LINEAR EQUALIZERS & NONLINEAR EQUALIZERS

Prepared by
Deepa.T, Asst.Prof. /TCE
 Equalizers

• The goal of equalizers is to

eliminate intersymbol
interference (ISI) and the additive
noise as much as possible.
• Intersymbol interference(ISI)
arises because of the spreading
of a transmitted pulse due to the
dispersive nature of the channel,
which results in overlap of
adjacent pulses.
• In Fig there is a four‐level pulse
amplitude modulated signal
(PAM), x(t). This signal is
transmitted through the channel
with impulse response h(t). Then
noise n(t) is added. The received
signal r(t) is a distorted signal.
 Classification of Equalization

• Equalization using
– MLSE (Maximum likelihood sequence
estimation)
– Filtering
• Transversal filtering
– Zero‐forcing equalizer
– Minimum mean square error (MSE) equalizer
• Decision feedback
– Using the past decisions to remove the ISI contributed by
them
• Adaptive equalizer

Lecture 6 3
 Categories of Equalization

• Equalizers are used to overcome the

negative effects of the channel. In
general, equalization is partitioned
into two broad categories;
1. Maximum likelihood sequence
estimation (MLSE) which entails
making measurement of channel
impulse response and then
providing a means for adjusting the
receiver to the transmission
environment.
(Example: Viterbi equalization)
2. Equalization with filters, uses filters
to compansate the distorted pulses.
The general channel and equalizer
pair is shown in Figure.2.
Depending on the time nature

• These type of equalizers can be grouped as preset or

adaptive equalizers.
• Preset equalizers assume that the channel is time invariant
and try to find H(f) and design equalizer depending on H(f).
The examples of these ADAPTIVE EQUALIZERS are zero
forcing equalizer, minimum mean square error equalizer,
and desicion feedback equalizer.
• Adaptive equalizers assume channel is time varying channel
and try to design equalizer filter whose filter coefficients
are varying in time according to the change of channel, and
try to eliminate ISI and additive noise at each time. The
implicit assumption of adaptive equalizers is that the
channel is varying slowly.
 Equalization

Step 1 – waveform to sample transformation Step 2 – decision making

Demodulate & Sample Detect

z (T ) Threshold m̂i
r (t ) Frequency Receiving Equalizing
comparison
down‐conversion filter filter

For bandpass signals Compensation for

channel induced ISI

Received waveform Baseband pulse

Baseband pulse Sample
(possibly distored)
(test statistic)

Lecture 6 6
 Equalization

• ISI due to filtering effect of the

communications channel (e.g. wireless
channels)
– Channels behave like band‐limited filters
jθ c ( f )
Hc ( f ) = Hc ( f ) e

Non‐constant amplitude Non‐linear phase

Amplitude distortion Phase distortion

Lecture 6 7
Equalization Techniques

Fig.3 Classification of equalizers

 Equalizer Techniques

•Linear transversal equalizer (LTE, made up of tapped delay

lines as shown in Fig.4)

Fig.4 Basic linear transversal equalizer structure

•Finite impulse response (FIR) filter (see Fig.5)

•Infinite impulse response (IIR) filter (see Fig.5)
 Equalization by transversal filtering

• Transversal filter:
– A weighted tap delayed line that reduces the effect of ISI
by properN
adjustment of the filter taps.
z (t ) = ∑ cn x(t − nτ ) n = − N ,..., N k = −2 N ,...,2 N
n=− N

x(t )
τ τ τ τ
c− N c− N +1 c N −1 cN

z (t )
∑
Coeff.
adjustment
Lecture 6 10
 Transversal equalizing filter …

• Zero‐forcing equalizer:
– The filter taps are adjusted such that the equalizer output is
forced to be zero at N sample points on each side:
Adjust ⎧1 k =0
z (k ) = ⎨
{cn }nN=− N ⎩0 k = ±1,...,± N

• Mean Square Error (MSE) equalizer:

– The filter taps are adjusted such that the MSE of ISI and noise
power at the equalizer output is minimized.
Adjust

{c n }
N [
min E ( z (kT ) − ak ) 2 ]
n=− N

Lecture 6 11
 Linear vs. Nonlinear Equalization Techniques

•Two general categories ‐ linear and nonlinear

equalization
Linear ‐No feedback path to adapt the equalizer, the
equalization is linear
Non linear ‐fed back to change the subsequent outputs
of the equalizer, the equalization is nonlinear
LINEAR EQUALIZERS
 Linear Equalizers
• Zero‐forcing
– Design E(z) so that ISI is totally removed.
• Minimum mean square error (MMSE)
– Design E(z) to minimize the mean square error (MSE)
MSE = ∑ ε = ∑ (ck − ck )
2 2
k
ˆ
k k

14
 Structure of a Linear
Transversal Equalizer [5]

N2
• dˆ k = ∑ C n* y k − n
n=− N1

[ ]
T πT
• E e(n) = ∫−π
2 No
2π T F(e ) + No
jωt 2
dω

F(e jωt ) :frequency response of the channel

No :noise spectral density

Structure of a Lattice Equalizer [6‐7]

Fig.7 The structure of a Lattice Equalizer

 Characteristics of Lattice Filter
•Advantages
¾Numerical stability
¾Faster convergence
¾Unique structure allows the dynamic assignment of the most
effective
length
•Disadvantages
¾The structure is more complicated
 Linear Equalizers
• System modeling
⎡ u (− K ) 0 0 ... 0 0 ⎤ ⎡e− K ⎤ ⎡cˆ− 2 K ⎤
K ⎢u (− K + 1) u (− K ) 0 ... ... ⎥⎢ M ⎥ ⎢ M ⎥
∑e u
...
ĉi = n i −n ⇒⎢ ⎥⎢ ⎥ = ⎢ ⎥
n=− K
⎢ ⎥⎢ M ⎥ ⎢ M ⎥
⎢ ⎥⎢ ⎥ ⎢ ⎥
⎣ 0 0 0 0 u ( K )⎦ ⎣ eK ⎦ ⎣ cˆ2 K ⎦

• Zero‐forcing solution
⎧1, k = 0
cˆk = ⎨
⎩0, k ≠ 0
• Example
– K = 1 and ui = {0, 0.2, 0.9, ‐0.3, 0.1}. Calculate {ei}.

18
 Linear Equalizers
• Zero‐forcing equalizer
– Noise enhancement

19
 Linear Equalizers
• MMSE

20
Linear Equalizers
• Minimum Mean Square Error (MMSE) Solution
ε i = ci − cˆi = ci − uTi e,
[ ] [( ) ] = E[c − 2c u e + u ee u ]
E ε i2 = E ci − uTi e
2 2
i i
T
i
T
i
H
i
*
i

Let : E [ε ] = 0 ⇒ −2 E [c u ]+ 2 E [u u ]e = 0
∂ 2 * * T

∂e
i i i i i

E [c u ] = u c = p, E [u u ] = u u = R
i
*
i
* * T
i i
* T

e = [uHu]‐1[uHc]
• Example
K = 3, ui = {0.0110, 0.0227, ‐0.1749, 1.0000, 0.1617, ‐0.0558, 0.0108}.
Calculate {ei} and then calculate ci i = ±0, ±1, ±2,…, ±6.
Col = [0.0110, 0.0227, ‐0.1749, 1.0000, 0.1617, ‐0.0558, 0.0108 zeros(1, 6)];
Row = [0.0110, zeros(1, 6)]; U = toeplitz(C, Row);
R = U'*U; P = U'*c; %c = [0 0 0 1 0 0 0]’;
e = [‐0.0116 0.0108 0.1659 0.9495 ‐0.1318 0.0670 ‐0.0269];21
 Linear Equalizers
• Algorithms
– To reduce the computation complexity of
−1
e=R p (~ (2K+1)3 operations!)

– Least‐mean‐square algorithm (LMS)

∇ˆ = −2pˆ + 2R
n n
ˆ e
n n

e = e − μ∇
n +1
ˆ
n n
• μ determines convergence and residual errors
– Recursive‐least‐squares algorithm (RLS)
R n−1−1uTn
kn =
λ + u n R n −1u n
* −1 T
, R −1
n =
1 −1
λ
R (
n −1 − k *
)
−1
n n n −1 ,
u R

e n = e n −1 + k n (cn − e n −1uTn )
• λ determines the tracking ability of the RLS equalizers.
22
NON LINEAR EQUALIZERS
 Nonlinear Equalization
•Used in applications where the channel distrotion is too severe
•Three effective methods [6]
¾Decision Feedback Equalization (DFE)
¾Maximum Likelihood Symbol Detection
¾Maximum Likelihood Sequence Estimator (MLSE)
 Nonlinear Equalization‐‐DFE
•Basic idea : once an information symbol has been detected and
decided
upon, the ISI that it induces on future symbols can be estimated
and
substracted out before detection of subsequent symbols
•Can be realized in either the direct transversal form (see Fig.8) or
N N
as aˆ
• d k = ∑ C n y k − n + ∑ Fi d k − i
2 3
*

lattice filter
n=− N 1 i =1

[ ] T Tπ
• E e(n) min = exp{ ∫− π ln[
2 No
2π T F(e ) + No
jωT 2
]dω}
Nonlinear Equalizer‐DFE

Fig.8 Decision feedback equalizer (DFE)

 Nonlinear Equalization‐‐DFE
•Predictive DFE (proposed by Belfiore and Park, [8])
•Consists of an FFF and an FBF, the latter is called a noise predictor
( see Fig.9 )
•Predictive DFE performs as well as conventional DFE as the limit
in the number of taps in FFF and the FBF approach infinity
•The FBF in predictive DFE can also be realized as a lattice structure
[9].
The RLS algorithm can be used to yield fast convergence
Nonlinear Equalizer‐DFE

Fig.9 Predictive decision feedback equalizer

 Nonlinear Equalizers
• Decision feedback equalizer (DFE)

29
 DFE Equalizer

Fig.5 Tapped delay line filter with both feedforward and feedback taps
 Nonlinear Equalization‐‐MLSE
•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
•Usually has a large computational requirement
•First proposed by Forney [10] using a basic MLSE estimator
structure and implementing it with the Viterbi algorithm
•The block diagram of MLSE receiver (see Fig.10 )
 Nonlinear Equalizers
• Zero‐forcing DFE

• MMSE DFE

32
 Nonlinear Equalizers
• Maximum‐likelihood sequence estimation
L
ui = ∑ f n ci − n + ni
n =0

1 ⎧⎪ 1 N
⎛ L
⎞
2
⎫⎪
pdf (u | c; f ) =
(2πσ )
2 N /2
exp⎨− 2
⎪⎩ 2σ n
∑ ⎜
i =1 ⎝
u i − ∑
n =0
f n i −n ⎟
c
⎠
⎬
⎪⎭
n
2
⎛ N L
⎞
MLSE ⇒ {cn } = arg min ∑ ⎜ ui − ∑ f n ci − n ⎟
ˆ
i =1 ⎝ n =0 ⎠

33
 Nonlinear Equalizers
• Maximum‐likelihood sequence estimation
– Example

34
 Nonlinear Equalizers
• Maximum‐likelihood sequence estimation

35
 Nonlinear Equalizer‐MLSE

Fig.10 The structure of a maximum likelihood sequence equalizer(MLSE)

with an adaptive matched filter
•MLSE requires knowledge of the channel characteristics in order
to compute the matrics for making decisions
•MLSE also requires knowledge of the statistical distribution of the
noise corrupting the signal