PERFORMANCE OF WORLDWIDE
INTEROPERABILITY FOR MICROWAVE
ACCESS (WIMAX) IN NON LINE OF SIGHT AND
MULTIPATH ENVIRONMENT
KHAIRUL BARIAH BINTI ABAS AZMI
UNIVERSITI TEKNOLOGI MALAYSIA
iii
To all my loving family members,
Especially to my beloved HUSBAND & my dearly children
Siti Sabariah
Siti Aishah
Muhammad Saifuddin
Muhammad Jazman
Thank you for all your support and understanding
iv
ACKNOWLEDGEMENT
In the name of Allah, the Most Beneficent and Most Merciful
First and foremost, I would like to extend my highest gratitude and thanks to
my supervisor, Prof. Dr. Tharek B. Abd. Rahman for his generous support,
comments, advice and guidance throughout the duration of my project. Without his
continuous support and interest, this thesis would not have been the same as
presented here.
The thanks also go to all my friends for their constant kind help and moral
support despite the hectic semester that we had to undergo. Special thanks to Kak
Anis, Ainor, Muza, Bambang and others who have provided assistance at various
occasions…, thanks for being such a wonderful companion.
Last but not least, my deepest appreciation to my dearest husband, children
and my lovely parents for all their assistance, love and care….your fully support and
encouragement is gratefully appreciated.
v
ABSTRACT
The requirement of BWA to operate in the non line of sight and multipath
environment has lead to the amendment of IEEE 802.16 standard. The amended
standard which is 802.16a covers fixed BWA in the spectrum of 2 to 11 GHz has
become a stepping stoned to the enhancement of the wireless broadband technology.
A discrete channel model based on the Stanford University Interim (SUI) model was
used in the standard. This project evaluates the effect of empirical models on the
802.16a physical layer performance at 3.5 GHz. The empirical propagation model
used in the project are COST 231[8] and ECC 33 model[9]. The comparison of the
system performance of the OFDM between SUI, COST 231 and ECC 33 model were
evaluated in terms of bit error rate (BER).
The simulation results of these
propagation model on 802.16a physical layer specification were used to validate their
applicability in urban, suburban and rural environment.
.
vi
ABSTRAK
Keperluan di dalam capaian jalur lebar tanpa wayar di dalam untuk
beroperasi di dalam perambatan tanpa pandangan penglihatan dan berbagai laluan
telah mendorong kepada perubahan di dalam piawai “IEEE 802.16”. Piawai yang
telah diubah adalah “IEEE 802.16a” yang mana ianya meliputi capaian jalur lebar
tanpa wayar tetap, ia beroperasi di dalam specktrum frekuensi 2 x 109 hingga 11 x
109. Piawai baru ini merupakan satu lonjakan paradigma kepada teknologi jalur
lebar tanpa wayar. Saluran perambatan yang digunakan di dalam piawai jalur lebar
tanpa wayar yang baru ini ialah “Standford University Interim (SUI)”. Projek ini
menilai tindakbalas saluran perambatan “empirical” pada system digital piawai jalur
lebar tanpa wayar yang baru ini. Saluran perambatan “empirical” pada piawai jalur
lebar tanpa wayar yang digunakan adalah perambatan “COST 231” dan “ECC 33”.
Perbandingan tindakbalas system “OFDM” terhadap perambatan di antara “SUI”,
“COST 231” dan “ECC 33” akan dinilai dalam bentuk kadar kesilapan bit yang
diterima. Keputusan simulasi daripada ketiga-tiga jenis perambatan akan digunakan
untuk mengesahkan kegunaan piawai yang baru ini pada keadaan di dalam bandar,
luar bandar dan kampung.
vii
TABLE OF CONTENTS
CHAPTER
1
TITLE
PAGE
TITLE
i
CERTIFICATION
ii
DEDICATION
iii
ACKNOWLEDGEMENTS
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF TABLES
xi
LIST OF FIGURES
xii
LIST OF ABBREVIATIONS
xv
LIST OF SYMBOLS
xvi
INTRODUCTION
1
1.1
Introduction
1
1.1.1 Wimax Spectrum allocation
2
1.2
Objectives
2
1.3
Scope of Work
3
1.4
Outline of Thesis
4
viii
2
3
WIMAX TECHNOLOGY & FIXED WIMAX
5
2.1
Introduction
5
2.2
NLOS Technology Solution
6
2.2.1 OFDM
8
2.2.2 Sub Channelization
9
2.2.3 Antennas for Fixed Wireless Applications
11
2.2.4 Transmit and Receive Diversity
11
2.2.5 Adaptive Modulation
12
2.2.6 Error Correction Techniques
15
2.2.7 Power Control
15
2.3
The 802.16a standard
16
2.4
Why 3.5 GHz?
23
2.5
Summary
23
PROPAGATION PATH LOSS MODEL &
24
EXISTING CHANNEL MODEL
3.1
3.2
Introduction
24
3.1.1 Theoretical Models or stochastic models
25
3.1.2 Empirical Models
26
3.1.3
26
Physical Models or deterministic models
Propagation model used in the project
28
3.2.1 Overview
28
3.2.2 COST 231
29
3.2.3 Electronic Communication
3.3
Committee – Report 33 (ECC 33)
31
Stanford University Interim (SUI) model
33
3.3.1 Introduction
33
3.3.2 SUI parameter
35
ix
3.3
Verification of performance for SUI
38
channel in the existing 802.16a
3.3.1 Introduction
38
3.3.2 BER performance of SUI
3.4
4
5
channel ( for 3 taps)
39
3.3.2.1 Introduction
39
3.3.2.2 BER Performance
39
Summary
SIMULATION SPECIFICATION
44
4.1
Introduction
44
4.2
Design specification
45
4.3
Simulation results representation
47
4.4
Parameter set up in the channel model
61
4.5
Summary
62
SYSTEM PERFORMANCE
63
5.1
Introduction
64
5.2
System performance results in term of
bit-error-rate (BER)
5.3
5.4
6
43
61
Analysis of results for all environments in
all channel model
82
Summary
88
CONCLUSION AND SUGGESTION
89
6.1
Conclusion
89
6.2
Suggestion for future work
91
6.3
Wimax in Malaysia
92
x
REFERENCES
93 - 95
xi
LIST OF TABLES
TABLE NO.
TITLE
PAGE
2.1
802.16a PHY Features
17
2.2
OFDM symbol parameters
18
3.1
Parameters for various terrain type
34
3.3
SUI model parameter1
35
3.4
SUI model parameter2
35
4.1
Parameter set up in the channel model
61
xii
LIST OF FIGURES
FIGURE NO.
TITLE
PAGE
2.1
LOS Propagation
6
2.2
NLOS Propagation
8
2.3
Single Carrier and OFDM
10
2.4
Single carrier and OFDM received signal
11
2.5
The effect of sub-channelization
12
2.6
Relative cell radii for adaptive modulation
14
2.7
Simulator front panel of 802.16a physical layer
11
2.8
Simulator Block Diagram of 802.16a physical layer
12
3.1
The generic structure of SUI channel models
36
3.2
BER curves for AWGN
39
3.3
BER curves for QPSK under various channel condition
40
3.4
Comparison of channel estimation method on QPSK
modulation in SUI-3
41
xiii
3.5
Performance improvement for equalization and coding
for SUI-3
42
4.1
Front panel of the amended simulator of 802.16a
48
4.2
Block diagram of the amended simulator of 802.16a
49
4.3
Front panel of the OFDM in 802.16a simulator
50
4.4
Block diagram of the OFDM in 802.16a simulator
51
4.5
Front panel of the channel model in 802.16a simulator
52
4.6
Block diagram of the channel model in 802.16a simulator
53
4.7
Front panel of the Rician Fading in 802.16a simulator
54
4.8
Block diagram of the Rician Fading in 802.16a simulator
55
4.9
Front panel of the one tap processing in 802.16a simulator
56
4.10
Block diagram of the one tap processing in 802.16a simulator
57
4.11
Front panel of the AWGN in 802.16a simulator
58
4.12
Block diagram of the AWGN in 802.16a simulator
59
4.13
Front panel of the channel model in 802.16a simulator
60
5.1
SUI (CPE antenna height = 10 m, URBAN)
64
5.2
SUI (CPE antenna height = 10 m, SUBURBAN)
65
5.3
SUI (CPE antenna height = 10 m, RURAL)
66
xiv
5.4
SUI (CPE antenna height = 10m , ALL ENVIRONMENT)
5.5
SUI (CPE antenna height = 6 m, no Ch. Estimation, URBAN) 69
5.6
SUI (CPE antenna height = 6 m, URBAN)
70
5.7
SUI (CPE antenna height = 6 m, SUBURBAN)
72
5.8
SUI (CPE antenna height = 6 m, ALL ENVIRONMENT)
73
5.9
COST 231 (CPE antenna height = 10 m, URBAN)
75
5.10
COST 231
(CPE antenna height = 10 m, ALL ENVIRONMENT)
5.11
68
76
COST 231
(CPE antenna height = 6 m, ALL ENVIRONMENT)
77
5.12
ECC 33 (CPE antenna height = 10 m, ALL ENVIRONMENT) 79
5.13
ECC 33 (CPE antenna height = 6 m, ALL ENVIRONMENT)
5.14
SUI
(CPE antenna height = 6 m, QPSK, ALL ENVIRONMENT)
80
85
xv
LIST OF ABBREVIATIONS
BWA
-
Broadband Wireless Access
FWA
-
Fixed Wireless Access
WIMAX-
-
Worldwide Interoperability for Microwave Access
OFDM -
-
Orthogonal Frequency Division Multiplexing
BER
-
Bit error rate
SUI
-
Stanford University Interim
ECC 33
-
Electronic Communication Committee Report 33
TDD
-
Time Division Multiplexing
FDD
-
Frequency Division Multiplexing
LAN
-
Local Area Network
LOS
-
Line of Sight
NLOS
-
Non Line of Sight
CPE
-
Customer Premise Equipment
xvi
LIST OF SYMBOLS
f
-
frequency
d
-
distance
hb
-
Access Panel antenna height above ground level in
metres
nCOST
-
Path loss exponent for COST 231
λ
-
wavelength
γ
-
path-loss exponent of SUI
CHAPTER 1
INTRODUCTION
1.1
Introduction
WiMax stands for Worldwide Interoperability for Microwave Access.
WiMax refers to broadband wireless networks that are based on the IEEE 802.16
standard, which ensures compatibility and interoperability between broadband
wireless access equipment. WiMax, which will have a range of up to 31 miles, is
primary aimed at making broadband network access widely available without the
expenses of stringing wires (as in cable-access broadband) or the distance limitations
of Digital Subscriber Line. The rapid demand of Wimax is increasing because of the
services provided which includes the broadband internet access, landline telephone
bypass, cable/satellite TV bypass and mobile data plus cell phone services.
2
1.1.1 Wimax spectrum allocation
The IEEE Wireless Metropolitan Area Networks (Wireless MAN) Standard
802.16a is a new standard for BWA. Announced on January 30, 2003, the extension
of the 802.16 standard covers fixed broadband wireless access in licensed and
unlicensed spectrum from 2 to 11 GHz. It is also known as 802.16-2004 standards.
802.16-2004 WiMAX only supports fixed access, and products are already available.
It provides service from a base station to a subscriber station, also known as
customer premise equipment (CPE). Some goals for Wimax include a radius of
service coverage of 6 miles from a Wimax base station for point-to-multipoint nonline-of-sight service. This service should deliver approximately 40 Mbps for fixed
and portable access applications. While for point-to-point line-of-sight provides 30
miles of coverage with 72 Mbps.[4].
1.2
Objectives
The objectives of this project is to model and simulate the IEEE 802.16a
(fixed access) OFDM physical layer at 3.5 GHz using LabView 8.0. The project will
focus on the channel model of the standard. The channel models used are COST 231,
ECC 33 and Stanford University Interim (SUI) model. The system performance of
the OFDM through these channel models were evaluated under urban, suburban and
rural environment.
3
1.3
Scope of Work
The scope of work of the project is to study the OFDM system through fixed
broadband wireless access by evaluating the effects of modified path loss parameter
in the 802.16a physical layer. This project will simulate the 256-point transform of
OFDM physical layer at frequency allocation of 3.5 GHZ. The project modeled,
simulated, and evaluated the performance of OFDM in different path loss algorithm
in terms of bit-error-rate (BER). By this performance study, it is hoped that it could
be a reference for network developer to deploy a fixed WiMax access in a different
path loss model in Malaysia.
4
1.4
Outline of the Thesis
The thesis comprises of six chapters and the overview of all the chapters is as
below:
Chapter 1:
This chapter provides the introduction, objective and scope of work
involved in accomplishing the project.
Chapter 2:
This chapter presents the literature reviews on Wimax technology and
description of 802.16a fixed Wimax simulator.
Chapter 3:
This chapter comprises the literature review on propagation path loss
used and the verification of the system performance in 802.16a
simulator.
Chapter 4:
This chapter describes the methodology used in setting up the path
loss parameter in the OFDM.
Chapter 5:
The simulation and results obtained are discussed in this chapter.
Chapter 6:
Conclusion of the project and suggestions for future work are
presented in this final chapter.
CHAPTER 2
WIMAX TECHNOLOGY
&
FIXED WIMAX
2.2
Introduction
The IEEE 802.16a standard was thought and designed to operate in the lower
frequency bands, 2-11 GHz. By doing this, the standard includes the ability to
support Non-Line-Of-Sight (NLOS), something that would be impossible with the
previous version of the standard due to high frequencies involved (10 – 66 GHz) and
the mandatory need of line-of-sight.[10]. OFDM is the key technology for the
broadband wireless networks and applications. OFDM is an emerging technology
allowing digital signals to be transmitted simultaneously on multiple RF carrier
waves, allowing non-line-of-sight operation. It is also resistance to multipath effects
that obstruct the deployment of wireless broadband system.
6
2.2
NLOS Technology Solution
The radio channel of a wireless communication system is often described as
being either LOS or NLOS. In a LOS link, a signal travels over a direct and
unobstructed path from the transmitter to the receiver. A LOS link requires that most
of the first Fresnel zone is free of any obstruction. In a NLOS link, a signal reaches
the receiver through reflections, scattering, and diffractions. The signals arriving at
the receiver consists of components from the direct path, multiple reflected paths,
scattered energy, and diffracted propagation paths. These signals have different
delay spreads, attenuation, polarizations, and stability relative to the direct path. The
multi path phenomena can also cause the polarization of the signal to be changed.
Figure 2.1
LOS Propagation
There are several advantages that make NLOS deployments desirable. For
instance, strict planning requirements and antenna height restrictions often do not
allow the antenna to be positioned for LOS. For large-scale contiguous cellular
7
deployments, where frequency re-use is critical, lowering the antenna is
advantageous to reduce the co channel interference between adjacent cell sites. This
often forces the base stations to operate in NLOS conditions. LOS systems cannot
reduce antenna heights because doing so would impact the required direct view path
from the CPE to the Base Station.
NLOS technology also reduces installation expenses by making under-theeaves CPE installation a reality and easing the difficulty of locating adequate CPE
mounting locations. The technology also reduces the need for pre installation site
surveys and improves the accuracy of NLOS planning tools.
The NLOS technology and the enhanced features in WiMAX make it
possible to use indoor customer premise equipment (CPE). This has two main
challenges; firstly overcoming the building penetration losses and secondly, covering
reasonable distances with the lower transmit powers and antenna gains that are
usually associated with indoor CPEs. WiMAX makes this possible, and the NLOS
coverage can be further improved by leveraging some of WiMAX’s optional
capabilities. This is elaborated more in the following sections.
8
Figure 2.2
NLOS Propagation
WiMAX technology, solves or mitigates the problems resulting from NLOS
conditions by using:
• OFDM technology.
• Sub-Channelization.
• Directional antennas.
• Transmit and receive diversity.
• Adaptive modulation.
• Error correction techniques.
• Power control.
9
2.2.1 OFDM
Orthogonal frequency division multiplexing (OFDM) also sometimes called
discrete multitone modulation (DMT), is a modulation technique for transmission
based upon the idea of frequency-division multiplexing (FDM), where each
frequency channel carries a separate stream of data. In OFDM the frequencies are
chosen so that the modulated data streams are orthogonal to each other, which
greatly simplifies the design of both the transmitter and the receiver, and also allows
high spectral efficiency. Although the principles and some of the benefits have been
known since 1960s, OFDM is made popular today by the lower cost and availability
of digital signal processing components.
One key principle of OFDM is it suffers less from intersymbol interference
caused by multipath, even though in low rate modulation schemes. It is more
effective to transmit a number of low-rate streams in parallel than a single high-rate
stream. OFDM achieves this by dividing the available frequency spectrum into
several sub-bands, and then transmitting a low-rate data stream over each sub-band
using a standard modulation scheme (PSK, QAM, etc.). This means that the effects
of the channel are roughly constant (flat) over a given sub-band, making equalization
far simpler at the receiver.
Standard techniques such as channel coding, power allocation and adaptive
modulation may be applied to each sub-band or across all sub-bands. Multiple
access is also possible, using either time, frequency or coding separation of the users.
10
OFDM technology provides operators with an efficient means to overcome
the challenges of NLOS propagation. The WiMax OFDM waveform offers the
advantage of being able to operate with the larger delay spread of the NLOS
environment. By virtue of the OFDM symbol time and use of cyclic prefix, the
OFDM waveform eliminates the inter-symbol interference (ISI) problems and the
complexities of adaptive equalization. Because the OFDM waveform is composed
of multiple narrowband orthogonal carriers, selective fading is localized to a subset
of carriers that are relatively easy to equalize, as more robust to multi-path time
dispersion [1]. Therefore it can be concluded that an OFDM enabled receiver can be
consider as a received input signal all sub-carrier signals, whether they are direct or
not, as soon as the time difference between them is not too long.
Figure 2.3
Single Carrier and OFDM
11
Figure 2.4
Single carrier and OFDM received signal
12
2.2.3 Sub Channelization
Sub Channelization in the uplink is an option within WiMAX. Without sub
channelization, regulatory restrictions and the need for cost effective CPEs, typically
cause the link budget to be asymmetrical, this causes the system range to be up link
limited. Sub channeling enables the link budget to be balanced such that the system
gains are similar for both the up and down links. Sub channeling concentrates the
transmit power into fewer OFDM carriers; this is what increases the system gain that
can either be used to extend the reach of the system, overcome the building
penetration losses, and or reduce the power consumption of the CPE. The use of sub
channeling is further expanded in orthogonal frequency division multiple access
(OFDMA) to enable a more flexible use of resources that can support nomadic or
mobile operation.
Figure 2.5
The effect of sub-channelization
13
2.2.3 Antennas for Fixed Wireless Applications
Directional antennas increase the fade margin by adding more gain. This
increases the link availability as shown by K-factor comparisons between directional
and omni-directional antennas.
Delay spread is further reduced by directional
antennas at both the Base Station and CPE [9]. The antenna pattern suppresses any
multi-path signals that arrive in the sidelobes and backlobes. The effectiveness of
these methods has been proven and demonstrated in successful deployments, in
which the service operates under significant NLOS fading.
Adaptive antenna
systems (AAS) are an optional part of the 802.16 standard.
These AAS have
beamforming properties that can steer their focus to a particular direction or
directions. This means that while transmitting, the signal can be limited to the
required direction of the receiver, like a spotlight. Conversely when receiving, the
AAS can be made to focus only in the direction from where the desired signal is
coming from. They also have the property of suppressing co-channel interference
from other locations. AASs are considered to be future developments that could
eventually improve the spectrum re-use and capacity of a WiMAX network.
2.2.4 Transmit and Receive Diversity
Diversity schemes are used to take advantage of multi-path and reflections
signals that occur in NLOS conditions. Diversity is an optional feature in WiMAX.
The diversity algorithms offered by WiMAX in both the transmitter and receiver
increase the system availability. The WiMAX transmit diversity option uses space
time coding to provide transmit source independence; this reduces the fade margin
requirement and combats interference. For receive diversity, various combining
techniques are exist to improve the availability of the system.
For instance,
maximum ratio combining (MRC) takes advantage of two separate receive chains to
help overcome fading and reduce path loss. Diversity has proven to be an effective
tool for coping with the challenges of NLOS propagation.
14
2.2.5 Adaptive Modulation
Adaptive modulation allows the WiMAX system to adjust the signal
modulation scheme depending on the signal to noise ratio (SNR) condition of the
radio link. When the radio link is high in quality, the highest modulation scheme is
used, giving the system more capacity. During a signal fade, the WiMAX system can
shift to a lower modulation scheme to maintain the connection quality and link
stability. This feature allows the system to overcome time-selective fading. The key
feature of adaptive modulation is that it increases the range that a higher modulation
scheme can be used over, since the system can flex to the actual fading conditions, as
opposed to having a fixed scheme that is budgeted for the worst case conditions.
Figure 2.6
Relative cell radii for adaptive modulation
15
2.2.6 Error Correction Techniques
Error correction techniques have been incorporated into WiMAX to reduce
the system signal to noise ratio requirements. Strong Reed Solomon FEC,
convolutional encoding, and interleaving algorithms are used to detect and correct
errors to improve throughput. These robust error correction techniques help to
recover errored frames that may have been lost due to frequency selective fading or
burst errors. Automatic repeat request (ARQ) is used to correct errors that cannot be
corrected by the FEC, by having the errored information resent. This significantly
improves the bit error rate (BER) performance for a similar threshold level.
2.2.7 Power Control
Power control algorithms are used to improve the overall performance of the
system, it is implemented by the base station sending power control information to
each of the CPEs to regulate the transmit power level so that the level received at the
base station is at a predetermined level. In a dynamical changing fading environment
this pre-determined performance level means that the CPE only transmits enough
power to meet this requirement. The converse would be that the CPE transmit level is
based on worst-case conditions. The power control reduces the overall power
consumption of the CPE and the potential interference with other collocated base
stations. For LOS the transmit power of the CPE is approximately proportional to it’s
distance from the base station, for NLOS it is also heavily dependant on the
clearance and obstructions.
16
2.3
The 802.16a standard
The IEEE Wireless Metropolitan Area Networks (WirelessMAN) Standard
802.16a is a new standard for BWA.[11]. The key difference between the initial
802.16 standard and the 802.16a is the modulation scheme or the transmission
technique used. This modulation implies important changes into the physical layer
and also in the final application market. 802.16 is applicable only for LOS system
due to the modulation used by the standard. However, the 802.16a uses OFDM,
which is very robust for multipath environments and it is adaptable in NLOS
schemes.
17
Table 2.1 below shows the features of the 802.16a standard and the advantages of the
standard for BWA network.
Table 2.1 : 802.16a PHY Features[12]
Feature
256 point FFT OFDM waveform
Benefit
Built in support for addressing
mutipath in outdoors LOS and NLOS
environments.
Adaptive Modulation and variable
Ensures a robust RF link while
error correction encoding per RF burst. maximizing the number of bits/second
for each subscriber unit.
TDD and FDD duplexing support
Address varying worldwide regulations
where one or both mat be allowed.
Flexible Channel sizes ( e.g 3.5 MHz,
Provides the flexibility necessary to
5 MHz , 10 MHz etc)
operate in many different frequency
bands with varying channel
requirements around the world.
Designed to support smart antenna
Smart antennas are fast becoming more
systems
affordable, and as these costs come
down their ability to suppress
interference and increase system gain
will become important to BWA
deployments.
18
The IEEE 802.16a standard specifies three air interfaces: single-carrier
modulation, 256-point transform OFDM, and 2048-point transform OFDMA. This
project simulates the 256-point transform OFDM physical layer, since this air
interface is mandatory for operation in license-exempt bands [8]. The physical layer
has the following parameters:
Table 2.2 : OFDM symbol parameters [8]
The objective of the existing project is to model and simulate the IEEE
802.16a OFDM physical layer using LabVIEW 7.2. The simulations will compare
the performance of different receiver models, while noting their complexity. The
developed simulation will also provide a framework for developing new receiver
models for 802.16a systems.
19
The screenshot shown below is the 802.16a OFDM system without coding in an
additive white Gaussian noise channel.[8]
Figure 2.7
Simulator front panel of 802.16a physical layer[8] using Labview of
7.0
The simulator runs in two modes where it toggles between interactive and
preset modes of the simulator. Interactive mode allows run time modification of the
system parameters while observing the outputs. Preset mode runs the simulator
under the given parameters in incremental levels of signal to noise ratio in order to
plot the BER versus Eb/No curves. These curves are accumulated for each run for
easy comparison of performance as long as the clear button is darkened.
20
Figure 2.8
Simulator Block Diagram of 802.16a physical layer[8] using
Labview 7.0
The block diagram is the programming part, which control and simulate the
output. The simulation is done in Labview 7.2 where it requires an installation of
Mathlab on the same PC to run properly.
21
Each block are modeled as follows:[31]
Channel:
A discrete channel model based on the Stanford University Interim (SUI) [7]
model will be used. This model have a set of six typical channels which were
selected for the three terrain types that are typical of the continental US [7]. These
wireless channels are characterized by path loss, multipath delay spread, fading
characteristics, Doppler spread, co-channel and adjacent channel interference, and
antenna gain reduction factor. The user will also be given the option of modifying
the channel characteristics while the simulation is running.
OFDM Modulator and Demodulator:
The OFDM modulator and demodulator blocks would be developed using
standard IFFT and FFT blocks provided in LabVIEW. Perfect clock, symbol, and
frame synchronization is assumed, and the effects of synchronization error will not
be discussed in this project. OFDM Modulation Schemes will be applied to the
modulator.
Channel Estimator:
The channel estimator will obtain a depiction of the channel state to combat
the effects of the channel using an equalizer. Two estimation algorithms, the Least
Squares (LS) and Linear Minimum Mean Squared Error (LMMSE)[16], will be
analyzed in terms of their performance and complexity.
Channel Equalizer:
The channel equalizer will use the output from the channel estimator to
rearrange the effects of the channel and improve the performance of the system. A
frequency domain equalizer will be thoroughly investigated in terms of its
performance and complexity. Also, the performance of the system with and without
the equalizer will be evaluated.
22
Error Control Coding:
Error control coding is essential for OFDM systems since it compensates for
the bit errors that are predictable in times of deep fade in the channel. The Reed
Solomon encoder and decoder, convolutional encoder and Viterbi decoder, and the
randomizer/de-randomizer and interleaver/de-interleaver will be implemented, and
its effect on the overall bit error rate (BER) performance will be analyzed.
Performance Analysis Blocks:
To determine the qualitative and quantitative performance of the system and
give a good intuitive understanding of the effects of certain parameters on the system,
LabVIEW blocks that display performance curves were developed for the simulation.
These include bit error rate testers, spectrum analyzers, and constellation plotters.
23
2.4
Why 3.5 GHz?
The frequency was chosen because it can provide widespread high
performance in BWA.[24] which covered non-cabled broadband. Its channel
bandwidth is 7 MHz and it is the most likely band used for FWA system mainly in
Europe. This frequency has been tested by Europe Communication Community for
FWA system with spectrum allocation of 14 MHz that support TDD or FDD. 3.5
GHz is in the Super High Frequency (SHF) where it is the frequency allocated for
microwave devices, wireless LAN and modern radars.[24] It is often considered not
to be part of radio spectrum and form their own microwave spectrum.
2.5
Summary
In this chapter, an introduction of 802.16a followed by its advantages and
disadvantages are introduced. The Wimax technology were elaborated, thus this
project will focus on OFDM technology. Next, the propagation path loss used in the
project and the verification of 802.16a system performance are discussed. Finally,
the detailed explanation of simulation set up will be explained.
CHAPTER 3
PROPAGATION PATH LOSS MODEL
&
EXISTING CHANNEL MODEL
3.1
Introduction
Propagation and channel models are fundamental tools for designing any
fixed broadband wireless communication system. It basically predicts what will
happen to the transmitted signal while it transit to the receiver. These models are
divided into three basic classifications: theoretical, empirical, and physical.
25
3.1.1 Theoretical Models or stochastic models
These models are based on some theoretical assumptions about the
propagation environment.
Theoretical models are not suitable for planning
communications systems to serve a particular area because there is no way to relate
the parameters of these models to physical parameters of any particular propagation
environment. They do not directly use information about any specific environment,
thus it can be useful for analytical studies of the behavior of communication systems
under a wide variety of channel response circumstances. An example of a theoretical
model is the “tapped delay line” model in which densely spaced delays and
multiplying constants and tap-to-tap correlation coefficients are determined on the
basis of measurements or some theoretical interpretation of how the propagation
environment affects the signal. On the other hand, it models the environment as a
series of random variables. These models are at the least accurate but require the
least information about the environment and as much less processing power to
generate predictions.
26
3.1.2 Empirical Models
Empirical models are based on observations or measurements. Measurements
are typically done in the field to measure path loss, delay spread, or other channel
characteristics.
Empirical models are widely used in mobile radio and cellular
system engineering. Many cellular operators have ongoing measurements or drivetest programs that collect measurements of signal level, call quality, and network
performance which are then used to refine empirical propagation models used in the
system-planning tool.
An example of an empirical model is the “Cost-231 Hata “model which was
devised as an extension to the “Hata-Okumura” model. The Hata-Okumura model is
developed for the 500 to 1500 MHz frequency range using measurements done by
Okumura and equations fitting to the path loss curves by Hata. [4] The Cost-231
model also has correction for urban, suburban and open areas. The basic path loss
equation for urban areas is:
3.1.3
Physical Models or deterministic models
Physical models are the most widely used propagation models for fixed
broadband wireless systems. They rely on the basic principles of physics rather than
statistical outcomes from experiments to find the EM field at a point. Databases of
terrain elevations, clutter heights, atmospheric refractivity conditions, and rain
intensity rates are all used in the design process. Physical models may or may not be
site specific.
Non site specific models uses physical principles of EM wave
propagation to predict signal levels in a generic environment in order to develop
some simple relationships between the characteristics of that environment. On the
other hand, when particular elements of the propagation environment between the
transmitter and receiver are considered, the modal is considered site specific.
An example of a physical model is the “Ray-tracing” model. Ray-tracing is
not a cohesive mathematical technique but a collection of methods based on
27
geometric optics (GO), the uniform theory of diffraction (UTD), and other scattering
mechanisms, which can predict EM scattering from objects in the propagation
environment.[4]
This collection of field calculation methods are drawn upon mainly because
none alone can successfully deal with all the geometric features of propagation
environments likely to be encountered in broadband communication systems.
However, if there is an incomplete or insufficiently refined description of the
propagation environment, Ray tracing does not provide a complete and accurate
calculation of the field at all locations in the environment.
28
3.2
Propagation model used in the project
3.5.1 Overview
Empirical models can be split into two subcategories namely, time dispersive
and non-time dispersive [6]. The former type is designed to provide information
relating to the time dispersive characteristics of the channel i.e., the multipath delay
spread of the channel. An example of this type are the Stanford University Interim
(SUI) channel models developed under the Institute of Electrical and Electronic
Engineers (IEEE) 802.16 working group [1].
Examples of non-timedispersive
empirical models are ITU-R [12], Hata [13] and the COST-231 Hata model [8]. All
these models predict mean path loss as a function of various parameters, for example
distance, antenna heights etc. [15]
Among the empirical model, the SUI, COST 231 and ECC 33 model are the
most promises model in applying the FWA system. These models were used by
researchers from the University of Cambridge, UK to validate the propagation path
loss model mentioned above for FWA system at 3.5 GHz frequency band.
A
comprehensive set of propagation measurements were taken at 3.5 GHz in
Cambridge, UK is used to validate the applicability of the three models mentioned
previously for rural, suburban and urban environments.[14]
29
3.2.2 COST 231[19]
A model that is widely used for predicting path loss in mobile wireless
system is the COST-231 Hata model . It was devised as an extension to the HataOkumura model. The COST-231 Hata model is designed to be used in the frequency
band from 500 MHz to 2000 MHz. It also contains corrections for urban, suburban
and rural (flat) environments. Although its frequency range is outside that of the
measurements, its simplicity and the availability of correction factors has seen it
widely used for path loss prediction at this frequency band.
The basic equation for path loss in dB is ,
PL = 46.3 + 33.9 log10(f) - 13.82 log10(hb) – ahm + (44.9 - 6.55 log10(hb)) log10 d
+ Cm
(3.1)
where, f is the frequency in MHz, d is the distance between AP and CPE antennas in
km, and hb is the AP antenna height above ground level in metres. The parameter
Cm is defined as 0 dB for suburban or open environments and 3 dB for urban
environments.
The parameter ahm is defined for urban environments as [3]
ahm = 3.20(log10(11.75hr))2-4.97, for f > 400 MHz
(3.1.1)
For suburban or rural (flat) environments,
ahm = (1.1 log10 f - 0.7)hr - (1.56 log10 f - 0.8)
where, hr is the CPE antenna height above ground level.
(3.1.2)
30
Observation of [7] to [9] reveals that the path loss exponent of the predictions made
by COST-231 Hata model is given by,
nCOST = (44.9 - 6.55 log10(hb))/10
(3.1.3)
To evaluate the applicability of the COST-231 model for the 3.5 GHz band,
the model predictions are compared against measurements for three different
environments namely, rural (flat), suburban and urban.
31
3.2.3 Electronic Communication Committee – Report 33 (ECC 33)[19]
The original Okumura experimental data were gathered in the suburbs of
Tokyo [4]. The authors refer to urban areas subdivided into ‘large city’ and ‘medium
city’ categories. They also give correction factors for ‘suburban’ and ‘open’ areas.
Since the characteristics of a highly built-up area such as Tokyo are quite different to
those found in typical European suburban areas, use of the ‘medium city’ model is
recommended for European cities [5], [6]. Although the Hata-Okumura model [8] is
widely used for UHF bands its accuracy is questionable for higher frequencies. The
COST-231 model extended its use up to 2 GHz but it was proposed for mobile
systems having omni-directional CPE antennas sited less than 3 m above ground
level. A different approach as taken in [23], which extrapolated the original
measurements by Okumura and modified its assumptions so that it is more closely,
represents a FWA system. The path loss model presented in [23], is referred to here
as the ECC-33 model. The path loss is defined as,
PL = Afs + Abm - Gb – Gr
(3.2)
where, Afs, Abm, Gb and Gr are the free space attenuation, the basic median path loss,
the BS height gain factor and the terminal (CPE) height gain factor. They are
individually defined as,
Afs = 92.4 + 20 log10(d) + 20 log10(f)
(3.2.1)
Abm = 20.41 + 9.83 log10(d) + 7.894 log10(f) + 9.56[log10(f)]2
(3.2.2)
Gb = log10(hb/200){13.958 + 5.8[log10(d)]2}
(3.2.3)
32
For medium city environments,
Gr = [42.57 + 13.7 log10(f)][log10(hr) - 0.585]
(3.2.4)
where, f is the frequency in GHz, d is the distance between AP and CPE in km, hb is
the BS antenna height in meters and hr is the CPE antenna height in meters. The
medium city model is more appropriate for European cities whereas the large city
environment should only be used for cities having tall buildings. The prediction
using ECC 33 models with the medium city option are compared with SUI and
COST 231. Thus, the comparison will focus on urban and suburban only.
33
3.3
Stanford University Interim (SUI) model[8]
3.3.1 Introduction
The Hata-Okumura path loss model is the most widely used model in signal
strength prediction and in simulations. This model is valid in the frequency range of
500 to 1500 MHz, receiver distances greater than 1 Km from the base station and for
base station antenna heights of greater than 30 meters. Path loss regions can be
divided into categories dependent on the factors discussed above. These categories
are made based on the terrain and the obstacles in the path of LoS propagation. The
median path loss (PL in dB) for a fixed close in distance do is given by:
PL = A + 10 γ log 10 (d/d0) + s
d>d0
(3.3)
A = 20 log 10 (4 π d0 /λ)
(3.3.1)
γ = (a – b hb + c / hb)
(3.3.2)
where λ is the wavelength, γ is the path-loss exponent, hb is the height of the base
station for hb between 10m and 8m, d0 is the close-in distance (chosen as 100 m),
a,b,c are constants dependent upon the nature of the terrain, s represents the
shadowing effect which has a lognormal distribution and has typical values of
standard deviation in the range of 8-10dB.
34
Table 3.1 : Parameters for various terrain type [12]
Table 3.1 shows the parameter value for different type of terrain in SUI. For
light to moderate urban areas, Type A is most commonly used. The above model is
without the correction terms. Including the terms, it is obtain that the correction
factors for the operating frequency and for the CPE antenna height model are [8] as
below :
Xf = 6.0 log 10 (f/2000)
Xh = -10.8 log 10 (hr/2000)
= -20.0 log 10 (hr/2000)
(3.3.3)
for Terrain type A and B
(3.3.4)
for Terrain type C
(3.3.5)
where f is the frequency in MHz and hr is the CPE (Customer Premises Equipment)
antenna height above the ground in meters. The SUI model is used to predict the
path loss in all three environment, namely rural, suburban and urban by setting the
propagation delay for all three environment.
35
3.3.2 SUI parameter
The parametric view of the SUI channels is summarized in the following tables:
Table 3.2 : Summary of parametric view of SUI channel[ ]
Terrain Type
SUI Channels
C
SUI-1, SUI -2
B
SUI-3, SUI-4
A
SUI-5, SUI-6
K-factor : Low
Table 3.3 : SUI model parameter1[8]
Doppler
Low
Low delay
Moderate delay
High delay
spread
spread
spread
SUI-3
High
SUI-5
SUI-4
SUI-6
K-Factor : High
Table 3.4 : SUI model parameter2 [8]
Doppler
Low
High
Low delay
Moderate delay
High delay
spread
spread
spread
SUI-1, SUI-2
36
The generic structure for the SUI Channel model is given below :[1]
Figure 3.1
The generic structure of SUI channel models[8]
37
The above structure is general for Multiple Input Multiple Output (MIMO)
channels and includes other configurations like Single Input Single Output (SISO)
and Single Input Multiple Output (SIMO) as subsets. The SUI channel structure is
the same for the primary and interfering signals.[8]
Input Mixing Matrix:
This part models correlation between input signals if multiple transmitting
antennas are used.
Tapped Delay Line Matrix:
This part models the multipath fading of the channel. The multipath fading is
modeled as a tapped-delay line with 3 taps with non-uniform delays. The gain
associated with each tap is characterized by a distribution (Ricean with a K-factor>0,
or Rayleigh with K-factor = 0) and the maximum Doppler frequency.
Output Mixing Matrix:
This part models the correlation between output signals if multiple receiving
antennas are used.
38
3.6
Verification of performance for SUI channel in the existing 802.16a
3.6.1 Introduction
The existing channel model is actually a power delay profile model, where it
model the multipath component as a tapped delay line which having stochastic
variable.
It used the multipath signals to determine or evaluate the path loss
performance.
For a FWA scenario, it is often has a case that the delayed multipath
components will be of quite low power compared to the direct path because
considerable power is lost undergoing phenomena such as reflection and diffraction.
If the multipath arrive with long delays relative to the first arriving path and
particularly if they contain significant power, the result is a phenomenon known as
Inter Symbol Interference (ISI) where copies of several delayed symbols are smeared
on to the symbol currently being detected. This interference greatly worsens the
BER performance. To overcome the limitations caused by ISI, the CPE is often
equipped with an equalizer, which gives rise to a significant increase in complexity
of the modem in the CPE.
Thus by the problem arise, the OFDM were used to overcome the problems
where it is equipped with equalizer, FFT transformation and error control coding.
Below are the results of BER performance for 802.16a physical layer.
The
verification of the analysis can be found in [8], [9] ,[15] and [19].
There are six typical channel of SUI where it is set up base on the multipath
delay spread parameter. These channels are characterized by path loss, multipath
delay spread, fading characteristics, Doppler spread co-channel and adjacent channel
interference and antenna gain reduction.
39
3.6.2 BER performance of SUI channel ( for 3 taps)
3.6.2.1 Introduction
The simulator can determine the qualitative and quantitative performance of
the system and give intuitive understanding of the effects of certain parameter on the
system. Labview blocks display a performance curve that was developed for the
simulation. These include the bit error rate testers, spectrum analyzers and
constellation plotters. But for this project the performance will be focusing on the bit
error rate testers
3.3.2.2 BER Performance[8]
Various simulation runs were performed on the developed simulator[8], and
the results are shown below.
Figure 3.2
BER curves for AWGN
40
BER performance of the OFDM system with sub carrier modulations of
QPSK, 16 and 64 QAM under AWGN without coding. It is observed that the curves
are similar to the case for single carrier modulation, with a slight SNR penalty due to
the addition of a cyclic prefix.
Figure 3.3
BER curves for QPSK under various channel condition
Figure 3.3 shows the case for QPSK subcarrier modulation operating in the 6
different channels. It can be observed that aside from SUI-1 and 2, the system is no
longer operable in the other channels without equalization. This is because SUI-1
and 2 are flat fading channels, and SUI-3 – 6 are frequency selective channels.
41
Figure 3.4
Comparison of channel estimation method on QPSK modulation in
SUI-3
Figure 3.4 shows the performance of the different channel estimation
methods on QPSK subcarrier modulation OFDM operating in SUI-3. Note that comb
type estimation is not acceptable since the delay spread in SUI-3 results in a
coherence bandwidth that is
less than the frequency spacing of the pilot carriers of one OFDM symbol. Block
estimation, on the other hand, estimates the channel quite well since an entire symbol
is used as pilot carriers at the start of each frame. This also exploits the fact that we
have a very slow varying channel.
Furthermore, we can observe that block
estimation coupled with a low pass filter improves the performance of the system,
since aliasing in the linear interpolation is effectively suppressed [7].
42
Figure 3.5
Performance improvement for equalization and coding for SUI-3
Figure 3.5 shows the incremental improvement from using no equalization, to
using least squares block equalization with low pass filtering, and finally adding Rate
1/2, constraint length 7 convolution error control coding to the system. We can see
that error control coding gives almost 10dB gain to the system at 10–4 BER. The
FBWA channel also suffers from effects due to Doppler shifts. The Doppler power
spectral density (PSD) of the scatter (variable) component is distributed around 0 Hz
43
3.7
Summary
This chapter tells in details the propagation path loss used in the project. By
the verification of the 802.16a results, the channel model parameters were used to
compare the system performance between different propagation path loss models.
The path loss value from the equation of SUI, ECC 33 and COST 231 were used in
characterizing the wireless channels. The wireless channel are characterized by path
loss, multipath delay spread, fading characteristics, Doppler spread, co-channel and
adjacent channel interference and antenna gain reduction factor.[8] The concept of
the SUI channel with 3 multipath taps were applied in this project. Further details of
the setup will be in the following chapter.
CHAPTER 4
SIMULATION SPECIFICATION
4.1
Introduction :
The wireless channels set in the OFDM are based on the statistical
characterization of the parameter i.e in terms of mean and variance. All channel
impairment parameter are dependent on several factors such as terrain, topography,
density of vegetation, antenna height, wind speed, beam width, time of year and
other factors. In this project these factors are categorized by the urban, suburban and
rural environment.
The concept used in the system is using the existing 802.16a channel concept
where it applies the three-tap channel model. The environments were set in these
three taps where the multipath components are set into 0 ns for urban, 400 ns for
suburban and 900 ns for rural. These excess delay follows the existing simulator
excess delay in [8].
As a result, the bit error rate (BER) performance for the
propagation model mentioned in chapter 3 will be evaluated.
45
4.2
Design specification
The IEEE 802.16a standard specifies three air interfaces: single-carrier
modulation, 256-point transform OFDM and 2048-point transform OFDMA. This
project simulates the 256-point transform OFDM physical layer, since this air
interface is mandatory for operation in license-exempt bands [5]. This physical layer
has the parameters as in table 2.2. The simulator was tested on 3.5 GHz frequency
with channel bandwidth of 7 MHz. This channel bandwidth was set in the channel
block in the programming part of Labview 8.0.
Basically, the wireless channel characterized by the multipath delay spread
component. It is a very important factors that need to take into consideration for the
simulation. The typical value of the RMS delay spread is used[8]. This project
applied the path loss value to the wireless channel by using the same concept. Below
are the multipath component used in the existing 802.16a physical layer.[8]
K-factor : The k-factor is defined as the ratio of the power in the fixed component
to
the power in the variable component. If it is zero, the channel is Rayleigh. For
larger
values, the channel is assumed to be Ricean in nature. As the objective of the project
is to look into the performance of the simulator in NLOS and multipath environment,
the k factor is set in Rayleigh condition.
Average Fade duration : It is the average period of time for which the received
signal is below a specified level. It is generally chosen with respect to the mean
signal
power level [20].
Level Crossing Rate : It is the number of times the signal crosses a given level per
second.
46
Doppler Spectrum : Due to motion of the mobile receiver as well as the nature of
the
path, the transmitted frequencies undergo Doppler frequency shifts. The Doppler
frequency is primarily influenced by wind speed and traffic density. For fixed
wireless
channels the Doppler Power Spectral Density of the variable component is mainly
distributed around f = 0Hz. The approximation for the shape of the spectrum is
given
below:
(4.1)
The function is parameterized by a maximum Doppler frequency fm .
Spatial Fading Characteristics : The spatial fading characteristics include the
co-channel interference, coherence distance and the antenna gain reduction factor.
47
4.3
Simulation results representation
As in [8], the 802.16a OFDM physical layer was successfully modeled and
simulated in Labview 7.0. Thus this projects also using the application of Labview
to evaluate the performance of the OFDM in different channel model. But the
Labview 8.0 were used because it is an enhancement of Labview 7.0 where it does
not need mathlab interfacing to operate.
Labview is a dataflow programming language that is suitable for
communication modeling and simulation.[21].
Labview provides a graphical
interface that is useful for interactive control of key simulation parameters, which
provides better intuitive understanding of the system. Additionally, the block
diagrams developed in it could be synthesized into Field Programmable Gate Arrays
(FPGAs) and programmable processors, which make it a great candidate for rapid
prototyping of these systems.
Labview system simulation was proved successfully implemented in [8]. It
provides a suitable environment for the effective analysis of the performance of the
802.16a OFDM physical layer. Various simulation parameters can be automatically
chosen by the simulator for predefined simulation runs, or interactively changed by
the user while a simulation is running. These parameters include a choice between
the various rates supported by the standard (QPSK, 16-QAM, and 64-QAM), the
length of the cyclic prefix, the different channel parameters, and the several channel
estimation techniques.
48
Figure 4.1
Front panel of the amended simulator of 802.16a
The simulator channel parameters were changed i.e changing the control file
that controls all blocks in the simulator. The urban, suburban and rural environments
were presented in a tap delay as presented in the original 802.16a simulator. The
channel model for each environment were placed at 0ns, 400ns and 900ns away from
the arriving path(the excess delay is 400ns). This excess delay follows the original
excess delay of 802.16a simulator. The antenna directivity of 300 vertical polarized
were used in the system which follows the setup of the existing 802.16a simulator.
This amended system is a little bit different from the existing simulator in term of the
selecting the channel tap. The existing simulator chooses all three taps to operate in
certain terrain condition. But, the amended simulator were given choices to select
the environment wanted by clicking the on/off button However, the simulator can
operate in all three taps simultaneously, nevertheless it will give unpredictable
outcome.
49
Figure 4.2
Block diagram of the amended simulator of 802.16a
Figure 4.2 shows the block diagram represents the programming part of
Labview 8.0. The block diagram contains of two main loop of programming. The
right hand side is for the controller setup and the left hand side is for the OFDM
setup. Once the Run button is click, the system gives one value and it goes through
the two main loop. When the channel parameters were chosen, the controller setup
will change the type of controller to gives one value to calculate the Eb/No. But, the
operation will only be done after one cycle of looping..
50
Figure 4.3
Front panel of the OFDM in 802.16a simulator
51
Figure 4.4
Block diagram of the OFDM in 802.16a simulator
From Figure 4.4, the TranParams component of the block diagram is the main
controller of the OFDM. It contains the multipath components, where this parameter
will be applied to each block. In this project, the channel model will be change
based on the path loss equation in Chapter 3.
52
Figure 4.5
Front panel of the channel model in 802.16a simulator
As can be seen in the figure above, there are control parameter at the input
and output of the channel model. These controllers are related to each other. Thus,
when applying the input control parameter, the output control parameter follows the
input control parameter. The Sampling Frequency component controls the frequency
bands used. The frequency bands of 3.5 GHz with 7 MHz channel bandwidth is
setup in the Frequency Sampling. The author has set the frequency sampling as the
bandwidth, therefore the bandwidth setup for 3.5 GHz frequency band were place.
53
Figure 4.6
Block diagram of the channel model in 802.16a simulator
The channel model consists of the Adaptive White Gaussian Noise, channel
model parameter and Rician Fading block diagram. The channel model contains of
the multipath components that were setup base on the propagation path loss equation.
The three tap processing parameters were set up in the Rician Fading block. It can
be seen that the TranParamsIn parameter being extract by the block diagrams. Each
parameter from TranParamIn have their own characteristic that will activate the each
block diagrams.
54
Figure 4.7
Front panel of the Rician Fading in 802.16a simulator
55
Figure 4.8
Block diagram of the Rician Fading in 802.16a simulator
The block diagram of the Rician Fading shows that the 3 tap processing done
in the simulator. The output of the block diagram is channel impulse and signal
value being filtered out by the filter set inside the block diagram. The tap, gives
values to the main programming to generate the bit-error rate (BER).
56
Figure 4.9
Front panel of the one tap processing in 802.16a simulator
57
Figure 4.10
Block diagram of the one tap processing in 802.16a simulator
The tap processing is controlled by the on/off button that we choose in the
front panel of the 802.16a simulator.
If the user choose one environment i.e
suburban, the system will activate the 2nd tap, thus it will give value to the tap gain
and tap index and the BER will be calculated. The system will also deactivate the
other tap and thus giving the value of tap gain and tap index. The taps are connected
to each other and it will gives a value to the main loop. The processes repeat until 8
cycles because the pilot symbol being injected to the system is 8 bit.
58
The AWGN plays an important role in the channel model where we can
measure the intersymbol interference by this component. The following figures are
the figures of AWGN.
Figure 4.11
Figure 4.12
Front panel of the AWGN in 802.16a simulator
Block diagram of the AWGN in 802.16a simulator
59
Figure 4.13
Front panel of the channel model in 802.16a simulator
60
Figure 4.14
Block diagram of the channel model in 802.16a simulator
The channel model set in the block diagram above is a very important
component in the channel model. It results due to the scatterings nature of the
environment.
The parameters set up in the block diagram are the multipath
components i.e based on the path loss equation of SUI, COST 231 and ECC 33.
These parameters were set in the main controller of OFDM.
61
4.4
Parameter set up in the channel model
The channel mentioned in chapter 3 is set up base on the delay in urban,
suburban and rural environment, mean power, k factor, Doppler shift and on/off
button. The path loss gain from equation (3.1) to (3.3.5) was used. The CPE antenna
height of 10m and 6m were used in the path loss equation to evaluate the
performance in different antenna height. These antenna heights were chosen because
it is used in the Cambridge experiment to evaluate the performance of FWA.
Therefore, to validate the applicability of FWA in OFDM, comparison of
performance in different CPE antenna heights referring to the Cambridge experiment
validate the results. The wavelengths of the SUI model used are 50 mm for CPE
antenna height of 10m and 5 mm for CPE antenna height of 6m. Note that in table
4.1 the distance and base station antenna height for each environment are different.
Below are the setup used in the channel model:
Table 4.1 : Parameter set up in the channel model
Urban
Suburban
Rural
SUI – type B
SUI – type B
SUI – Type C
COST 231 – urban
COST 231 - suburban
COST 231 - rural
correction factor equation
correction factor equation
correction factor equation
used
used
used
ECC 33 – medium city
ECC 33 - medium city
correction factor equation
correction factor equation
used
used
BS antenna height = 17
BS antenna height = 38
BS antenna height = 15
m
m
m
Distance = 1 km
Distance = 2 km
Distance > 5 km
62
4.5
Summary
This chapter tells about the methodology of the OFDM system. The front
panel is the user interface where user can change or choose the parameter required.
The block diagram is the programming parts that control the front panel. The block
diagram were controlled by a controller namely TranParam.ctl.
This controller
control every block of the OFDM. Changing the controller will affect the whole
system, therefore the parameter set in the controller must be set globally where it can
applied to the whole block of the OFDM.
It is noted that the ECC 33 path loss equation does not have any correction
factor for rural environment. Further explanation about the statement can be found in
[14]. For rural environment of SUI, type C was used because it is considered as
‘flat’ environment. [19]. For the urban and suburban environment, SUI type B were
used to represent the urban and suburban environment.
Based on the parameters set in the channel model as above, the system
performance in terms of bit-error-rate (BER) of SUI, COST 231 and ECC 33 were
evaluated at different CPE antenna height. From here, we can see the relation of the
propagation model equation with the system performance. Further details of the
results will be explained in chapter 5.
CHAPTER 5
SYSTEM PERFORMANCE
5.4
Introduction
Propagation models are used extensively in network planning, particularly for
conducting practicability studies and during initial deployment. They are also very
useful for performing interference studies as the deployment proceeds. Empirical
propagation models have shore up in both research and industrial communities owing
to their speed of execution and their limited confidence on detailed knowledge of the
terrain. Thus, the empirical models used in the project have been used to see the
performance for FWA in urban, suburban and rural environment. The empirical
model used a statistical analysis to predict the FWA access. Verification of system
performance in the existing 802.16a physical layer can be found in chapter 3. The
concept of the system performance is based on the bit error rate (BER) versus Eb/No
(dB) which is the energy per bit to noise density. As the power increase, the Eb/No
increase making less noise, less error and less loss.
The prediction of system
performance in this fixed Wimax simulator will be based on the graph drawn by the
system.
64
5.5
System performance results in term of bit-error-rate (BER)
SUI
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
none
CPE antenna height
:
10m
Environment
:
Urban
Figure 5.1
SUI (CPE antenna height = 10 m, URBAN)
65
SUI
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
none
CPE antenna height
:
10m
Environment
:
Suburban
Figure 5.2
SUI (CPE antenna height = 10 m, SUBURBAN)
66
SUI
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
none
CPE antenna height
:
10m
Environment
:
Rural
Figure 5.3
SUI (CPE antenna height = 10 m, RURAL)
From the Figures of 5.1, 5.2 and 5.3, it can be seen that the system
performance similar to the performance of the existing simulator. The SUI-1 and
SUI-2 which is type C in the existing simulator has much better performance than the
one set in the rural environment as in Figure 5.3. This may caused by the k factor
67
setup in the parameter channel. The existing simulator used various value of k factor
to determine the performance but for this project the k factor value is set into
Rayleigh condition which is k=0 to represent the NLOS condition.
68
For the QPSK modulation techniques, with no channel estimation and at CPE
antenna height of 10m, the overall results of all type of environment are as follows:
Urban
Suburban
Rural
Figure 5.4
SUI (CPE antenna height = 10m , ALL ENVIRONMENT)
The system performance at the CPE antenna height of 10m in SUI model
does not need any equalization. The channel may be corrected by the path loss
equation and at this height of antenna there may be less obstruction.
69
SUI
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
none
CPE antenna height
:
6m
Environment
:
Urban
Figure 5.5
SUI (CPE antenna height = 6 m, no Ch. Estimation, URBAN)
70
SUI
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
LS Comb & LS Block
CPE antenna height
:
6m
Environment
:
Urban
LS Comb
LS Block
Figure 5.6
SUI (CPE antenna height = 6 m, URBAN)
From the simulation, it is shown that the channel estimation of LS Block
gives better performance of the channel model. This block pattern leads to excellent
frequency resolution. The block pattern eliminates other components which caused
71
multipath with high delays. The simulator use a low pass filter in the block LS
estimation to eliminate the noise.
72
SUI
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
no channel estimation & LS Block
CPE antenna height
:
6m
Environment
:
Suburban
No Ch. Estimation
LS - Block
Figure 5.7
SUI (CPE antenna height = 6 m, SUBURBAN)
It can be seen that for antenna height of 6m, the SUI channel needs channel
equalization and channel estimation of LS Block.
73
SUI
Modulation Techniques
:
64 QAM
Cyclic Prefix
:
¼
Channel Estimation
:
LS Block
CPE antenna height
:
6m
Environment
:
Suburban
Urban
Suburban
Rural
Figure 5.8
SUI (CPE antenna height = 6 m, ALL ENVIRONMENT)
The modulation techniques used is 64 QAM, it seems like the modulation
techniques influence the performance of the digital system.
The modulation
techniques used for Figure 5.7 represent 8 bit/symbol being transmit to the OFDM.
74
The major advantage of the OFDM is it transmits four frequencies at one time where
carrier frequency is in these frequencies.
It can also overcome the effect of
frequency selectivity in transmission channel. From the figure above, the BER
worsen, it may caused by the CPE antenna height of 6m where obstruction may
happen at this level.
75
COST 231
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
All estimation techniques
CPE antenna height
:
10 m
Environment
:
Urban
No Ch. Estimation
LS Comb
LMMSE
LS Block
Figure 5.9
COST 231 (CPE antenna height = 10 m, URBAN)
The results show the performance in various channel estimation. The LS
Block plays an important role in improving the performance of the system.
76
COST 231
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
LS Block
CPE antenna height
:
10 m
Environment
:
All environment
Urban
Suburban
Rural
Figure 5.10
COST 231 (CPE antenna height = 10 m, ALL ENVIRONMENT)
77
COST 231
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
LS Block
CPE antenna height
:
6m
Environment
:
All environment
Urban
Suburban
Rural
Figure 5.11
COST 231 (CPE antenna height = 6 m, ALL ENVIRONMENT)
From Figures 5.8 and 5.9, it can be seen that the suburban environment gives
better performance in both antenna height. For antenna height of 10m, the suburban
and urban environment gives the same BER value of 10-1 at 20 dB gain. However,
78
the performance is improved in CPE antenna height of 6m where the BER value goes
to 10-2 at 20 dB gain. The rural environment shows worst results. This may be due
to the frequency coverage for COST 231 is only applicable up to 2 GHz.
79
ECC 33
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
LS Block
CPE antenna height
:
10 m
Environment
:
All environment
Urban
Suburban
Figure 5.12
ECC 33 (CPE antenna height = 10 m, ALL ENVIRONMENT)
80
ECC 33
Modulation Techniques
:
QPSK
Cyclic Prefix
:
¼
Channel Estimation
:
LS Block
CPE antenna height
:
6m
Environment
:
All environment
Urban
Suburban
Figure 5.13
ECC 33 (CPE antenna height = 6 m, ALL ENVIRONMENT)
The Figures of 5.12 and Figure 5.13 gives better results for suburban
environment. This may caused by the frequency coverage of the ECC 33 is up to
GHz.
Moreover it is actually a propagation model used to promote Wimax
81
deployment. It is based on the European propagation path. However, the rural
correction factors equation of this model has not been developed yet.
82
5.3
Analysis of results for all environments in all channel model
Performance of the 802.16a simulator for CPE antenna height of 10m at QPSK
modulation techniques and LS Block channel estimation
Urban
Suburban
Rural
Figure 5.4
SUI (CPE antenna height = 10m , ALL ENVIRONMENT)
83
Urban
Suburban
Rural
Figure 5.12
ECC 33 (CPE antenna height = 10 m, ALL ENVIRONMENT)
84
Urban
Suburban
Figure 5.12
ECC 33 (CPE antenna height = 10 m, ALL ENVIRONMENT)
85
Performance of the 802.16a simulator for CPE antenna height of 6m on QPSK
modulation techniques and LS Block channel estimation
Urban
Suburban
SUI
Rural
Figure 5.14
SUI (CPE antenna height = 6 m, QPSK, ALL ENVIRONMENT)
86
Urban
Suburban
Rural
Figure 5.11
COST 231 (CPE antenna height = 6 m, ALL ENVIRONMENT)
87
Urban
Suburban
Figure 5.13
ECC 33 (CPE antenna height = 6 m, ALL ENVIRONMENT)
88
5.4
Summary
As can be seen in the figures above, the system performance shows good
results. ECC 33 gives the best performance in the suburban environment for both
CPE antenna height. It may caused by the correction factor equation in the path loss
equation or by the frequency coverage of the model. The SUI model only applicable
for 1 to 4 GHz coverage, while COST 231 applicable up to 2 GHz coverage and ECC
33 applicable up to GHz frequency coverage. Even though COST 231 frequency
range is outside the measurements, its simplicity and availability of the correction
factors has seen widely used for path loss prediction at 3.5 GHz frequency band.
Moreover, it also contains corrections for urban, suburban and rural environment.
Thus by the results, we can see the relation between propagation path loss and
system performance. The more frequency coverage of the path loss, the system
performance will be better. However, other factors should also be considered to
evaluate the system performance. The details conclusion of the system performance
will be explained in the next chapter.
CHAPTER 6
CONCLUSION AND SUGGESTION
6.4
Conclusion
The aim of this project was to simulate and evaluate the performance of
different channel model of 802.16a physical layer.
The results obtained from
Chapter 3 proved that the simulator can be used to simulate the OFDM system
performance of Wimax. The results obtained from Chapter 5 proved that the path
loss propagation model has been successfully simulated at 3.5 GHz frequency band.
The proposed propagation models are COST 231 and ECC 33.
For the existing 802.16a simulator, the channel model proposes different sets
of parameters for three types of terrains, the categories are not specified in a
particularly systematic manner and do not explicitly include urban and suburban
environment. The BER performance results of the existing 802.16a physical layer
were compared with the theoretical results presented in various paper s [8] But for
this project, it is focused the urban, suburban and rural environment.
90
The ECC 33 model shows good performance in suburban environment. It is
recommended for urban and suburban environment for Wimax application.
Moreover the frequency coverage of this model is in GHz which it is applicable for
Wimax application. For antenna height of 10m, it gives good results in suburban
environment where the BER is at 10-4 at 20 dB gain. But for antenna height of 6m,
the BER performances worsen. It may caused by the obstruction in between the
propagation path loss. The SUI gives good performance than other propagation
model at antenna height of 6m.
By these analysis, it can be concluded that for the ECC 33, it is best apply at
the CPE antenna height of 10m and the SUI are best apply at the CPE antenna height
of 6m. The COST 231 model is not suitable for fixed Wimax application because it
does not gives much different whether in different antenna height or different
environment. However, studies shown that COST 231 is being used as a propagation
model for 802.20 standard. The equation of COST 231 may be used in Wimax
application provided that the correction factors of path loss being changed.
The performance of the OFDM can be improves by using LS Block channel
estimation. It shows that the ability of OFDM can improved the system performance
by changing the channel estimation. Thus, OFDM are recommended to be used in
Wimax application.
The Labview system simulation was successfully implemented [8].
It
provides a suitable environment for the effective analysis of the performance of the
802.16a OFDM physical layer. This baseband model shows various simulation
parameters that can be automatically chosen by the simulator for predefined
simulation runs, or interactively changed by the user while a simulation is running.
These parameters include a choice between the various rates supported by the
standard (QPSK, 16 QAM and 64 QAM), the length of the cyclic prefix, the different
channel parameters and the channel estimation techniques.
91
6.5
Suggestion for future work
The digital system performance of OFDM has been successfully designed on
three type of channel model which is SUI, COST 231 and ECC 33. The frequency
band used are 3.5 GHz. It is recommended that the simulator being tested on
different frequency band i.e 2.3 GHz and 2.5 Ghz which is the allowable frequency
of Wimax in Malaysia.
It is recommended to apply the channel model that suits Malaysia’s
geographical environment because the propagation path loss used are tested and
developed in Europe and US. Other than that, it is recommended to use the simulator
in other type of propagation model that is other than empirical model i.e the Longley
Rice model.
The simulator could be applied in different CPE antenna height. The antenna
height used in this project is 10m and 6m because it is commonly used in predicting
the FWA access.
92
6.6
Wimax in Malaysia
On 16th March 2007, the Energy, Water and Communications Minister, Datuk
Seri Dr. Lim Keng Yek has announced that the Wimax service in Malaysia will be
operated at 2.3 GHz spectrum in licensed band and 2.4 GHz in unlicensed band. The
list of the Wimax license winners are also announced on that day. The Wimax
license winners are GreenPacket, Redtone, YTL and Asiapace. It was unexpected
news from the Telco’s company. The results from the Malaysian Communications
and Multimedia Commission (MCMC) are clearly against the incumbent Telcos and
existing ISPs. Non of the incumbents manage to successfully bid the license, that
includes Telekom Malaysia, Maxis, DiGi and Time Telekom. The question is, how
do these small companies run or deploy Wimax infrastructure in Malaysia. The
existing ISPs have the infrastructure but does not owned the licensed spectrum.. It
will create a competition between mobile technologies against Wimax. One of the
solution is, these unpopular license winners should cooperate with the existing ISPs
to provide broadband wireless services in Malaysia. Thus, by this cooperation it will
lead Malaysia towards the 2020 vision.
REFERENCES
1.
Theodore S. Rappaport, Wireless Communications Principles and Practice.
2nd ed. United States of America : Prentice Hall. 2002
2.
Henry L. Bertoni, Radio Propagation for Modern Wireless Systems. 1st ed.
United States of America : Prentice Hall. 2000
3.
H. R. Anderson. Fixed Broadband Wireless System Design. John Wiley & Co.
2003
4.
Y. Okumura. Review of the Electrical Communications Laboratory.Field
strength and it’s variability in VHF and UHF land-mobile radio services. Vol.
16. September-October 1968
5.
S.R. Saunders. Antenna and Propagation for Wireless Communciation
Systems. John Wiley & Sons Ltd. 1999
6.
J.D. Parsons. The Mobile Radio Propagation Channel. Second Edition. John
Wiley & Sons Ltd. 2000
7.
S. Coleri, M. Ergen, A. Puri and A. Bahai. IEEE Transaction on Broadcasting.
Channel estimation techniques based on pilot arrangement in OFDM systems.
Vol. 48, no. 3, pp. 223-229, Sep. 2002
8.
Alden J. Doyle, Kyungtae Han, Suvid Nadkarni, Kalpana Seshadrinathan,
Raghuveer Simha, Ian C. Wong. “Performance Evaluation of IEEE 802.16a
physical
layer
using
simulation”.
EE381K-11
(14980)
,
Wireless
Communication Project Rep.2004
9.
L.J Greenstein, V. Erceg, Y.S. Yeh and M.V Clark. IEEE Trans. Veh.
Technol. A new path-gain/delay-spread propagation model for digital
cellular channels. May 1997 vol. 46 : 2,
10.
J.W. Porter and J.A Thweatt. Microwave Propagation Characteristics in the
MMDS Frequency Band. ICC 2000 Conference Proceedings, pp. 1578-1582
94
11.
L.J. Greeenstein, S. Ghassemzadeh, V. Erceg and D.G Michelson. Ricean Kfactors in narrowband fixed wireless channels: Theory, experiments and
statistical models. WPMC 99 Conference Proceedings. Amsterdam:
September 1999
12.
V.Erceg, et. Al.. IEEE Journal on Selected Areas in Communication.
Empirical Based Path Loss Model For Wireless Channel in Suburban
Environments. Jul. 1999
13.
David Huo. IEEE 802.20 Working Group on Mobile Broadband Wireless
Access . Issues on Path Loss Model. 15 May 2005
14.
Draft ECC 33 Report. The Analysis Of The Coexistence Of FWA Cells In The
3.4 to 3.8 GHz Bands. 2005
15.
H.Cordovo, P. Boets and L. Van Biesen. Insight Analysis Into Wimax
Standard and Its Trends. B-1050 Brussels, Belgium
16.
V.Erceg, V. S Hari, M.S Smith, D.S Baum. Channel models for fixed wireless
application. Contribution to IEEE 802.16.3, Jul. 2001
17.
Recommendation ITU-R P.1546. Method for point-to-area predictions for
terrestrial services in the frequency range 30 MHz to 3000 MHz. tech. rep.
International Telecommunication Union, 2001
18.
M.Hata. IEEE Transactions on Vehicular Technology Empirical formula for
propagation loss in land mobile radio services., vol. VT-29, pp. 317-325,
September 1981
19.
JNL,V.S. Abhayawardhana, I.J. Wassell, D. Crosby, M.P. Sellars, M.G.
Brown, Comparison of Empirical Propagation Path Loss Models for Fixed
Wireless Access Systems. Cambridge, 2004
20.
T.S. Rappaport. Wireless Communications, Principles and Principles.
Prentice Hall, New Jersey. 2002
21.
P. Das and D. Koch. IEEE Proceedings of Southeast Con. On the use of
Visual Programming Languages for Communication System simulation. Apr.
1991
22.
V.Erceg, K.V.S. Hari, M.S. Smith, D.S Baum, “ Channel Model For
Fixed Wireless Applications”, IEEE 802.16 Broadband Wireless Access
Working Group, IEEE 802.16.3c-01/29r4
95
23.
Ofcom Ref: AY4463, “A Study On The Efficient Dimensioning Of
Broadband Wireless Access Networks”, fin. Rep., Cambridge University
Engineering Dept. & Cambridge Broadband Ltd. , 2004
24.
Electronic Communication Committee (ECC) within the European
Conference of Postal and Telecommunications Administration (CEPT), “The
analysis of the coexistence of FWA cells in the 3.4 – 3.8 GHz band,” tech.
Rep., ECC Report 33, May 2003
25.
COST Action 231, “Digital mobile radio towards future generation systems,
final report,” tech. Rep. European Communities, EUR 18957, 1999
26.
H.R. Anderson, Fixed Broadband Wireless System Design. John Wiley & Co.
2003