Diversity in MW
Diversity in MW
This Thesis is a part (30 ECTS) of Master of Science degree (120 ECTS) in Electrical
Engineering emphasis on Telecommunication
February 12
Abstract
Different diversity techniques such as Maximal-Ratio Combining (MRC), Equal-Gain
Combining (EGC) and Selection Combining (SC) are described and analyzed. Two branches
(N=2) diversity systems that are used for pre-detection combining have been investigated and
computed. The statistics of carrier to noise ratio (CNR) and carrier to interference ratio (CIR)
without diversity assuming Rayleigh fading model have been examined and then measured
for diversity systems. The probability of error ( ) vs CNR and ( ) versus CIR have also
been obtained. The fading dynamic range of the instantaneous CNR and CIR is reduced
remarkably when diversity systems are used [1]. For a certain average probability of error, a
higher valued average CNR and CIR is in need for non-diversity systems [1]. But a smaller
valued of CNR and CIR are compared to diversity systems. The overall conclusion is that
maximal-ratio combining (MRC) achieves the best performance improvement compared to
other combining methods. Diversity techniques are very useful to improve the performance of
high speed wireless channel to transmit data and information. The problems which considered
in this thesis are not new but I have tried to organize, prove and analyze in new ways.
3
ACKNOWLEDGEMENT
I am grateful to my thesis supervisor, Dr. Magnus Nilsson for his sincere help. The way he
encouraged me to finish this thesis, was really great. It was impossible to finish this thesis
without his support.
Contents
Abstract …………………………………………………………………………2
Acknowledgement ………………………………………………………...3
Chapter 1 Introduction
1.1 Background …………………………………………………………….9
Networks ………………………………………………………………16
3.2.3 Interference
References ………………………………………………………………71
8
Chapter 1
Introduction
1.1 Background:
Communications between creatures is very important in our world. Wired communication and
wireless communication are the major types of communications effective in the present world.
Some examples of wireless communications are ‘Hand held CB (Citizens’ Band) radio’ and
‘cellular phone’. There are two main categories of Wireless communications such as mobile
communications and fixed wireless communications. Customer needs and technology
requirements, these two play vital role to make their own market in the wireless
communications [2]. Mobility or non-tethered communications are required for the mobile
communications. Mobility of the mobile communications ensures to communicate anytime
and anywhere. Mobile communications technology must be able to allow roaming. Roaming
is the ability to provide usability to the mobile phone users while outside their own network.
However, fixed wireless is just an alternative to wired communications. The fixed wireless
users do not get the mobility facility but they need cost effective communications from their
fixed locations. The alternative of wireless in the means of providing service is the only
means. When the customers attempt to communicate from the remote locations, Satellite is
the only alternative. Wireless communication is a revolution of human communications. The
technology to send idea and information to remote location could not even be imagined before
wireless communication technology was invented. The wireless communication technology
makes that dream true of the human being. Wireless hand held devices such as mini-
computers and phones deliver the world to our fingertips whatever our locations are, with
enough speed and flexibility. Wireless cellular networks are growing rapidly around the world
day by day un-imaginably and it seems this trend will to continue for several years. Over the
last few years, the subscribers of wireless communications have an exponential growth. The
ongoing progress in radio technology provides more and more new and improved services.
Current wireless services include transmission of fax, voice and low-speed data flexibly [1].
Usability of more band width in communications with consuming interactive multimedia
services like video-on-demand like TV viewing and internet access for transferring essential
data is supported in wireless communications.
9
Chapter 2
Overview of Wireless Communications
The basics of wireless communication systems are described in chapter2, based on their
services such as speech communication, data transmission, etc. The generations of
communication technologies such as 1st generation, 2nd generation, 2nd –plus generation, 3rd
generation and 4th generation are described in this chapter. Multiple access formats for
wireless communication and their operations are also described briefly at the last section of
this chapter.
Europe USA
Circuit switched connections are used in the second generation technologies for transporting
data and providing data transmission rate in between 9.6 to 14.4 Kbps [11]. These
technologies implement a high level flow control and error connection to provide reliable data
transfer. With second generation systems, a single cellular channel can be shared by multiple
users. This system function reduces congestion and provides access for more users. The
multiple access methods to the users provide extensive coverage with a proved and reliable
communications infrastructure in 2G. The standards within 2G in use worldwide are the
following:
13
Wired broadband connections and portable bandwidth are the same level with 3G technology
and the data transfer rates is up to 2Mbps in fixed applications. But data transfer rate is 128
Kbps in a car and 384 Kbps when a device remains stationary. When the targeted speed is
achieved fully, wireless technology will be more attractive to users those are interested in
internet browsing, gaming with wireless connection, and listening to music with broadband
mobile connection [3]. Voice and text messaging are available with the present mobile
networks. On the other hand, 3G networks are allowed us more faster and complex data
transmission such as streaming video and audio, satellite navigation, video conferencing and
sharing with interactive application. Packet switched data access provided with this kind of
networks to the internet end-to-end IP connection. A mobile phone has been connected
automatically to the internet when the cell activated. Then subscribers get the similar facilities
to today’s fixed-line internet connection with significant add-ons such as location-based and
highly personalized services. Some of these features already achieved and some are on the
road to come to public hands.
15
3G technologies allow handsets to connect to network permanently and use only available
capacity when they receive or transmit packets. Subscribers are expected to pay for the
volume of data transfer ignoring the length of talk time.
Although 3G technology seems complicated still this technology makes our life very easy and
flexible. Much of those we have already been getting. Much of these facilities we have
already been getting with limited conditioning usability. The quality media information and
entertainment will be at our fingertips irrespectively location and time. The end user device
will not be just a mobile phone but a terminal after achieving all those functionalities with 3G
wireless communications.
Standards:
The 3G technology began to standardize in 1990s under the supervision of the International
Telecommunication Union (ITU). The goal of 3G was full interoperability and inter-activity
of mobile systems providing capability of value-added services. The ITU called for Radio
Transmission Technology (RTT) in the year 1998 with proposals for IMT-2000 (International
Mobile Communication for the year 2000) [13]. The ITU IMT-2000 standards are currently
separated into two major organizations reflecting the two 3G camps: 3GPP (3G partnership
project for wideband CDMA standards based on backward compatibility with GSM and IS-
136/PDC) and 3GPP2 (3G partnership Project for cdma2000 standards based on backward
compatibility with IS-95). The eventual 3G evolution for 2G CDMA systems leads to
cdma2000. CDMA 2000 are developed in several research based on IS-95 and IS-95B
technologies. The eventual 3G evolution for GSM, IS-136, and PDC system leads to
wideband CDMA (W-CDMA) (Wideband Code Division Multiple Access) also called
Universal Mobile Communication Service (UMTS). W-CDMA was invented based on the
network fundamental of GSM as well as improved version of GSM and IS-136 through
EDGE. It is fair to say that these two major 3G technology camps, cdma2000 and W-CDMA,
is remain popular throughout the early part of the 21st century. In January 1998, The ETSI
incorporated W-CDMA standard in 1998 was also incorporated with terrestrial Radio Access
specifications, and W-CDAM and UMTS are used synonymously sometimes [13]. IMT-2000
ensures that these technologies can work in different networks even in the GSM as well as
American ANSI networks. Most of the major network operators in Europe and Asia are
committed to the W-CDMA standard for 3G mobile communications while other parts of the
world implements other standards. In Asia Pacific and North America, the next generation
wireless network is going to be mainly based on CDMA2000, and China, the largest market
for mobile communication in the world, will be using TD-SCDMA standard for 3G networks.
16
Availability:
Upgrading 2G to 3G technology requires a big amount of capital investment. Some European
Government auctioned off radio spectrum bands to accommodate the 3G networks. In the UK,
for example, five 3G mobile licenses were auctioned off at a total of $35 billion USD with the
expecting cost for each license between $4 billion and $9 billion USD to form out their 3G
network [3]. For this reason carriers have opposed to upgrade networks because they wanted
to be ensured and see a real demand for high speed wireless data, and many of them viewed
that 2.5G is more than just an interim solution following the reason that it delivers significant
bandwidth improvements at a lower cost.
However, major wireless service providers estimated the high costs of deploying 3G services
and warned about technical difficulties such as 3G handset and network infrastructure
foundation, a few are already working on W-CDMA in Europe and Japan, And some other
country also applying the technology step by step from lower income to high income
countries. A market structure and status of deployment of mobile internet technologies in
some of these countries are shown in table 3. NTT DoCoMo in Japan released a third
generation phone service FOMA (Freedom of Mobile multimedia Access) in major urban
location of the country. The receiving data of FOMA data at 384 Kbps and transmitting data
rate is 64 Kbps. Subscribers get access for almost everything in a limit such as movie trailers,
sports highlights, music, video clips and news feeds.
The strategies group predicted that there would be 9.5 million 3G mobile high-speed data
subscribers by 2005 but it crossed the limit of the predictors. Within 6 years of 3G
introduction, it gains 402 million subscribers by 2008, and 30 million subscribers are added in
every quarter. According to the UMTS Forum prediction by 2010 data services will represent
$300 billion or 66% of all worldwide 3G revenues. Japan, South Korea and the USA already
the number of 3G subscribers have already surpassed that of 2G subscribers [14].
There are two key elements which are required to deliver a legitimate 4G network. First is the
ability to roam across different wireless network standards with the one device; and the
second, and most obvious, is a higher level of bandwidth [3].
17
Though function of 4G is not defined fully yet, still it can predict that 4G technology will
replace 3G, and it will use a combination of Wi-MAX and Wi-Fi. Standard bodies or carriers
have not defined definitely what exactly 4G will be but it is expected that IP end to end and
streaming quality with high quality will be featured in 4G technologies.
There are two competing 4G standards working mainly. One of them is a joint effort by
Hewlett Packard and another is Japan’s NTT DoCoMo to create Moto-Media. Moreover, the
wireless World Research Forum specifications with the backing of some of the Europe’s
largest phone markets.
Most of the wireless observer bodies made a question that is how the 4G market going to be,
and when can the industries reasonably expected to invest in a new network. Some of the
analysts have estimated that the 4G mobile systems would have more 50 million subscribers
by the end of year 2007, and it would account for 14 percent of total mobile data revenues.
But it is estimated that the subscriber for 3G/4G who use WCDMA/HSDPA, TD-SCDMA,
mobile Wi-MAX, EV-DO and LTE networks was 230 million in 2007 and rose to 375 million
in 2008 [15]. But most of the analyst estimate the technology to be ready around 2008-2010.
Nokia and Samsung have teamed up to create 4G wireless equipment; a move demonstrates
the support for the 4G mobile systems.
Table 3: Comparison of Data Services for 2G, 2.5G and 3G Networks [1] [2] [3]
Web browsing Short text 100KB web page takes 100KB web page
screens takes Approx.2
approx. 30 sec to sec to download
download
(User)
2
1
Frequency
Available
Bandwidth
Picture 2: FDMA schemes in which different users are assigned different frequency bands
User
2
1
Time
Available
Bandwidth
Picture 3: TDMA schemes in which each user occupies a cyclically repeating time slot
TDMA shares a single carrier frequency amongst many users, where each user uses non-
overlapping time slot. In the case of FDMA, if a channel is not in use then the corresponding
time slot sits idle and can’t be used by other users. In TDMA system data transmission for
users is not continuous but occurs in bursts transmission [3]. Synchronization overhead is
required in TDMA systems Because of this burst transmission. Moreover, it is important to
use guard slots for separating users. In general, TDMA mobile systems are more complex
than the FDMA systems.
Each user of a multiple access system like FDMA or TDMA is limited by bandwidth and time
allocated to it, the degradation caused by background noise, multipath fading and shadowing
effects. Another weakness is that its relatively low frequency reuse factor of FDMA and
TDMA.
A common performance degradation source to all multiple access system is fading in its
terrestrial environments. Fading is occurs by interference between different versions of the
transmitted signals which arrive at the receiver at very minor different time. This phenomenon
is very severe when each channel is allocated narrow bandwidth with the FDMA systems.
Spread spectrum communication system is applied with Code Division Multiple Access in
which multiple users have access to the same frequency band. Spread spectrum refers to
power spreading over a given transmission bandwidth in DS-CDMA (Direct-Sequence-
CDMA). This is accomplished by spreading the base-band binary data by means of high
speed pseudo noise (PN) code (called the chip rate). The composite high speed data are then
modulated and demodulated over the air.
In direct sequence CDMA (DS-CDMA) systems, the narrowband message signal is multiplied
by a very large-bandwidth signal which is called the spreading signal. The same carrier
frequencies are being used by all users in DS-CDMA systems and may transmit
simultaneously. Each user has its own spreading signal which is almost orthogonal to all other
users’ spreading signals. A correlation operation is performed by the receiver for detecting the
message addressed to a given user. The signals from other users appear as noise because of
de-correlation operation. The receiver requires the spreading signal for detecting the message
signal used by the transmitter. Each user operates independently to the other users with no
message which is called uncoordinated transmission.
CHAPTER 3
Performance Degradation Factors of Wireless
Communications
Local to base
A typical wireless propagation environment at outdoor is seen in figure 2.1, where the mobile
wireless node is communicating with base station or a wireless access point. The transmitted
signal from the mobile may travel two ways. The signal may reach the access point directly
which is called line of sight (LOS) [16] or it may reach the access point through multiple
reflections on local scatters (buildings, mountains, bridges, trees etc.). As a result the received
signal has multiple random attenuation and of course with some delays. Moreover, the
mobility or movement of either the nodes or the scattering objects may cause these random
fluctuations to vary with time. Time variation results in random increasing or decreasing of
the transmitted signal strength over time period. In conclusion, an undesired interference may
be occurred by a shared wireless propagation because of simultaneous transmissions to the
transmitted signal. The combined effects of these discussed factors put wireless in a
challenging communication environment.
The difference from wired communication such as twisted pair, coaxial cable, optical fiber, is
that the transmission channel can’t be predicted or random and it can vary over as very short
time scales as microseconds [16]. The consequence gives the limited power resources which
leads the communication systems to new challenges in signal transmission. Moreover, the
equitable sharing of limited resources is an important challenge while the medium is
frequently shared by several users. In addition, freedom of mobility is an important factor to
the users and for that, the end users need to be located in order to information to them. This
situation makes the network topology very complex and a challenging dealing is needed for
24
the challenging and complex wireless network. Finally, wireless devices are needed to
connect with an internet network infrastructure which is wired connected. As result a newly
challenge creates which is interoperability of these disparate communication media.
Services are increasing dramatically because of users demand. Excess services lead air time
usage. Moreover, easy usability and cheaper services creates more and more users in wireless
communication industries. But the radio spectrum resources are limited, so system capacity is
a challenge to the wireless device and infrastructure designers. Some more challenges include
like-
1. Unfriendly medium due to presence of noise, interference, multipath and time variations.
Path loss
Shadowing loss
Channel spreading
Channel fading
Interference
Noise
Different copies of signal undergo different attenuation, distortion, phase shift and delays
during transmission. The overall performance of system degraded severely due to the above
problem.
25
The path loss in power varies gradually because of the signal attenuation which is determined
by the geometry of the path creates from received power in its whole area. Simply, a signal
attenuates according to propagation in a real channel. If a radio wave transmitted from a point
source in free space the path loss is power is,
Where is wavelength of the signal; and d is distance between source and receiver.
The power of the signal decays is the squire of the distance. The same thing happens in
landline wireless communication also.
The mean power of a signal decays as the nth power of the distance:
3.2.2 Shadowing:
Second factor is shadowing loss which effects on wireless channel. It is very difficult to
model Shadowing Loss. Shadowing is happen due to absorption of radio wave in the
propagation by scattering structures. In a normal sense, shadowing causes by the land surface
and obstructions between the transmitter and receiver in communication system. It remains
independent in concern of distance between receiver and transmitter. It is a random variable
which varies according to types of environment where radio wave propagates. Experiment
shows that it is modeled as well by a log normal distribution, for example, the attenuation
behavior is like, ASL [16]. Where log (A) always follows the well known Gaussian
distribution. It can be said in other way like that, the measured attenuation is Gaussian
distribution in decibels. Moreover, Variation of time occurs when scattering environment is
changed. It can be said by an example like that that when a mobile moves to a corner of a
street, the time variation of shadowing loss is very difficult t model, and there are very few
experiment to study this situation. If we think about the quality and compare with path loss
and fading, we can describe that, shadowing loss varies faster than the path loss, but it is
slower than the fading (loss) [16] [3].
If a mobile unit is on moving, it does not affect the short term characteristics of the
shadowing. It’s because of the relatively large obstacles. However, shadowing behavior is
determined from the nature of the terrain surrounding of the base station, the mobile unit and
the height of the antenna. Shadowing adds some additional fluctuations around the system. As
a result, the received power as mean varies around the system area. Normally, shadowing is
modeled as a random process which is slowly time-varying and also multiplicative. If we
avoid all the other impairments, the received signal is [16]:
If we use an observation interval, we might let g(t) is a constant g. Normally, this s modeled
as a log normal random variable which has a density function as:
Here ln (g) is a Gaussian random variable and is mean value. Here is the variance. The
variance and the mean of the shadowing loss are measured in decibels. , the function of
the terrain and the antenna height, can be ranged between 4 dB to 12 dB in cellular and micro-
cellular environment.
27
3.2.3 Interference:
The source of interference in a radio system can be originated at the system itself or it can be
located at external source. The self-centered interference can be divided into two types: inter-
cell and intra-cell interface. The inter-cell comes from other mobile stations and it normally
approximately 60% of total interference in a system. In the uplink case, the intra-cell interface
comes from other mobile station cells. If a communicating system uses multi-option access
technique, the users of the same cell or the neighboring cells of the system can be interfaced
with each other. Moreover, multiple propagation paths in a multi-paths channels interface
with each other. There are two types of interference, such as:
Inter-symbol-interference (ISI)
Co-channel-interference (CCI)
Amplitude
Time
Symbol Time
Amplitude
Time
ISI can be controlled by slowing down the transmit data. When the received signal damps
down, then the next pulse of information should send [1]. The time it takes to damp is called
delay spread, and the original time of pulse is called symbol time. To suppressed ISI, we can
select a space-time filter that equalizes the channel. Moreover it can be controlled by using a
maximum likelihood sequence detector.
Receiver 1 Receiver 2
Interference
Signal
Cell 2 Cell 1
3.2.4 Noise:
There are two types of noise which have a role play in radio wave propagation, natural noise
and man-made noise. The main source of 1st type of noise is the ignition systems of vehicles.
However, natural noise source such as galactic noise, solar, atmospheric noise, has less effect
in land-mobile communication systems but lots in radio channel. The noise generated in the
components in the communication systems also corrupt the received signals besides the
natural and man-made noise. The characteristics of different noise sources are simply
different but the noise process most commonly and frequently modeled with Additive White
Gaussian Noise (AWGN). If the receiver components introduced thermal noise and dominates
the noise process then AWGN assumption the component is valid [3]. The AWN should be
taken into account when considering a system with very few users as it will impose a lower
bound on the bit error rate. However, if we want to think about the system capacity then we
consider the state in which the system has large number of users. In this case the noise is not
important factor like interference introduced by the other users in the system.
30
Doppler spread
Delay spread
If surrounding scatters are present with multi-directions and a pure tone is spread over a finite
bandwidth, then Doppler power spectrum is the Fourier transform of the received signal in
time domain and the support of the Doppler power spectrum is the Doppler spread. If the
scatters equally distributed in angle, then the Doppler power spectrum is given by classical
spectrum.
S (f) =
, ----------------------------------- (3.6)
S(f)+B(f - ) ……………………………………………………………….(3.7)
The result of Doppler spread in the channel characteristics is that it change the channel
characteristics sharply in time, it gives rise to the so called time selectivity. The fading
channel can be considered as constant during the coherent time and the coherent time is
inversely proportional to the Doppler spread. Typical values of Doppler spread are 10 to 250
Hz (Suburban areas), 10 to 20 Hz (Urban areas), 10 to 100 Hz (Office areas). A person
walking at 10km/h induce a maximum Doppler spread of 22 Hz at 2.4 GHz. (Reference:
Broadband wireless access by Benny Bing, page- 5). A Doppler spread makes good channel
tracking which is an important feature of receiver design.
In mobile communication, there is not only multipath propagation exists but also its time
varying. The phenomena results a time-varying fading channel. It’s very difficult to
communicate through this kind of fading channel in communication systems. But there is
some special technique may be taken to achieve satisfactory performance. In wireless
communication, the received signal comes from both direct part and the path of scattering,
32
reflections and diffractions. Because of the propagation loss, the effect of the terrain
configuration implement small-scale long term fading which is also called shadowing fading
and it changes with the atmosphere and electrical constants. Natural and man-made structures
such as buildings, traffic, motion, trees, hills and the other nearby environment would cause
the multi-path fading on the received signal called short-term fading.
Transmitted impulse
Average received power
t Tm
≈ ……………………………………………… (3.9)
34
Signal tone f
Doppler shift
…………………………………… (3.10)
Here and are received and transmitted powers, is wavelength of the radio wave, d is
the range are gains of transmitter and receiver respectively. The main path is
accompanied sometimes by a surface reflected path which may destructively interfere with
the primary path. There is a model called path loss model is developed to handle this effect.
The model is [2]:
……………………………………. (3.11)
Here are the effective heights of transmitter and receiver respectively. This path loss
model built in according to an inverse fourth power law and the path loss exponent may vary
from 2.5 to 5 depending on the environment.
35
However, considering the same parameters if the bandwidth of the transmitted signal is large
then different frequency component may undergo of different fading. This kind of channel is
called frequency selective fading. The symbol duration is small comparing to coherent
bandwidth due to their reciprocal relationship. Delays between different paths can be large
respect to the symbol duration. It is then expected to receive multiple copies of signal.
When the symbol duration is small comparing to coherent time , then the channel is
called slow fading channel. This type of channel often modeled as time invariant channels
36
over a number of symbol intervals. Moreover, slow varying channel parameters can be
estimated by different types of estimation techniques.
When there is a close or smaller coherent time to/than symbol duration, the channels is fast
fading or time-selective fading. It is still a difficult phenomenon to estimate the parameters of
the channel in a fast fading channel.
If there is a line-of-sight (LOS) propagation path in a dense urban environment, then there is a
zero mean Gaussian distribution which is leading to a Rayleigh distributed attenuation and the
channels is called Rayleigh fading channel. The envelope fluctuations of the signal follow the
Rayleigh PDF in Rayleigh fading channel and the signal comes from almost the directions
with the same average power. The Rayleigh fading is predominant and worst case in typical
land mobile communication systems. Multipath fading phenomena occurs in three situations.
If the mobile unit and the nearby scattering objects all moving, If the mobile unit is standing
but the nearby scattering objects are moving and if the mobile unit and the nearby scattering
objects all standing.
38
CHAPTER 4
Diversity Technique for Wireless Channel
Several types of diversity techniques are described in this chapter. In section 4.3, several
types of diversity techniques such as Multipath Diversity, Spatial Diversity, Time Diversity,
Polarization Diversity, Angle diversity, Antenna Diversity are described. Every diversity
technique has some branches. Those are also introduced with the respective diversity
technique description. For example, Transmit diversity has two branches such as close-loop
transmit diversity and open-loop transmit diversity.
One of these problems is that the dynamic range of the transmitter and the required
transmitting power is extremely high if it’s intended to fully compensate the fading. This is
impossible because of the radiation power limitations, the cost and the size of the amplifiers,
and the limited battery power in the portable unit. Moreover, excess transmitted power
increases the interference level at the other channels and users in the system unit. Another
problem in power control (PC) approach that a feedback link is needed for the channel unless
the operation of the radio channel is in time division duplex (TDD) mode. In a TDD system,
the same frequency band is used for the downlink transmission from the base station to
mobile unit and for the uplink transmission from the mobile unit to the base station. As a
result, the transmitted signal undergoes the fading channel as the received signal due to its
reciprocal characteristic of the channel, the transmitted power of transmitter is adjusted
according to the received signal power. The feedback information usage decreases throughout
the channel and increases the complexity in the system [4]. Even an appropriate feedback link
may not available in some application.
Using PC the fading can’t be overcome completely but the attenuation may compensate
considerably. It can mention that large-scale fading can be compensated as well in the uplink
of a system, for example CDMA. But stringent power control is required in prevention near-
far problem in the system. The rate of large-scale fading is simply slow, as a result it can be
tracked well and the delay in the feedback of the power control commands can be neglected
39
comparing with the rapid fading. On the other hand, small-scale fading can result in such
rapid variations in the signal power that even the power control can’t follow them.
Another approach to minimize fading effect in a system is to supply multiple replicas of the
transmitted signal to the receiver which already have passed through different fading
channels. The result of this approach is that the probability that all replicas of the signal will
fade simultaneously is reduced [1]. This is called diversity and it is effectively and commonly
used to overcome degradation in performance due to interference and fading. If there is D
number of fading channels and the probability of any one channel may fade under some
threshold is P, then the probability of all D signals which fade below the threshold is . The
number of diversity channel D is called diversity order in the system.
Multipath/frequency diversity
Spatial/space diversity
Temporal/time diversity
Polarization diversity
Angle diversity
Antenna diversity
40
Moreover, frequency hopping (FH) might be use to achieve such kind of diversity instead of
sending multiple frequency replicas over different affecting frequencies. Frequency band can
be changed many times per symbol in fast frequency hopping (FFH) and this results
frequency diversity on each transmitted symbol. This process is very beneficial in an
environment where there is a partial band jamming in channel [5]. However, a different
frequency band is used for a burst of symbol in slow frequency hopping (SFH). If SFH is
combined with time domain coding and interleaving, it is seen as like an additional block
interleaving in the frequency domain. This kind of spreading of information in a frequency
domain introduces frequency diversity benefit.
Frequency diversity can also be implied as in the case of multipath diversity. Transmission of
a wideband signal is given by the following equation where the bandwidth is more than the
coherence bandwidth of the previously used channel and this results a frequency
selective fading [3].
In a sufficient wide signal bandwidth, multipath components can resolve. In the result, it is
possible to obtain different independently fading signal. The number of resolvable multipath
given by the following equation is used to approximate the maximum achievable diversity
order for multipath diversity:
L= [ ]+1
In general, the information signals are modulated through different carriers M in frequency
diversity scheme. It is important that different signals undergo independent fading. The
carriers should be separated by at least coherent bandwidth from each other. L copies of
signals are optimally combined at the receiver to make a statistic decision. The maximum
ratio combiner is the optimal combiner.
41
Ant Ant
Tx Rx
Comp Sw
mod ∑
Tx Rx
On the other hand, physical complexity restricts its application widely. Like several receiving
antennas use in space diversity, several transmission antennas also can be used to send several
copies of transmitted signals. This kind of diversity can be employed combating frequency
and time selective fading both.
There are two types of spatial diversity techniques such as receive diversity and transmit
diversity.
Intentional redundancy is introduced into the transmitted signal to achieve time diversity in
the temporal domain. Redundancy can be done by repetition of channel coding. To make
repetition coding, information bearing signals are transmitted in several time slots. But the
separation between time slots should be more or equal than the coherent time of the channel
to obtain independent faded signals which helps to gain full diversity advantages. Moreover, it
43
Channel coding, when there is an error control capability, may be performed separately or
with modulation. If we want to perform coding and modulation separately, then we can use
conventional error control coding to achieve redundancy in the form of extra symbols. The
advantage to this case is that the transmission of extra symbols leads to lower efficiency in
system bandwidth. On the other hand, if we want to perform coding and modulation together
(the process also called as coded modulation) then the redundancy comes from the channel-set
expansion. But the most beneficial process is ceded modulation because it is bandwidth
efficient since it allows error control coding without increasing the bandwidth.
If the successive channel symbols are independent and the channel is Rayleigh fading then the
effective minimum time diversity of the channel code is equal to its lowest hamming distance
in signal symbols. But the problem is that if the fading is not very fast then there are no
successive and independent channel symbols. As a result we can’t achieve full diversity in
the system. Longer fade duration with slower fading results burst errors if there is no use of
interleaving or effective burst error corresponding codes. The purpose of the interleaving is to
impose further spread out in time coded signal symbols to achieve independent errors at the
place of burst error. Therefore, a code is used which is capable to correct errors.
In general, a desired signal is transmitted in M different periods of time in time diversity. For
example, every symbol is transmitted M times. As it is mentioned earlier that intervals
between the transmitted symbols should be at least coherence time to make ensure that
different copies of the same symbol undergo independent fading. Maximum ratio combiner
can be used to obtain optimal combiner [10]. If we send the same symbol M times then it
applies the (M, I) repetition code. We can also use non-trivial coding. Error control coding
and interleaving is an effective way to combat time selective or fast fading.
Ant Ant
D
SW Rx Sw
D Tx
D Comp
Spatial diversity may achieve by using multiple antennas with independent polarizations in
the same location instead of multiple antennas in use in different locations. This is the method
of polarization diversity. If an implementation of spatial diversity with small dimensions is
desired, this is very attractive process. Normally two orthogonally polarized antennas are used
on horizontal and vertical planes or with a slope of to employ polarization diversity [6].
Experiments show that polarization diversity may obtain in dense scattering environments
when there is line of sight (LOS) and non-line of sight (non-LOS) situations.
A multiple fixed narrow-beam antenna or an array antenna, which is fully adaptive, transmits
and receives much more energy between the mobile terminal and base station comparing to a
wide beam antenna system [9]. On the other hand, a beam forming system using multiple
antennas may experiences multipath fading sometimes when the multipath components are
applied from very close angles. The base station antenna system in a dense urban environment
experiences antenna gain reduction and the multipath components will spread in wider angles.
In this case, it is wise to use angle diversity to avoid deep multipath fading by choosing the
best beam to collect energy. The size of the angle diversity is very smaller comparing to the
space diversity system where we need the wide separation between the receiving antennas.
45
This diversity system needs a number of directional antennas those responds independently to
wave propagation. An antenna response to a wave propagates at a specific angle and receives
a faded signal which is uncorrelated with other signals.
Multipath components in a cluster those have individual arrival angles travel through different
paths and employ different fading. Very basic procedure to obtain angle diversity is to fix
antennas with narrow beam widths different sector in the system. Then the arriving multi-
paths from the different beam directions are resolved and combined advantageously. This
procedure not only creates diversity but also increases the antenna gain and reduces
interference by providing angular discrimination. It is mentioned earlier that one more method
of achieving diversity is to use antenna array with adaptive beam forming which termed as
path diversity sometimes.
In the feedback schemes, the scheme uses either direct or indirect feedback from the receiver
to the transmitter to adjust the channel conditions. Indirect feedback can be used in time
division duplex (TDD) systems due to reciprocal channel in uplink and downlink. The same
diversity combing system, used for received signals from multiple antennas, can be converted
to distribute and weight for the transmission signals from multiple antennas. Indirect feedback
accuracy depends on the channel change during the dwell time between the uplink and
downlink transmission. In the direct feedback the receiver provides the transmission
information about the preferred weighting of the transmission antennas based on the
measurement of the receiver. The feedback information amount is a tradeoff between antenna
weighting accuracy and data throughput reduction of the uplink.
The next category schemes use feed forward information to compensate the receiver response
in the channel. The pilot symbol insertion is the commonly used method to obtain channel
estimation for coherent detection. For spreading the information across antennas, this
diversity schemes use feed forward information and linear processing at the transmitter. In the
reception section, the schemes estimate the state of the channel based on pilot symbols using
maximum-likelihood decoding sequence.
The last schemes category does not need feedback or feed forward information. Non-coherent
or differential detection is used generally in these schemes because coherent detection without
feed forward may be impossible due to fast fading. The phase sweeping transmit diversity is
one example of blind schemes. We may need to combat burst error due to slow fading by
using two transmit antennas, which are uncorrelated spatially, to generate force fading. After
that the interleaving depths and short block codes are used for making correction of forward
error. Quaternary differential phase shift keying (QDPSK) with differential detection is used
in these schemes category.
Again there are two main classes of transmit diversity, they are:
The close-loop (CL) uses a feedback channel to send Channel state information (CSI)
required at the receiver. The receiver sends back to transmitter which uses for signal design
while the Open-loop does not need CSI. On the condition of availability of CSI, which is ideal
meant error free and instantaneous, at the transmitters, close-loop techniques have an SNR
advantage of 10 ( ) over open-loop techniques due to the factor of ‘array gain’ [5].
But there are some other factors which are responsible to degrade the performance of close-
loop techniques such as channel estimation errors at the receiver, errors in feedback link (for
noise, interference, quantization effects), and feedback delay. Amongst them, feedback delay
causes a mismatch between the available CSI and the actual CSI. All of these factors
combined with extra bandwidth and system complexity resources. The process make open-
loop technique more attractive as a robust means for improving downlink performance for
high mobility applications while close-loop techniques become attractive under low-mobility
applications.
Delay diversity is one simple example for open-loop diversity. In a delay diversity scheme
employed for a single base station, where the symbols are transmitted different times through
multiple antennas, creates an artificial multipath distortion. It is essential to resolve the
multipath distortion. For that, a Maximum Likelihood Sequence Estimator (MLSE) or a
Minimum Mean Squire Error (MMSE) equalizer is used then obtain the targeted diversity
gain. So the main theme of delay diversity is that to transmit of the same information for M
antennas simultaneously with individual delay and the operating flat fading channel
intentionally into a frequency selective channel.
The capacity of receiver diversity with possible combining can be presented by following
equation:
Where = average signal to noise ratio (SNR) at each of the receiver branches.
The power gain of the Rayleigh fading channel is Random variable with unit variance. The
probability density function (PDF) of is a central chi-squire with variance and zero
mean. Representation of random variable means the variation of received power from
the branches of fading diversity. In the result, the fading channel capacity is a random
variable.
However, in the transmit diversity, the total transmitted power distributed amongst the
transmitted antennas. As a result the average signal-to-noise-ratio (SNR) of transmitted signal
components from the branches of diversity is a fraction of total SNR at the receiver.
Therefore, if the number of transmitting antenna is J then the capacity of the transmit diversity
is:
Here 1, using the large number rule. So, the capacity becomes as Gaussian
capacity . The fact is applicable when the number of transmitting antennas
increases. The practical gain in term of capacity is very little if more than four transmit
antenna and only one receive antenna is used.
If we use combined transmit-receive diversity then the capacity is absolutely greater than if
we use only receive diversity. The minimum level of capacity gain for a combined transmit-
receive diversity when J M is:
Here the transmitted signal components are un-coupled and there is no interference in the
received signal. It is observable that the capacity of combined transmit-receive diversity
increases linearly as a function of J when J M. It is feasible to obtain almost all the diversity
capacity by using six transmit antennas and two receive antennas.
49
Chapter 5
Combining Techniques of Diversity
If we want to get benefit from diversity technique then we must need to combine some
diversity technique to get advantage. Therefore, diversity combining concepts are described
in this chapter, in section 5.2, we described Maximal-ratio Combining (MRC), in section 5.3,
we described Equal-gain Combining (EGC) and in section 4.4, we described Selection
Combining (SC). Above three combining systems are our main focuses and applied in
experiments to improve performance in wireless communication systems. Block diagram of
combining methods are drawn in this chapter. Switched combining method, Periodic
combining method, Phase-sweeping methods are described shortly and their diagrams also
drawn in this chapter.
The combining processes which use to combine multiple diversity branches in the reception,
has two classes such as post-detection combing and pre-detection combining. The signals
from diversity branches are combined coherently before detection in pre-detection combining.
However, signals are detected individually before combining in post-detection. The
performance of communication system is the same for both combining techniques for
coherent detection. However, the performance of communication system is better by using
pre-detection combining for non-coherent detection. It does mean that there is no effect in
performance by the type of combining procedure for the coherent modulation case. The post-
detection combining is not complex in non-coherent detection, results very common in use.
There is a difference in system performance when used pre-detection combining and post-
detection combining for non-coherent detection such as frequency modulation (FM)
discriminator or differential detection schemes. Moreover, the terms pre-detection and post-
50
detection are also indicates the time of combining means when the combining is performed,
before or after the hard decision.
conjugate of channel gain [1]. As a result the phase-shifts are compensated in the diversity
channels and the signals coming from strong diversity branches which has low level noise, are
weighted more comparing to the signals from the weak branches with high level of noise. The
term in weighting can be neglected conditioning that has equal value for all d. Then
the realization of the combiner needs the estimation of gains in complex channel and it does
not need any estimation of the power of noise.
It is feasible to employ MRC in transmission process of transmit diversity. But in this case the
transmitter should get proper feedback information about the sub-channels state between
single receive antenna and multiple transmit antennas. However, it is not feasible to weight
transmissions from multiple antennas optimally for every receiving antenna, in a combined
transmit-receive diversity channel. Moreover, if interference is limited in a communication
system, then there is a scheme which combines the diversity branches in order to maximize
the signal-to-interference-plus-noise ratio may allow much better performance than MRC
provides. The assumption is valid for spatially white Gaussian noise if we can observe noise
power at the receiver where just thermal noise is accounted. If we use the same type antenna
elements then the thermal noise power is uncorrelated and equal for each branch.
51
Ant
Rx
Ant
Det
Det
Rx
Q C
Q A
A
B
B
t
t
C
The EGC can employ in the reception of diversity with coherent modulation. The envelope
gains of diversity channels are neglected in EGC and the diversity branches are combined
52
here with equal weights but conjugate phase. The structure of equal-gain combining (EGC) is
as following since there is no envelope gain estimation of the channel.
Rx
Det
Det
Rx
The general form of selection combining is to monitor all the diversity branches and select the
best one (the one which has the highest SNR) for detection. Therefore we can say that SC is
not a combining method but a selection procedure at the available diversity. However,
measuring SNR is quite difficult because the system has to select it in a very short time. But
selecting the branch with the highest SNR is similar to select the branch with highest received
power when average power of noise is the same on each branch. Therefore, it is practical to
select the branch which has the largest signal composition, noise and interference. If there is
an availability of feedback information about the channel state of the diversity branch the
selection combining also can be used in transmission.
53
Rx
Comp Det
Rx
Performance improvement obtained by the switching method leys on the value of threshold
selection, the delay of time that creates from the loop of feedback of monitoring estimation,
switching and decision. Moreover, phase transients and envelope of a carrier can reduce the
improvement of performance. In the system of angle modulation, for example, GSM, the
phase transient is responsible to create errors in detection stream of data. In this case, a pre-
detection band pass filter may be used to remove envelope transients.
Ant
Switch
Ant
Rx Det
Comp
Fixed
Threshold
(a)
54
Ant
Switch
Ant
Rx Det
Comp Estimation
(b)
Picture 15: Switching combining methods with fixed threshold (a) and variable threshold (b).
Ant
Switch
Ant
Rx Det
Oscillator
Ant
Ant
Rx Det
Ant
Sweep signal
CHAPTER 6
Performance Analysis of Different Combining Techniques
--------------------------- (6.1)
Where is depends on the type of channel estimation and can be interpreted as cross
correlation coefficient. cab be expressed using following equation in the case of coherent
detection in conditioning perfect channel estimation.
------------------------------------------------------------ (6.2)
is received SNR for every diversity channel in average. The equation of M-array Phase
Shift Keying (PSK) in the corresponding SNR per bit is:
----------------------------------------------------------------- (6.3)
A QPSK signal can be expressed as two Binary Phase Shift Keying (BPSK) signals in phase
quadrature [3]. As a result, if there is no cross talk or interference between signals on two
quadratures then the bit error probability for QPSK and BPSK is similar. If there is no
additional interference in an AWGN channel then the noise in the in-phase and quadrature
components is independent. Therefore, the equation of bit error probability of coherent BPSK
is used to find the bit error probability of coherent QPSK or the reverse way. On the other
57
hand, if the bit error probability is a function of , the received SNR per channel in average
must be normalized by the number of bits per symbol in QPSK.
The bit error probability in AWGN channel is expressed by the following equation:
------------------------------------------- (6.4)
The equation (5.1) can’t apply if the SNR in all branches are identical. If the mean SNR of all
diversity branches are distinct then the bit error probability for coherent QPSK is defined by:
----------------------------- (6.5)
Here
-------------------------------------------- (6.6)
The estimated correlation between two diversity branches is equal. In such dual diversity, the
bit error probability of QPSK and BPSK with MRC for correlated fading is:
--------------------- (6.7)
------------------------------------------- (6.8)
In the case of CIR, is the ratio of squired signal envelope to squired interference envelope.
Both of the envelopes are independent and Rayleigh distributed random variables. The pdf of
( 0) is F-distribution and measured by the equation,
----------------------------------------- (6.9)
58
Let the CNR of each diversity branch , where i=1, 2 and let the CNR of combined branch
.
Rayleigh fades are independent on the diversity branches and faded signals have equal
average power such as .
Here (i=1,2) means the average value of (i=1,2). The equations of pdf for the three
combining methods of are:
It is approximated that << . The probability that does not exceed the specified value of ,
can be calculated by the integral,
---------------------------------------- (6.12)
Maximal-ratio combining -------------------------- (6.13a)
Equal gain combining ------------------------------ (6.13b)
Selection or switching combining ------------- (6.13c)
This has been assumed a perfect-pilot co-phasing scheme for maximal ratio and equal gain
combining methods. However, comparably larger desired signal power algorithm assumed for
59
selection combining method. The mutual correlation within four Rayleigh fades which is
visible on desired signals and the interference which is undesired found at respective branches
are negligible. As like as average CNRs, average CIRs are equal to each. Therefore,
. The value of out-age is,
---------------------------------- (6.14)
The resultant study of the simplest MSK diversity systems can be useful in advance
modulated wireless diversity structures.
-------------------------- (6.15)
-------------------------------- (6.16)
----------------------------------- (6.17)
In the similar way following equation is obtained for maximal-ratio combining using the
equation (5.10a),
60
] ------------------------------- (6.18)
These relationships can also be used for the case of M-branches diversity reception instead of
2-branch.
------------------------------------------ (6.19a)
--------------------------------------------------- (6.19b)
-------------------------------------- (6.19c)
Now we substitute equations from (6.19) into equation (6.17) to find the equations for
in selection method,
-------------------------------- (6.20a)
----------------------------------------------------------- (6.20b)
-------------------------------------- (6.20c)
Again, we substitute equations from (6.19) into equation (6.18) to find in the case
maximal ratio combining method,
---------------------------- (6.21a)
--------------------------------------------------------------------- (6.21b)
---------------------------------------- (6.21c)
-------------------------------------------------- (6.22a)
------------------------------ (6.22b)
------------------------------------------- (5.22c)
Chapter 7
Computational Method and Computational Result
7.1 Computational Method:
The parameters used in the equations and computations are:
= A specified value of
= A specified value of
= Probability of error
63
Program 1:
Equation (6.12) for equations (6.8), (6.11a) and (6.11c) gives graphs in terms of probability
distribution of versus the ratio ( / ) in dB.
2
10
no diversity
1 selection diversity
10
maximal ratio combining
0
10
-1
10
probability
-2
10
-3
10
-4
10
-5
10
-20 -15 -10 -5 0 5 10
10log10(Eb/No)(dB)
Program2:
Equation (6.14) for equations (6.9), (6.13a) and (6.13c) gives graphs in term of probability
distribution of versus the ratio (/ ) in dB.
1
10
no diversity
selection diversity
0
10 maximal ratio combining
-1
10
probability
-2
10
-3
10
-4
10
-5
10
-20 -15 -10 -5 0 5 10
10log10(Eb/No)(dB)
Program3:
Equation (6.19) and (6.20) gives graphs for average -versus-CNR performance for the
selection method.
4
10
discriminator detection
differential detection
2
10 coherent detection
0
10
average Pe
-2
10
-4
10
-6
10
-8
10
-20 -15 -10 -5 0 5 10
average CNR in dB
Program4:
Equation (6.19) and (6.20) gives graphs for average -versus-CNR performance for
maximal-ratio-combining method.
4
10
discriminator detection
differential detection
2
10 coherent detection
0
10
average Pe
-2
10
-4
10
-6
10
-8
10
-20 -15 -10 -5 0 5 10
Average CNR in dB
Figure 7.4: Average Pe-versus-CNR performance for the maximal ratio combining method.
67
Program5:
Equations (6.19a), (6.19c) and (6.22a, (6.22c) gives graphs for average -versus-CIR
performance for maximal ratio combining method.
4
10
discriminator detection
differential detection
2
10 coherent detection
0
10
average Pe
-2
10
-4
10
-6
10
-8
10
-20 -15 -10 -5 0 5 10
average CIR in dB
Figure 7.5: Average Pe-versus-CIR performance for the maximal ratio combining method.
68
The intention of this study is to investigate two branches (N=2) of diversity systems using
pre-detection combining which is very useful and effective in radio mobile communications.
The first examination is done to study the statistics of CNR and CIR without diversity
concerning a Rayleigh fading model. The same experiment is done with diversity. The study
for -versus-CNR and -versus-CIR performance improvements also has been done.
It can be concluded from figure 7.1 that the fading dynamic range of the instantaneous CNR
is reduced remarkably by the use of the diversity techniques and there is only a slight
difference among the performance improvements of the combining methods. is the average
value of . We considered average CNR at point 0 in X-axis.
In figure 7.2, it is shown that the fading dynamic range of the instantaneous CIR is reduced
remarkably by diversity techniques. There is only a slight difference in performance
improvements of the various combining methods. The is the average value of . The
probability is the probability that CIR is less than a determined value of . We considered
average CIR at point 0 in Y-axis.
Figure 7.3 and 7.4 indicate that for a certain average Pe, a higher amount of average CNR is
required for non-diversity system, while a smaller amount of CNR is required for a two
branch diversity technique. Diversity improvement effect is shown in case of an MSK
modulation system in nonselective Rayleigh-fading environment.
Figure 7.5 indicates that as in figure 7.3 and 7.4, in order to improve the average Pe at a
particular time, it is only necessary to increase the CIR by a smaller factor when two-branch
diversity improvement is available, whereas an increase of higher CIR is required in the case
of no diversity reception. In this case MSK and FQPSK, modulated and coherently
demodulated GMSK systems are considered in the Rayleigh-fading environment.
We can see that the maximal-ratio combining (MRC) achieves the best performance
improvement comparing to other methods.
69
Chapter 8
Discussion and Conclusion
8.1 Discussion:
Communication is very important in this world. We have two types of communication such as
wired communication and wireless communication. Wireless communication is divided into
mobile communications and fixed wireless communications. Each type of communication has
huge demand according to customers need in the market. Wireless communication have
passing a revolutionary era to transmit data and information to remote area which we would
not even imagine. Wireless data transmission gives us every opportunity to get all feasible
necessary access to the world wherever we are and wherever we need from. It has gained
exponential growth of subscribers during last 10 to 12 years and continues to expand
everyday with new technology invention.
Though there are many performance degradation factors in wireless communication channels
but fading problem is the major impairment problem. Fading means the loss of propagation
experienced by the radio signal on forward and reverse links. The signals which is received by
mobile terminals come from several propagation paths those are called multi-path
propagation. Reflection, diffraction and scattering of radio waves in natural and human made
structures are the main causes to introduce multipath in propagation. The received signal at
the receiver suffers magnitude and phase variation due to multiple paths of propagation.
Those interfere themselves constructively and destructively. This variation of received signal
is called multipath fading.
To improve the performance of those fading channels, diversity techniques are used. In
diversity technique, the receiver is supplied multiple replicas of transmitting signals instead of
one signal that have passed over different fading channels. As a result, the probability that all
replicas of signals will fade simultaneously is reduced considerably.
For getting full benefit, uncorrelated faded signals are collected from diversity branches and
combine in such manner that can improve the performance of communication systems. This is
called diversity combining method and also used to optimize received signal power or signal-
to-noise ratio (SNR). The main idea of diversity combining is to combine several copies of
transmitted signals which were gone through independent fading in order to increase overall
received SNR. This combining method can be used in receiver mainly.
70
The performance improvement of a diversity system means its ability to reduce the
permissible values of carrier to noise ratio (CNR) and carrier to interference ratio (CIR) those
are required to obtain specified probability of error ( . The performance improvement
also can be evaluated in terms of statistical reduction of fading dynamic range of the average
CNR and CIR.
8.2 Conclusion:
This paper describes the characteristics of the signal propagation and performance
degradation issues in wireless communication channels. There are many impairment factors
and fading is the major factor amongst them which degrades the performance of wireless
communication system severally. Several types of diversity techniques are described which
can be use to mitigate fading problem in the channels. Diversity is used to supply the receiver
with several replicas of the same signal. Diversity combining techniques are used to improve
the channels’ performance without transmitting any additional power. If high received signal
are de-correlated, the gain of diversity is also high. The performance of diversity system is
expressed by its ability to reduce the value of carrier to noise ratio (CNR) and carrier to
interference ratio (CIR) at a specified probability of error ( . The performance evaluated in
terms of statistical reduction of the fading dynamic range of the average CNR and CIR. We
have concluded that the fading dynamic range of the instantaneous CNR and instantaneous
CIR is reduced remarkably by the use of diversity techniques. We have also seen there is
only a slight difference between the combining methods. For a certain average of Pe, a higher
amount of CNR and CIR is required for non-diversity systems, while a smaller amount of
CNR and CIR is required for a diversity technique. We see that maximal-ratio combining
(MRC) achieves the best performance improvement comparing to the other methods.
References:
[1] Dr. Kamilo Feher, Wireless Digital Communication-Modulation and Spread spectrum
Application. Prentice Hall of India, New Delhi, 2001.
[2] Rappaport, Theodore S., Wireless Communications, Principles and practice, prentice hall
PTR, 1996.
[3] Saleh Faruque, Cellular Mobile Systems Engineering, Artech House, Boston, London,
1996.
[6] Carl B. Dietrich, Kai Dietze, J. Raandall Nealy, Warren L. Stutzman, ‘Spatial,
Polarization and Pattern Diversity for Wireless Handheld Terminals’, IEEE Transaction on
Antennas and Propagation, Vol 49, No 9, September 2001.
[7] P.L. Perini and C.L. Holloway, ‘Angle and Space Diversity Comparisons in different
Mobile Radio environments’, IEEE transaction on Antenna and Propagation, Vol 46, June
1998.
[9] Xexian Li and Matti Latva-aho, ‘ Spatial Transmit diversity Techniques for MC-CDMA
systems’, Centre for Wireless Communication, University of Oulu, Finland, Sept. 2002.
[10] J.H Winters, ‘The impact of Antenna Diversity on the Capacity of Wireless
Communication System,’ IEEE Transaction of Communications, Vol 42, No. 2/3/4, April
1994.
[11] William C.Y Lee, ‘Mobile Cellular Telecommunications’, McGraw-Hill Book Company,
1982.
[12] http://en.wikipedia.org/wiki/GSM
[14] http://www.itfacts.biz/402-mln-3g-subscribers-worldwide/7775).
[15] http://findarticles.com/p/articles/mi_m0EIN/is_2008_Feb_28/ai_n24354710/)
72
[16] Suhas N. Diggavi, Naofal Al-Dhahir, A. Stamoulis and A.R. Caldarbank. '' Great
Expectations: The value of Spatial Diversity in Wireless Networks''.
[17] http://www.freepatentsonline.com/7206583.html
[18] Asha Malhotra. 'Cellular Radio Performance Engineering', Prentice Hall of India, 2004.
The End
73
74