Equalization and Diversity

Outline
• Introduction
• Fundamentals of Equalization
• Survey of Equalization Techniques
• Linear Equalizers
• Nonlinear Equalization
• Algorithms for Adaptive Equalization
• Fundamentals of diversity
• Survey of Diversity Techniques
• Frequency/Time/Space/Polarization Diversity
• Selection/MRC/EGC Combining
• RAKE Receiver
• Interleaving
 Introduction
• The properties of mobile radio channels:
• Multipath fading -> time dispersion, ISI
• Doppler spread -> dynamical fluctuation
These effects have a strong negative impact on the bit error rate of any
modulation.
• Mobile communication systems require signal processing
techniques that improve the link performance in hostile
mobile radio environments.
• Three popular techniques:
• Equalization: compensates for ISI
• Diversity: compensates for channel fading
• Channel coding: detects or corrects errors
• These techniques can be deployed independently or jointly.
Transmitted signal: s(t)
Channel model: h(t) =   k  (t − k )
k

Received signal: r(t) = s(t)  h(t) =   k s(t − k )

k
Ts Ts

1
3
2 4
(1) Equalization

• If the modulation bandwidth exceeds the coherence bandwidth

of the radio channel, ISI occurs and modulation pulses are
spread in time.

• Equalization compensates for intersymbol interference (ISI)

created by multipath within time dispersive channels.
• An equalizer within a receiver compensates for the average
range of expected channel amplitude and delay characteristics.

• Equalizers must be adaptive

• since the channel is generally unknown and time varying.
(2) Diversity

• Usually employed to reduce the depth and duration of the

fades experienced by a receiver in a flat fading (narrowband)
channel.
• Without increasing the transmitted power or bandwidth.
• Can be employed at both base station and mobile receivers.
• Types of diversity:
• .antenna polarization diversity
• .frequency diversity
• .time diversity.
• For example, CDMA systems often use a RAKE receiver,
which provides link improvement through time diversity
• Spatial diversity is the most common one.
• While one antenna sees a signal null, one of the other
antennas may see a signal peak.
(3) Channel Coding
• Used to Improve mobile communication link performance by
adding redundant data bits in the transmitted message.
• At the baseband portion of the transmitter, a channel coder
maps a digital message sequence into another specific
code sequence containing a greater number of bits than
originally contained in the message.

• The coded message is then modulated for transmission in the

wireless channel.

• Coding can be considered to be a post detection technique.

• Because decoding is performed after the demodulation portion

• two general types of channel codes:

• block codes
• convolutional
codes.
 Channel coding is generally treated independently from the
type of modulation used
but this has changed recently with the use of trellis coded
modulation schemes that combine coding and modulation to
achieve large coding gains without any bandwidth expansion.

Notes
• The three techniques of equalization, diversity, and
channel coding are used to improve radio link performance
(i.e. to minimize the instantaneous bit error rate)
• but the approach, cost, complexity, and effectiveness of each
technique varies widely in practical wireless communication
systems.
 Fundamentals of Equalization

• Intersymbol interference (ISI)

• Caused by multipath propagation (time dispersion) ;
• Cause bit errors at the receiver;
• The major obstacle to high speed data transmission
over mobile radio channels.

• Equalization
• a technique used to combat ISI;
• can be any signal processing operation that minimizes ISI;
• usually track the varying channel adaptively.
Operating modes of an adaptive equalizer
• Training (first stage)
• A known fixed-length training sequence is sent by the
transmitter so that the receiver's equalizer may average to a
proper setting.
• The training sequence is designed to permit an equalizer
at the receiver to acquire the proper filter coefficients in the
worst possible channel conditions
• The training sequence is typically a pseudorandom binary
signal or a fixed, prescribed bit pattern.
• Immediately following the training sequence, the user data
is sent.
• The time span over which an equalizer converges is
a function of
• the equalizer algorithm
• the equalizer structure
• the time rate of change of the multipath radio channel.
• Equalizers require periodic retraining in order to maintain
effective ISI cancellation.
Operating modes of an adaptive equalizer

• Tracking (second stage)

• Immediately following the training sequence, the user data is
sent.

• As user data are received, the adaptive algorithm of the

equalizer tracks the changing channel and adjusts its filter
characteristics over time.

• commonly used in digital communication systems

• where user data is segmented into short time blocks.

• TDMA wireless systems are particularly well suited

for equalizers.
• data in fixed-length time blocks,
• training sequence usually sent at the beginning of a block
Communication system with an adaptive equalizer

• Equalizer can be implemented at baseband or at IF in a receiver.

• Since the baseband complex envelope expression can be
used to represent bandpass waveforms and, thus, the channel
response, demodulated signal, and adaptive equalizer
algorithms are usually simulated and implemented at baseband

• Block diagram of a simplified communications system using

an adaptive equalizer at the receiver is shown in next page
Communication system with an adaptive equalizer

Modulator Transmitter Radio

x(t) Channel RF Front End

Detector IF Stage
nb (t) + Matched Filter
f (t)
y(t)
Adaptive Decision
heq (t) Equalizer Maker d (t)
d̂ (t)
Σ
e(t)
Relevant equations

y(t) = x(t)  f (t) + nb (t)

d̂(t) = x(t)  f (t)  h (t) + n (t)  h (t)
eq b eq

heq (t) =  ck  (t − nTs )

k

To eliminate ISI, we must have

heq (t)  f (t) =  (t) H eq ( f ) =

1
F( f )
an equalizer is an inverse filter of the channel.
In frequency selective channel, enhances the frequency components
with small amplitudes, attenuates the strong frequencies therefore
provide a flat, composite, received frequency response and linear
phase response.
A Generic Adaptive Equalizer

yk yk-1 yk-2
Z-1 Z-1 Z-1 Z-1

w0 w1 w2 wN
d̂ k
Adaptive algorithm that updates the weights Σ
ek
Σ
Prior knowledge: d k
A Generic Adaptive Equalizer

• A transversal filter with

• N delay elements
• N+1 taps
• N+1 tunable complex multipliers
• N+1 weights:
• These weights are updated continuously by the
adaptive algorithm
• either on a sample by sample basis or on a block by block
basis.
• The adaptive algorithm is controlled by the error signal ek.
• ek is derived by comparing the output of the equalizer with
some signal which is either an exact scaled replica of the
transmitted signal xk or which represents a known property of
the transmitted signal.
 A Generic Adaptive Equalizer
• A cost function is used
• the cost function is minimized by using ekThe, and the weights
are updated iteratively.
• For example, The least mean squares (LMS) algorithm
can serve as a cost function.
• Iterative operation based on LMS
• New weights = Previous weights + (constant) x (Previous error) x (Current input
vector)
• Where
• Previous error = Previous desired output — Previous actual output
• This process is repeated rapidly in a programming loop while the
equalizer attempts to converge
• Upon reaching convergence, the adaptive algorithm freezes the
filter weights until the error signal exceeds an acceptable level
or until a new training sequence is sent.
 A Generic Adaptive Equalizer
• Techniques used to minimize the error
• gradient
• steepest decent algorithms

• Based on classical equalization theory, the most common

cost function is MSE
• MSE----mean square error (MSE) between the desired signal
and the output of the equalizer
Denoted by E[e(k)  e (k)]
*
 A Generic Adaptive Equalizer
Blind algorithms
• more recent class of adaptive algorithms
• able to exploit characteristics of the transmitted signal
and do not require training sequences.
• provide equalizer convergence without burdening the
transmitter with training overhead
• able to acquire equalization through property restoral
techniques of the transmitted signal,
• Two techniques:
• the constant modulus algorithm (CMA)
• used for constant envelope modulation
• forces the equalizer weights to maintain a constant envelope
on the received signal
• spectral coherence restoral algorithm (SCORE).
• exploits spectral redundancy or cyclostationarity in
the transmitted signal
Equalizers in a Communications
Receiver
• Because noise is present, an equalizer is unable to
achieve perfect performance.
• Therefore, the instantaneous combined frequency
response will not always be flat, resulting in some finite
prediction error.
2
• The mean squared error (MSE) E [ek ] is one of the
most important measures of how well an equalizer
works.
• Minimizing MSE E [ek2] tends to reduce the bit error rate.
• For wireless communication links, it would be best to
minimize the instantaneous probability of error instead of MSE
• generally results in nonlinear
equations much more difficult to
solve in real-time
Survey of Equalization Techniques
• Equalization techniques can be subdivided into two
general categories:
• linear equalization
• The output of the decision maker is not used in the
feedback path to adapt the equalizer.
• nonlinear equalization
• The output of the decision maker is used in the
feedback path to adapt the equalizer.

• Many filter structures are used to implement linear

and nonlinear equalizers

• For each structure, there are numerous algorithms

used to adapt the equalizer.
 Classification of equalizers

Equalizer

Linear Nonlinear

Types
DFE ML Symbol MLSE
Detector

Transversal Lattice Transversal

Structures Transversal Lattice Channel Est.

Zero forcing Gradient RLS LMS Gradient RLS LMS

LMS RLS RLS
Algorithms RLS Fast RLS Fast RLS
Fast RLS Sq. root RLS Sq. root RLS
Sq. root RLS
Most common structure:
---- Linear transversal equalizer (LTE)
• made up of tapped delay lines, with the tappings
spaced a symbol period (Ts) apart
• the transfer function can be written as a function of the delay
operator − jTs or Z −1
Assuming that the delay elements have unity gain and delay Ts,
of a linear

Basic linear transversal equalizer structure

Most common structure:
---- Linear transversal equalizer (LTE)

Two types of LTE

• finite impulse response (FIR) filter

• The simplest LTE uses only feedforwZ ar−1d taps
• Transfer function is a polynomial in
• has many zeroes but poles only at z = 0
• Usually simply called a transversal filter

• Infinite impulse response (IIR) filter

• has both feedforward and feedback taps
• transfer function is a rational function of Z-1
with poles and zeros.
• tend to be unstable when used in channels
where the strongest pulse arrives after an echo pulse
(i.e., leading echoes)
• rarely used.
Tapped delay line filter with both feedforward and feedback taps (IIR)
 Linear Equalizers
Transversal filter implementation (LTE)

Input

Output

This type of equalizer is the simplest. Threshold Detector

 Linear Equalizers
• current and past values of the received signal are linearly
weighted by the filter coefficient and summed to produce the
output,
• If the delays and the tap gains are analog, the continuous output
of the equalizer is sampled at the symbol rate and the samples
are applied to the decision device.
• Implementation is usually carried out in the digital domain where
the samples of the received signal are stored in a shift register.

• The output before decision making (threshold detection)

The minimum MSE it can achieve

 Linear Equalizers
Lattice filter implementation

Numerical stable, faster convergence, Complicated

 Linear Equalizers
• Two main advantages of the lattice equalizer
• numerical stability
• faster convergence

• Unique structure allows dynamic assignment of the

most effective length
• When channel is not very time
dispersive Only a fraction of the stages
are used.
• channel becomes more time dispersive
• Length can be increased without stopping the operation

• Drawback: more complicated than LTE

 Nonlinear Equalization
• Linear equalizers do not perform well on channels which have
deep spectral nulls in the passband.
• In an attempt to compensate for the distortion, the linear
equalizer places too much gain in the vicinity of the spectral
null, thereby enhancing the noise present in those
frequencies.

• Nonlinear equalizers are used in applications where

the channel distortion is too severe for a linear
equalizer to handle.

• Three very effective nonlinear equalizer

• Decision Feedback Equalization (DFE)
• Maximum Likelihood Symbol Detection
• Maximum Likelihood Sequence Estimation (MLSE)
 Decision Feedback Equalization (DFE)

Basic idea:
• once an information symbol has been detected, the ISI
that it induces on future symbols can be estimated
and subtracted out before detection of subsequent
symbols.

• DFE Can be realized in either the direct transversal

form or as a lattice filter.

• The LTE form consists of a feedforward filter (FFF)

and a feedback filter (FBF).
• The FBF is driven by decisions on the output of the detector,
and its coefficients can be adjusted to cancel the ISI on the
current symbol from past detected symbols.

• The equalizer has N1 + N2 + I taps in FFF and N3 taps in FBF

Decision Feedback Equalization (DFE)

Input

Output

Feedforward Filter

Feedback Filter
Decision Feedback Equalization (DFE)

The output of DFE

The minimum mean square error of DFE

•It can be seen that the minimum MSE for a DFE is always
smaller than that of an LTE
Unless F (e jT ) is a constant, where adaptive equalization is not
needed
• If there are nulls in the F (e jT ) , a DFE has significantly
smaller minimum MSE than an LTE.
 Decision Feedback Equalization (DFE)

Conclusion

• an LTE is well behaved when the channel

spectrum is comparatively flat
• a DFE is more appropriate for severely distorted
wireless channels.

• If the channel is severely distorted or exhibits nulls

in the spectrum
• the performance of an LTE deteriorates and the mean
squared error of a DFE is much better than a LTE.
• Also, an LTE has difficulty equalizing a nonminimum
phase channel
• where the strongest energy arrives after the first arriving signal
component.
Another form of DFE----predictive DFE
• also consists of a feed forward filter (FFF) as
in the conventional DFE.

• Difference: the feedback filter (FBF) is

driven by an input sequence formed by the difference of
the output of the detector and the output of the feed
forward filter.
• the FBF here is called a noise predictor because it predicts
the noise and the residual ISI contained in the signal at the
FFF output and subtracts from it

• The predictive DFE performs as well as the

conventional DFE as the limit in the number of taps in
the FFF and the FBF approach infinity.

• The FEF in the predictive DFE can also be realized

as a lattice structure
Another form of DFE----predictive DFE
 Maximum Likelihood Sequence
Estimation (MLSE) equalizer
• The MSE-based linear equalizers are optimum with respect to the
criterion of minimum probability of symbol error when the channel
does not introduce any amplitude distortion.
• Yet this is precisely the condition in which an equalizer is needed
for a mobile communications link.

• MLSE uses various forms of the classical maximum

likelihood receiver structure.

• the MLSE tests all possible data sequences (rather than

decoding each received symbol by itself), and chooses the data
sequence with the maximum probability as the output.
• A channel impulse response simulator is used within the
algorithm,

• Drawback: An MLSE usually has a large

computational requirement
• especially when the delay spread of the channel is large.
Maximum Likelihood Sequence
Estimation (MLSE) equalizer

Matched
Filter

Delay

Channel
Estimator
 Maximum Likelihood Sequence
Estimation (MLSE) equalizer
• The MLSE can be viewed as a problem in estimating the
state of a discrete time finite state machine
• The channel has ML states, where M is the size of the symbol
alphabet of the modulation.

• An M L trellis is used by the receiver to model the channel

over time.
• The Viterbi algorithm then tracks the state of the channel by the
paths through the trellis.

• The MLSE is optimal in the sense that it

minimizes the probability of a sequence error.
 Maximum Likelihood Sequence
Estimation (MLSE) equalizer
NOTES:

• The MLSE requires knowledge of the channel characteristics

• in order to compute the metrics for making decisions.

• The MLSE also requires knowledge of the statistical

distribution of the noise corrupting the signal
• the probability distribution of the noise determines the form of the
metric for optimum demodulation of the received signal.

• The matched filter operates on the continuous time

signal, whereas the MLSE and channel estimator rely on
discretized (nonlinear) samples.
 Algorithms for Adaptive Equalization
• Equalizer requires a specific algorithm to update the
coefficients and track the channel variations.
• Since it compensates for an unknown and time-varying channel

• This section outlines three of the basic algorithms for

adaptive equalization.
• Though the algorithms detailed in this section are derived for the
linear, transversal equalizer, they can be extended to other
equalizer structures, including nonlinear equalizers.
 Algorithms for Adaptive
Equalization
Factors determining the performance of an algorithm:

• Rate of convergence (fast or slow?)

• Defined as the number of iterations required for the
algorithm, in response to stationary inputs, to converge close
enough to the optimum solution.
• A fast rate of convergence allows the algorithm to adapt
rapidly to a stationary environment of unknown statistics.
• Furthermore, it enables the algorithm to track statistical
variations when operating in a nonstationary environment.

• Misadjustment (precise or not?)

• Provides a quantitative measure of the amount by which
the final value of the mean square error, averaged over an
ensemble of adaptive filters, deviates from the optimal
minimum mean square error.
 Algorithms for Adaptive
Equalization
Factors determining the performance of an algorithm:

• Computational complexity (simple or complex?)

• Number of operations required to make one
complete iteration of the algorithm.

• Numerical properties (stable or not?)

• When an algorithm is implemented numerically,
inaccuracies are produced due to round-off noise and
representation errors in the computer.
• These kinds of errors influence the stability of the algorithm.
 Algorithms for Adaptive
Equalization
Practical considerations for choice of an equalizer structure and
its algorithm

• The cost of the computing platform (affordable or not?)

• especially when used in user equipments

• The power budget (power limited applications or else?)

• In portable radio applications, battery drain at the subscriber unit
is a paramount consideration

• The radio propagation characteristics (fast fading &

time delay spread?)
• The speed of the mobile unit determines the channel fading rate
and the Doppler spread, which is directly related to the
coherence time of the channel
 Algorithms for Adaptive
Equalization
Three classic equalizer algorithms

• Zero Forcing Algorithm (ZF)

• Least Mean Square Algorithm (LMS)
• Recursive Least Squares Algorithm (RLS)

Please read references for detailed

information on a specific algorithm.
Zero Forcing (ZF) Algorithm
Criterion:
to force the samples of the combined channel and equalizer
impulse response to zero at all but one of sample points in the
tapped delay line filter.

Disadvantage:
may excessively amplify noise at frequencies where the
folded channel spectrum has high attenuation.

Suitability:
Wireline communications
1 1
H eq ( f ) = , f 
H ch ( f ) 2T
 Least Mean Square (LMS)
Algorithm
Criterion:

to minimize the mean square error (MSE) between the desired

equalizer output and the actual equalizer output.
Minimize  = E[ek  ek ]
*

Must be solved iteratively

Simplest algorithm, requires only 2N + I operations per iteration.

• The LMS equalizer maximizes the signal to distortion ratio

at its output within the constraints of the equalizer filter length.

a step size  is used to control the convergence rate and the

stability
 Least Mean Square (LMS)
Algorithm
Disadvantage: low convergence rate.
Because of the only one parameter 
Especially when the eigenvalues of the input covariance matrix
RNN have a very large spread, i.e, max / min  1

• If an input signal has a time dispersion characteristic that

is greater than the propagation delay through the equalizer,
then the equalizer will be unable to reduce distortion.

• To prevent the adaptation from becoming unstable, the value of 

• is chosen from

where i is the ith eigenvalue of the covariance matrix RNN.

The step size  can be controlled by the total input power in

order to avoid instability in the equalizer [Hay86].
since
 Recursive Least Squares (RLS)
Algorithm
• RLS is Proposed to improve the convergence rate of LMS algorithm.

• Error measures expressed in terms of a time average of the

actual received signal instead of a statistical average.

•  is the weighting coefficient that can change the performance of

the equalizer.
• If a channel is time-invariant,  can be set to 1. Usually 0.8-1.
• The value of  has no influence on the rate of convergence, but
does determine the tracking ability.
• The smaller the  , the better the tracking ability of the equalizer.
• However, if  is too small, the equalizer will be unstable
 Recursive Least Squares (RLS)
Algorithm
Advantage: high convergence rate

Disadvantage: sometimes unstable

The RLS algorithm described above, called the

Kalman RLS algorithm
Uses 2.5N2 + 4.5N arithmetic operations per
iteration.
7.8.4 Summary of equalization
algorithms
 There are number of variations of the LMS and RLS algorithms
 RLS algorithms have similar convergence and tracking
performances, which are much better than the LMS algorithm.
Usually have high computational requirement and complex
program structures.
Some RLS algorithms tend to be unstable.

About FTF
 Among the RLS algorithms, fast transversal filter (FTF)
algorithm requires the least computation
 a rescue variable can be used to avoid instability.
However, rescue techniques tend to be a bit tricky for widely
varying mobile radio channels.
FTF is not widely used.
Comparison of Various Algorithms for Adaptive Equalization
[Pro9l]
7.9 Fractionally Spaced
Equalizers(FSE)
 In the presence of channel distortion, the matched filter prior to
the equalizer must be matched to the channel and the corrupted
signal.
Usually get the suboptimal result because the channel response is
unknown.
This results in a significant degradation in performance.

 FSE is based on sampling the incoming signal at least as fast as

the Nyquist rate.

 The FSE compensates for the channel distortion before aliasing

effects occur due to the symbol rate sampling.
FSE incorporates the functions of a matched filter and equalizer
into a single filter structure.

 Simulation results demonstrate the effectiveness of the FSE over

a symbol rate equalizer.(see [Ste94])
7.10 Fundamentals of Diversity
Techniques
 Random nature of radio propagation:
 Multipath propagation
 Independent fading of each Multipath component
 If one radio path undergoes a deep fade, another
independent path may have a strong signal

 Diversity exploits the random nature of radio propagation by

finding independent signal paths for communication, so as to
boost the instantaneous SNR at the receiver.

Path 2

Transmitter Path 1 Receiver

Path 3
7.10 Fundamentals of Diversity
Techniques
 Diversity is a powerful communication receiver technique
that provides wireless link improvement at relatively low cost.

 Requires no training

 In virtually all applications, diversity decisions are made by the

receiver, and are unknown to the transmitter.

Two types of diversity

 Microscopic diversity small scale fading
 Macroscopic diversity large scale fading
7.10 Fundamentals of Diversity
Techniques
Microscopic diversity
 Small-scale fades: deep and rapid amplitude fluctuations over
distances of just a few wavelengths.
caused by multiple reflections from the surroundings in the vicinity
of the mobile.
results in a Rayleigh fading distribution of signal strength over small
distances.

 Microscopic diversity techniques can exploit the rapidly

changing signal.
For example, use two antennas at the receiver (separated by a
fraction of a meter), one may receive a null while the other receives a
strong signal.
By selecting the best signal at all times, a receiver can mitigate
small-scale fading effects
Called antenna diversity or spacediversity

 Samples: Rake receiver, MIMO transmission

7.10 Fundamentals of Diversity
Techniques
Macroscopic diversity

 Large-scale fading:caused by shadowing due to variations in

both the terrain profile and the nature of the surroundings.
In deeply shadowed conditions, the received signal strength at a
mobile can drop well below that of free space.
log-normally distributed with a standard deviation of about 10 dB in
urban environments.

 Macro-scope diversity: By selecting a base station which is not

shadowed when others are, the mobile can improve substantially
the average ratio on the forward link.
It is the mobile that takes advantage of large separations between
the serving base stations.
7.10 Fundamentals of Diversity
Techniques
Macroscopic diversity

 Macroscopic diversity is also useful at the base station receiver.

By using base station antennas that are sufficiently separated in
space, the base station is able to improve the reverse link by selecting
the antenna with the strongest signal from the mobile.

 Used to combat slow fading (shadowing)

 Samples: Base-station handoff in cellular networks

7.10 Fundamentals of Diversity
Techniques
Macro-scope diversity

Base station Base station

Mobile
7.10 Fundamentals of Diversity
Techniques
 Strategies used in diversity techniques
 Selection diversity
 Maximal ratio combining diversity
 Equal-gain combining diversity
 Hybrid schemes

 Practical considerations
 effectiveness, complexity, cost, and etc.
7.10.1 Derivation of Selection
Diversity improvement
 Consider M independent Rayleigh fading channels available
areceiver.
Each channel is called a diversity branch.
7.10.1 Derivation of Selection
Diversity improvement
 Further assume that each branch has the same average SNR
given by

Where we assume  2 = 1 .
 If each branch has an instantaneous SNR = i , then the pdf of  i
is

where  is the mean SNR of each branch.

 The probability that a single branch has SNR less than some
threshold y  is
7.10.1 Derivation of Selection
Diversity improvement
 Now, the probability that all M independent diversity branches
receive signals which are simultaneously less than some
specific SNR threshold is

This is the probability of all branches failing to achieve SNR = i .

 If a single branch achieves SNR   ,then the probability that
SNR   for one or more branches is given by

This is the probability of exceeding a threshold when selection

diversity is used.
7.10.1 Derivation of Selection
Diversity improvement
How to determine the average signal-to-noise ratio of the
received signal when diversity is used?

 First of all, find the pdf of  (the instantaneous SNR when M

branches are used).Thus we compute the derivation of CDF
PM () ,

 Then, we can compute the average SNR,  ,

where x =  /  .
The above equation can be evaluated to yield the average
SNR improvement offered by selection diversity.
 M
1
=
 k
7.10.1 Derivation of Selection
Diversity improvement
 Selection diversity offers an average improvement in the
link margin without requiring additional transmitter power
or sophisticated receiver circuitry.
The diversity improvement can be directly related to the average
bit error rate for various modulations.

 Selection diversity is easy to implement because all that is

needed is a side monitoring station and an antenna switch at
the receiver.

 However, it is not an optimal diversity technique because it

does not use all of the possible branchessimultaneously.
Maximal ratio combining uses each of the M branches in a co-
phased and weighted manner such that the highest achievable SNR
is available at the receiver at all times.
7.10.1 Derivation of Selection
Diversity improvement
Example
Assume four branch diversity is used, where each branch
receives an independent Rayleigh fading signal. If the average
SNR is 20 dB, determine the probability that the SNR will drop
below 10 dB. Compare this with the case of a single receiver
without diversity.
Solution
7.10.2 Derivation of Maximal Ratio
Combining Improvement
In maximal ratio combining, the voltage signals ri from each of
the M diversity branches are co-phased to provide coherent voltage
addition and are individually weighted to provide optimal SNR.
7.10.2 Derivation of Maximal Ratio
Combining Improvement
1) The SNR out of the diversity combiner:
 If each branch has gain Gi , then the resulting signal envelope
applied to the detector is

 Assuming that each branch has the same average noise power
N, the total noise power NT applied to the detector is simply the
weighted sum of the noise in each branch. Thus

which results in an SNR applied to the detector, M , given by

7.10.2 Derivation of Maximal Ratio
Combining Improvement
 Using Chebychev's inequality, M is maximized
when Gi = ri / N , which leads to

(7-66)

 Conclusion:
The SNR out of the diversity combiner is simply the sum of
the SNRs in each branch.
7.10.2 Derivation of Maximal Ratio
Combining Improvement
2) The pdf of M

 According to Chapter 3, M is a Chi-square distribution of 2M

Gaussian random variables. Thus,the pdf for  M is

(7-68)

3) The CDF of  M

 According to the abovementioned pdf, The probability thMat

is less than some SNR threshold is
7.10.2 Derivation of Maximal Ratio
Combining Improvement
4) The average SNR out of the diversity combinerM,

 M can be calculated by using the pdf of M (Eq. (7.68)). But the

direct way is to calculate it from Eq. (7-66).

 That is to say, the average SNR,  M , is simply the sum of the

individual  ifrom each branch.

The control algorithms for setting the gains and phases for
maximal ratio combining receivers are similar to those required in
equalizers and RAKE receivers.
Maximal ratio combining can be applied to virtually any diversity
application, although often at much greater cost and complexity than
other diversity techniques.
7.10.3 Practical Space Diversity
Considerations
 Space diversity (also known as antenna diversity), is one of the
most popular forms of diversity used in wireless systems.

 The signals received from spatially separated antennas on the

mobile would have essentially uncorrelated envelopes for
antenna separations of one half wavelength or more.

 Space diversity can be used at either the mobile or base station,

or both.

Since the important scatterers are generally on the ground in the

vicinity of the mobile, when base station diversity is used, the
antennas must be spaced considerably far apart to achieve
decorrelation (several tens of wavelengths).
7.10.3 Practical Space Diversity
Considerations
general block diagram of a space diversity scheme
7.10.3 Practical Space Diversity
Considerations
Space diversity reception methods can be classified into four
categories

 1. Selection diversity
 2. Feedback diversity
 3. Maximal ratio combining
 4. Equal gain diversity
7.10.3 Practical Space Diversity
Considerations
(1) Selection Diversity

 The simplest diversity technique.

 The receiver branch having the highest instantaneous SNR is
connected to the demodulator.
 The antenna signals themselves could be sampled and the best
one sent to a single demodulator.
 In practice, the branch with the largest (S + N) /N is used, since
it is difficult to measure SNR.
 A practical selection diversity system cannot function on a
truly instantaneous basis, but must be designed so that the
internal time constants of the selection circuitry are shorter
than the reciprocal of the signal fading rate.
7.10.3 Practical Space Diversity
Considerations
(2) Feedback or Scanning Diversity
 Very similar to selection diversity
 The M signals are scanned in a fixed sequence until one is
found to be above a predetermined threshold.
 This signal is then received until it falls below threshold and
the scanning process is again initiated.
 The resulting fading
statistics are
somewhat inferior to
those obtained by
the other methods.
 Advantage: very
simple to implement
(only one receiver is
required).
7.10.3 Practical Space Diversity
Considerations
(3) Maximal Ratio Combining
 The signals from all of the M branches are weighted and then
summed.
 The individual signals must be co-phased before being
summed.
requires an individual receiver and phasing circuit for each
antenna element.
 Output SNR equal to the sum of the individual SNRs.
 Advantage: produces an output with an acceptable SNR even
when none of the individual signals are themselves acceptable.
 Gives the best statistical reduction of fading of any known
linear diversity combiner.
7.10.3 Practical Space Diversity
Considerations
Maximal Ratio Combiner
7.10.3 Practical Space Diversity
Considerations
(4) Equal Gain Combining

In certain cases, it is not convenient to provide for the variable

weighting capability required for true maximal ratio combining. In
such cases, the branch

 Equal gain combining diversity sets all weights to unity but the
signals from each branch are co-phased.
 The possibility of producing an acceptable signal from a
number of unacceptable inputs is still retained,
 The performance is only marginally inferior to maximal ratio
combining and superior to selection diversity.
 7.10. 4 PolarIzation Diversity
At the base station, space diversity is considerably less practical .

 polarization diversity only provides two diversity branches, but

allows the antenna elements to be co-located.
 Measured horizontal and vertical polarization paths between a
mobile and a base station are reported to be uncorrelated.
 Decorrelation for the signals in each polarization is caused by
multiple reflections.
 The reflection coefficient for each polarization is different,
which results in different amplitudes and phases for each, or at
least some, of the reflections.
 After sufficient random reflections, the polarization state of the
signal will be independent of the transmitted polarization.
In practice, however, there is some dependence of the received
polarization on the transmitted polarization.
7.10.5 Frequency Diversity
 Transmits information on more than one carrier frequency.
frequencies separated by more than the coherence bandwidth of
the channel will not experience the same fades.

 Frequency diversity is often employed in microwave LOS links.

 In practice, 1:N protection switching is provided by a radio licensee,

When diversity is needed, the appropriate traffic is simply switched
to the backup frequency.

 Disadvantage: not only requires spare bandwidth but also

requires that there be as many receivers as there are channels
used for the frequency diversity.
for critical traffic, the expense may be justified.
7.10.5 Frequency Diversity
 New OFDM modulation and access techniques exploit
frequency diversity by providing simultaneous modulation
signals with error control coding across a large bandwidth.

 If a particular frequency undergoes a fade, the composite

signal will still be demodulated.
7.10.6 Time Diversity
 Time diversity repeatedly transmits information at time
spacings that exceed the coherence time of the channel
Multiple repetitions of the signal will be received with independent
fading conditions.

 One modem implementation of time diversity involves the use

of the RAKE receiver for spread spectrum CDMA, where the
multipath channel provides redundancy in the transmitted
message.
7.11 RAKE Receiver
 In CDMA spread spectrum systems, the spreading codes are
designed to provide very low correlation between successive
chips.

 If the multipath components are delayed in time by more than a

chip duration, they appear like uncorrelated noise at a CDMA
receiver, and equalization is not required.

 However, since there is useful information in the multipath

components, CDMA receivers may combine the time delayed
versions of the original signal transmission in order to improve
the signal to noise ratio at the receiver

 A RAKE is employed to do this:

It attempts to collect the time-shifted versions of the original
signal by providing a separate correlation receiver for each of
the multipath signals.
7.11 RAKE Receiver
The RAKE receiver is essentially a diversity receiver
designed specifically for CDMA, where the diversity is provided
by the fact that the multipath components are practically
uncorrelated from one another when their relative propagation
delays exceed a chip period.

An M branch (M-finger) RAKE receiver implementation. Each correlator detects a time shifted
version of the original CDMA transmission, and each finger of the RAKE correlates to a portion
of the signal which is delayed by at least one chip in time from the other fingers.
7.12 Interleaving
 Interleaving is used to obtain time diversity in a digital
communications system without adding any overhead.
useful technique in all second and third generation digital cellular
systems.

 It is typical for many speech coders to produce several

"important" bits in succession.

 Interleaver is employed to spread these bits out in time

so that if there is a deep fade or noise burst, the important bits
from a block of source data are not corrupted at the same time.

 Then the error control coding (called channel coding)can be

used to correct these errors.
usually error control coding can deal with random errors.
7.12 Interleaving
Two types of interleaver:
1) Block structure
 Formats the encoded data into a rectangular array of m rows
and n columns, and interleaves nm bits at a time.
 Usually, each row contains a word of source data having n bits.
An interleaver of degree m (or depth m) consists of m rows.
7.12 Interleaving

 Source bits are placed into the interleaver by sequentially

increasing the row number for each successive bit, and filling
the columns.

 The interleaved source data is then read out row-wise and

transmitted over the channel.
This has the effect of separating the original source bits by m bit
periods.

 At the receiver, the de-interleaver stores the received data by

sequentially increasing the row number of each successive bit,
and then clocks out the data row-wise, one word (row) at a time.
7.12 Interleaving
Delay introduced by interleaving

 There is an inherent delay associated with an interleaver .

since the received message block cannot be fully decoded until
all of the nm bits arrive at the receiver and are de-interleaved.

 Human speech is tolerable to listen to until delays of greater

than 40 ms occur.
all of the wireless data interleavers have delays which do not
exceed 40ms.

 The interleaver word size and depth are closely related to

(a) the type of speech coder used
(b) the source coding rate
(c) the maximum tolerable delay
7.12 Interleaving
2) Convolutional structure

 Can be used in place of block interleavers in much the same

fashion.
 Ideally suited for use with convolutional codes.
7.13 Fundamental of
Channel Coding
 Channel coding protects digital data from errors by selectively
introducing redundancies in the transmitted data.

 Two types of Channel codes

1) error detection codes
2) error correction codes.

 The basic purpose of Channel Coding:

Introduce redundancies in the data to improve wireless link
performance.

 Cost: Increases the bandwidth requirement for a fixed source

data rate.
This reduces the bandwidth efficiency of the link in high SNR
conditions.
But provides excellent BER performance at low SNR values.
7.13 Fundamental of Channel
Coding

Note:

Channel Coding will be detailed in another course.