3/4/2018

EE140 Introduction to
Communication Systems
Lecture 15

Instructor: Xiliang Luo

ShanghaiTech University, Spring 2018

1

Architecture of a (Digital) Communication System

Transmitter

A/D Source Channel

Source Modulator
converter encoder encoder

Absent if
source is Noise Channel
digital

D/A Source Channel

User Detector
converter decoder decoder

Receiver

1
 3/4/2018

Contents

• Review: Signal design with zero ISI

• Tx/Rx filter design with channel response

• Equalization

Signal Design for Bandlimited Channel

Zero ISI
• Nyquist condition for Zero ISI for pulse shape
1 0
0 0
or ∑ T
• With the above condition, the receiver output
simplifies to

2
 3/4/2018

Nyquist Condition: Ideal Case

• Nyquist’s first method for eliminating ISI is to use
1 | | /
/
0

• = Nyquist bandwidth
• The minimum transmission bandwidth for zero ISI.
A channel with bandwidth can support a max.
transmission rate of 2 symbols/sec
5

Achieving Nyquist Condition

• Challenges of designing such or
– ( ) is physically unrealizable due to the abrupt transitions
at
– decays slowly for large t, resulting in little margin of
error in sampling times in the receiver
– This demands accurate sample point timing - a major
challenge in modem / data receiver design
– Inaccuracy in symbol timing is referred to as timing jitter

3
 3/4/2018

Practical Solution: Raised Cosine Spectrum

• is made up of 3 parts: passband, stopband,
and transition band. The transition band is shaped
like a cosine wave.

Time-Domain Pulse Shape

• Taking inverse Fourier transform

4
 3/4/2018

Contents

• Review: Signal design with zero ISI

• Tx/Rx filter design with channel response

• Equalization

Optimum Transmit/Receive Filter

• Recall that when zero-ISI condition is satisfied by
with raised cosine spectrum , then the
sampled output of the receiver filter is
(assume (0)=1)
• Consider binary PAM transmission:
• Variance of noise
with ∗ ∗


Error Probability can be minimized through a proper choice

of H f and H f so that / is maximum (assuming H f
fixed and P(f) given)

10

5
 3/4/2018

Optimal Solution
• Compensate the channel distortion equally between
the transmitter and receiver filters

/
for

• Then, the transmit signal energy is given by

By Parseval’s theorem

• Hence

11

Optimal Solution (cont’d)

• Noise variance at the output of the receive filter is

2 2

 ,

Performance loss due to channel distortion

• Special case: 1 for |f|
– This is the ideal case with “flat” fading
– No loss, same as the matched filter receiver for AWGN
channel

12

6
 3/4/2018

Exercise
• Determine the optimum transmitting and receiving
filters for a binary communications system that
transmits data at a rate R=1/T = 4800 bps over a
channel with a frequency response ;

|f| ≤W where W= 4800 Hz

• The additive noise is zero-mean white Gaussian

with spectral density ⁄2 10 Watt/Hz

• Determine the required transmit energy to achieve

10

13

Solution
• Since W = 1/T = 4800, we use a signal pulse with a
raised cosine spectrum and a roll-off factor = 1
• Thus,
1 | |
1 cos | | cos
2 9600
• Therefore

cos 1 , for |f| 4800

9600 4800
• One can now use these filters to determine the
amount of transmit energy required to achieve a
specified error probability

14

7
 3/4/2018

Performance with ISI

• If zero-ISI condition is not met, then

• Let

• Then

15

Performance with ISI

• Often only 2 significant terms are considered.

• Finding the probability of error in this case is quite

difficult. Various approximation can be used
(Gaussian approximation, Chernoff bound, etc.)

• What is the solution?

16

8
 3/4/2018

Monte Carlo Simulation

17

Monte Carlo Simulation (cont’d)

• If one wants to be within 10% accuracy, how
many independent simulation runs do we need?

• If ~10 (this is typically the case for optical

communication systems), and assume each
simulation run takes 1 msec, how long will the
simulation take?

18

9
 3/4/2018

Contents

• Review: Signal design with zero ISI

• Tx/Rx filter design with channel response

• Equalization

19

What is Equalizer
• We have shown that
– By properly designing the transmitting and receiving filters,
one can guarantee zero ISI at sampling instants, thereby
minimizing .
– Appropriate when the channel is precisely known and its
characteristics do not change with time
– In practice, the channel is unknown or time-varying

• Channel equalizer: a receiving filter with adjustable

frequency response to minimize/eliminate inter-
symbol interference

20

10
 3/4/2018

Equalizer Configuration

• Overall frequency response

• Nyquist criterion for zero-ISI

• Ideal zero-ISI equalizer is an inverse channel filter

with
1
∝ | | 1/2
21

Linear Transversal Filter

• Finite impulse response (FIR) filter

• are the adjustable 2N+1 equalizer coefficients

• N is sufficiently large to span the length of ISI
22

11
 3/4/2018

Zero-Forcing Equalizer
• : received pulse from a channel to be equalized

1, 0
• ∑
0, 1, … ,
To suppose 2N adjacent interference terms

23

Zero-Forcing Equalizer (cont’d)

• Rearrange to matrix form

24

12
 3/4/2018

Example

25

Solution

26

13
 3/4/2018

Solution
• The inverse of this matrix (e.g by MATLAB)

• Therefore
0.117, 0.158, 0.937, 0.133, 0.091
• Equalized pulse response

• It can be verified that

0 1.0 0, 1, 2
27

Solution
• Note that values of for 2 or 2 are not
zero. For example:

28

14
 3/4/2018

Minimum Mean-Square Error Equalizer

• Drawback of ZF equalizer
– Ignores the additive noise
– May even amplify the noise power

• Alternatively,
– Relax zero ISI condition
– Minimize the combined power in the residual ISI and
additive noise at the output of the equalizer

• MMSE equalizer:
– a channel equalizer that is optimized based on the
minimum mean-square error (MMSE) criterion

29

MMSE Criterion

• The output is sampled at :

• Let A = desired equalizer output

≜ Minimum

30

15
 3/4/2018

MMSE Criterion (cont’d)

where
E is taken over
and the additive noise

• MMSE solution is obtained by 0

, 0, 1, … ,

31

MMSE Equalizer vs. ZF Equalizer

• Both can be obtained by solving similar equations

• ZF equalizer does not consider effects of noise

• MMSE equalizer designed so that mean-square error

(consisting of ISI terms and noise at the equalizer
output) is minimized

• Both equalizers are known as linear equalizers

32

16
 3/4/2018

Decision Feedback Equalizer (DFE)

• DFE is a nonlinear equalizer which attempts to
subtract from the current symbol to be detected the
ISI created by previously detected symbols

33

Examples of Channels with ISI

34

17
 3/4/2018

Frequency Response

Channel B tends to significantly enhance the noise when a

linear equalizer is used (since linear equalizers have to
introduce a large gain to compensate for channel null).

35

Performance of MMSE Equalizer

36

18
 3/4/2018

Performance of DFE

37

Maximum Likelihood Sequence Estimation

(MLSE)

• Let the transmitting filter have a square root raised cosine

frequency response
| |
0
• The receiving filter is matched to the transmitter filter with
| |
0
• The sampled output from receiving filter is

38

19
 3/4/2018

MLSE
• Assume ISI affects finite number of symbols, with
0
• Then, the channel is equivalent to an FIR discrete-
time filter

39

Performance of MLSE

40

20
 3/4/2018

More about Equalization

41

Thanks for your kind attention!

Questions?

42

21