. It implements equations 2.2-2.7 to calculate the

results of the simulation. Figure 5.4 shows the BERToo1 parameters for the case of 64

QAM over Rayleigh fading channel. We simulate the model shown in 5.2 for an Eb / N
0

range of 0 to 60 with samples plotted every three steps. The simulations are stopped when

1000 errors are reached or 1e9 bits are processed. For higher orders of QAM the

simulation limits are increased to increase the accuracy of the fitted curve.
 30

i Bit Error RateAnalysisTool Jl3i1

File Edit Window Help

Confidence Level Fit Plot BER Data Set EbfN (dB) BER # of Bits

SemianaIytic riont

E!Nrange: o:3:60J

Sin,ulation MATLAB file or Simulink model: op\MyThesis\Sirnulations\Rayleigh\Rayleigh64QAM.mdl Browse..

BER variable name: BER

Simulation limits:

Number of errors: 1000 —

or
Number cf bits: 1

L_i L_J

Figure 5.4 MATLAB Communications Toolbox BERTooI

Here we analyze results of MATLABs l3ERTool on the models created with

SIMULINK. The results of the BER simulation for even and odd symbol, QAM

constellations are shown in Figures 5.5 and 5.6 respectively. Analyzing the plots at a

BER of iO, it can be seen that for each increase in the number of bit per symbol an

additional —2-3dB of EbNO is required to achieve that same performance.

 31

100

1 o2

+
cc
uJ
aD

* 64QAMM=6
1o6 +
o
256 QAtvl, rv18
1024QAMM=10
.\..
s> 4096 QAM M=12

10.8
0 5 10 15 20 25 30 35
Eb/NO (dB)

Figure 5.5 Simulated BER performance of even bit symbol constellations in the
presence of Additive White Gaussian Noise

100

.z-e--._,r—
t -

- -

N. : + 8L.
1 o.2
+
:Q
+ N, :

cc
ui I s
0] .

‘H

* 128 QAM M7
+ 512 QAMI M=9 .

0 2048 QAM M1 1
I)
8192 QAM M=13

I I

0 5 10 15 20 25 30 35
Eb/NO (dE)

Figure 5.6 Simulated BER performance of odd bit symbol constellations in the
presence of Additive White Gaussian Noise
 32

The BER plots of the simulation of QAM over Rayleigh fading channel with

AWGN is shown in Figures 5.7 and 5.8. Analyzing the plots it can be seen that to achieve

a BER of 1 o’ requires an additional 1 dB of Lb N

0 per increase in the number of bits per

symbol to maintain performance. We also the difference in dB is relatively uniform at all

values of BER.

0 I

•1
* 64 QAM, M6
-1
+ 2E6QAM,M=8 1
10 E!HH!.H 0 1024 CAM, M=10
:::::L 0 4096 CAM, M=12 :.

10 1

io

2 7
Eb/No (dB)
Figure 5.7 Simulated BER performance of even bit constellations in the presence of
Rayleigh flat fading channel with additive white Gaussian noise
 33

10°
* 120 QAM, M7
612 QAM,M9
Q’ 6
.
.
0 2040 QAM, M11
:!. S192QAM,M13

11

0 10 20 30 40 60 60 70
Eb/No (dE l)
Figure 5.8 simulated BER performance of odd bit constellations in the presence of
Rayleigh flat fading channel with additive white Gaussian noise

The BER plots of the simulation of QAM over Rician fading channel with a K factor of 5

and AWGN is shown in Figures 5.9 and s.io. Analyzing the plots it can be seen that to

achieve a BER of 1 o requires an additional l dB of EhNO per increase in the number of

bits per symbol to maintain performance.

 34

10 1:1::::::: :1:::::: 11111 1:::::::

* 64 CAM, M=6
256 CAM M=8
10 ‘
C 1024Cv1M=10
:::::::.::.. $096 CAM, M=12
2 .

10 H iHHLHç

io 1 ii n.:::
:::.:::::::::::::::::::::::.::.:::..c ::..:.c::::::.:::::::.:::::::::.
iu
Cl]
..
10
4
N
.5 .
: \.
10 Hi!i HHWHH11H1HHHH111H1HH w
4
t ‘IHHH
+
I . .
+

0 10 20 30 40 50 60 70
Eb/NO (dO)
Figure 5.9 Simulated BER performance of even bit constellations in the presence of
Rician flat fading chaimel with additive white Gaussian noise

10 . :.: -

WEE.HHHHE * 128CAM,M=7
1
+ 512 CAM, M=9
10 Hk;ftNUHmHHHHHHW 0 1 Mu
2048 CAM
+
+ 8132 CAM M=13
-2
10

.4..
io_ .,‘?i—.l.

ci:
+
: .
io

:- :-
10

- .-

10

uo
0 10 20 30 40 50 60 70
Eb/NO (dO)

Figure 5.10 Simulated BER performance of odd bit constellations in the presence of
Rician flat fading channel with additive white Gaussian noise
 CHAPTER 6

Discussion

6.1 Conclusion

In this paper, we have simulated rectangular QAM schemes of the order 64, 128.

256, 512, 1024, 2048, 4096, and 8192 over AWGN, Rayleigh. and Rician channels. We

also performed theoretical calculations to obtain maximum throughput. Our results show

that the throughput of the system is increased when transmitting more bits per symbol.

This increase in throughput comes at the expense of higher BER. which can be mitigated

by increasing the energy per bit to noise power spectral density ratio of the system. It is

clear that there has to be a compromise between higher throughput, higher BER, and the

energy output. Based on our results, it is important to further investigate the

implementation of high order QAM in wireless broadband systems. This research will be

continued to reinforce the results in this paper.

6.2 Future Work

Our results show that further investigation is needed to determine the practicality

implementing high order QAM in wireless broadband systems. In the future we will add

empirical models to the simulation. This will allow us to examine the performance of

high order QAM in the presence of large-scale path loss. Additionally, we will to

incorporate Orthogonal Frequency Division Multiplexing (OFDM), which is being highly

adopted by modem broadband communications systems. into the system

35
 36

model to get a better idea of the perfonnance. Also, it may be interesting to investigate

the effects more optimal QAM constellations would have in our model.
 37

REFERENCES

L Rappaport, T.S., Wireless Communications Principles and Pratice. 2nd ed.

Prentice Hall Communications Engineering and Emerging Technologies Series,

ed. T.S. Rappaport2002, Upper Saddle River, Nj 07458: Prentice Hall.

2. Global Positioning System Standard Positioning Service Performance Standard,

Defense, Editor 2008.

3. Haykin, S. and M. Moher, Modern Wireless Communications2005, Upper Saddle

River, NJ: Pearson Prentice Hall.

4. Cisco Systems mc, Cisco Visual A/etworking Index: Global Mobile Data Traffic

Forecast Update, 2011-2016, 2012: USA.

5. Lathi, B.P. and Z. Ding, Modern Digital and Analog Communication Systems. 4th

ed. Oxford Series in Electrical and Computer Engineering2009: Oxford

University Press.

6. Nuaymi, L., WiMAX Technology for Broadband Wireless Access2007, West

Sussex, England: John Wiley & Sons Limited.

7. Ho, S.W. Adaptive Modulation (QPSK, QAM). 2004.

8. Tan, M. and W. Chen, Performance Comparison and Analysis ofPSK QAM, in

International Conference on Wireless Communications, Networking and Mobile

Computing20 11: Wuhan, China.

9. Huang, K.C. and Z. Wang, Millimeter Wave Communication Systems20l 1: John

Wiley & Sons.

 38

10. Baohong, L., W. Yanyan, and L. Tingjie, Performance Comparison ofHigh Rate

Ultra- Wideband OFDM System between Two Different Modulation Techniques,

in International Conference on Wireless Communications, Networking and

Mobile Computing. 2006: Wuhan, China.

11. Webb, W.T. and L. Hanzo, Modern Quadrature Amplitude Modulationl994,

London, England: Pentech Press Limited.

12. Cahn, C.R., Performance ofdigital phase modulation communication systems.

IRE Transactions on Communication Systems, 1959. 7(1): p. 3-6.

13. Cahn, C.R., Combined digital phase and amplitude modulation communication

system. IRE Transactions on Communication Systems 1960. 8(3): p. 150-155.

14. Hancock, J.C. and R.W. Lucky, PerfOrmance of combined amplitude and phase

modulated communications system. IRE Transactions on Communication

Systems, 1960. 8(4): p. 232-237.

15. Campopiano, C.N. and B.G. Glazer, A coherent digital amplitude and phase

modulation system. IRE Transactions on Communication Systems, 1962. 10(1): p.

90-95.

16. Hanzo, L.L., et al., Quadrature Amplitude Modulation: From Basics to Adaptive

Trellis-Coded, Turbo-Equalised and Space-Time Coded OFDM, CDM4 and MC

CDMA Systems. 2nd ed2004.

17. Lucky, R.W. and J.C. Hancock, On the optimum performance ofN-ary systems

having two degrees offreedom. IRE Transactions on Communication Systems,

1962. 10(2): p. 185-192.

 39

18. Foschini, G., R. Gitlin. and S. Weinstein, Optimization oftwo-dimensional signal

constellations in the presence ofgaussian noise. IEEE Transactions on

Communications, 1974. 22(1): p. 28-3 8.

19. Thomas, C., M. Weidner, and S. Durrani, Digital amplitude-phase keying with M

ary alphabets. IEEE Transactions on Communications, 1974. 22(2): p. 168-180.

20. Simon, M. and J. Smith, Carrier synchronization and detection of QASK signal

sets. IEEE Transactions on Communications, 1974. 22(2): p. 98-106.

21. Simon, M. and J. Smith, Ojfset quadrature communications with decision

feedback carrier srnchronization. IEEE Transactions on Communications, 1974.

22(10): p. 1576-1584.

22. Smith, J.G., Odd-bit quadrature amplitude-shfl keying. IEEE Transactions on

Communications, 1975. 23(3): p. 385-389.

23. Steele, R., C.-E.W. Sundberg, and W.C. Wong, Transmission of log-PCM via

QAM over Gaussian and Rayleigh Jading channels. TEE Proceedings-F

Communications, Radar, anf Signal Proceedings, 1987. 134(6): p. 539-556.

24. Sundberg, C.-E.W., W.C. Wong, and R. Steele, Logarithmic PCM weighted QAM

transmission over Gaussian and Rayleigh fading channels. TEE Proceedings-F

Communications. Radar, anf Signal Proceedings, 1987. 134(6): p. 557-570.

25. Simon. M.K. and M.-S. Alouini, Digital Communication over Fading Channels.

2nd ed. Wiley Series in Telecommunications and Signal Processing2004, New

York. NY: Wiley-IEEE Press.

 40

26. Proakis, J.G. and M. Salehi, Digital Communications, ed. 5th2007, New York,

NY: McGraw-Hill.

27. M4TLAB, 2012, The Math Works Inc.: Natick, Massachusetts.

28. Cho, K., On the general BER expression ofone- and two-dimensional amplitude

modulalions. IEEE Transactions on Communications, 2002. 50(7): p. 1074-1080.

29. Shannon, C.E., A matematical theory of communication. Part 1. Bell System

Technical Journal, 1948. 27: p. 379-423.

30. Shannon, C.E., A matematical theory ofcommunication. Part 2. Bell System

Technical Journal, 1948. 27: p. 623-656.

31. Shannon, C.E., Communication in the presence of noise, in Proceedings, Institute

ofRadio Engineersl949. p. 10-2 1.

32. IEEE, Air Interface for Broadband Wireless Access Systems, in Physical layer

(PHY)2 009.

33. 3GPP, Universal Mobile Telecommunications System (UMTS), in Requirements

for Evolved UTRA (E-UTRA) and Evolved UTRAN (E-UTRAN)2010.

34. Gray, F., Pulse Code Communication, U.S.P. Office, Editor 1947: United States.

35. Freund, J.E., Mathematical Statistics with Applications, 2004, Irwin Miller:

Upperaddle River, New Jersey.