EE5713 : Advanced Digital Communications Week 12, 13: Inter Symbol Interference (ISI) Nyquist Criteria for ISI Pulse Shaping and Raised-Cosine Filter 20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 1 Eye Pattern Equalization (On Board)
35
Embed
Inter Symbol Interference (ISI) Nyquist Criteria … : Advanced Digital Communications Week 12, 13: Inter Symbol Interference (ISI) Nyquist Criteria for ISI Pulse Shaping and Raised-Cosine
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
EE5713 : Advanced Digital Communications
Week 12, 13: � Inter Symbol Interference (ISI)
� Nyquist Criteria for ISI
� Pulse Shaping and Raised-Cosine Filter
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 1
� Eye Pattern
� Equalization (On Board)
Baseband Communication System� We have been considering the following baseband system
� The transmitted signal is created by the line coder according to
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 2
to
where an is the symbol mapping and g(t) is the pulse shapeProblems with Line Codes
� One big problem with the line codes is that they are not bandlimited
� The absolute bandwidth is infinite
� The power outside the 1st null bandwidth is not negligible. That is, the power in the sidelobes can be quite high
∑∞
−∞=
−=n
bn nTtgats )()(
� If the transmission channel is bandlimited, then high frequency components will be cut off
– Hence, the pulses will spread out
– If the pulse spread out into the adjacent symbol periods, then it is said that intersymbol interference (ISI) has occurred
Intersymbol Interference (ISI)
Intersymbol Interference (ISI)
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 3
� Intersymbol interference (ISI) occurs when a pulse spreads out in such a way that it interferes with adjacent pulses at the sample instant
� Causes
– Channel induced distortion which spreads or disperses the pulses
– Multipath effects (echo)
– Due to improper filtering (@ Tx and/or Rx), the received pulses overlap one another thus making detection difficult
Pulse spreading
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 4
another thus making detection difficult
� Example of ISI
– Assume polar NRZ line code
– Input data stream and bit superposition
Inter Symbol Interference
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 5
� The channel output is the sum of the contributions from each bit
Note:
� ISI can occur whenever a non-bandlimited line code is used over a bandlimited channel
� ISI can occur only at the sampling instants
� Overlapping pulses will not cause ISI if they have zero
ISI
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 6
� Overlapping pulses will not cause ISI if they have zero amplitude at the time the signal is sampled
ISI Baseband Communication System Model
channel,theofresponseImpulse)(
r,transmittetheofresponseImpulse)(where
==
th
th
C
T
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 7
receivertheofresponseImpulse)(
channel,theofresponseImpulse)(
==
th
th
R
C
∑ −=∞
−∞=nTn nTthats ),()(
∑∞
−∞=
==+−=n
sCTTn fTththtgwheretnnTtgatr /1),(*)()(),()()(
∑∞
−∞=
+−=n
een tnnTthaty )()()( ),(*)(*)()( ththththwhere RCTe =)(*)(*)()( ththtntn RCe =
� Note that he(t) is the equivalent impulse response of the receiving filter
� To recover the information sequence {an}, the output y(t) is sampled at t = kT, k = 0, 1, 2, …
� The sampled sequence is
∑∞
−∞=
+−=n
een kTnnTkThakTy )()()(
AWGN term
ISI Baseband Communication System Model
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 8
or equivalently
– h0 is an arbitrary constant
−∞=n
∑ ∑∞
−∞=
∞
≠−∞=−− ++=+=
n knnknknkknknk nhaahnhay
,0
,..2,1,0),(),(where ±±=== kkTnnkThh ekek
AWGN term
Effect of other symbols at the sampling instants t=kT
Desired symbol scaled by gain parameters h0
Signal Design for Bandlimited Channel
Zero ISI
� To remove ISI, it is necessary and sufficient to make the term
∑∞
≠−∞=
+−+=knn
eenk kTnnTkThaahkTy,
0 )()()(
0,0)(0≠≠=− handknfornTkTh
e
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 9
� Nyquist Criterion
– Pulse amplitudes can be detected correctly despite pulse spreading or overlapping, if there is no ISI at the decision-making instants
Nyquist Criteria� Nyquist’s three criteria
– Pulse amplitudes can be detected correctly despite pulse spreading or overlapping, if there is no ISI at the decision-making instants
• 1: At sampling points, no ISI
• 2: At threshold, no ISI
• 3: Areas within symbol period is zero, then no ISI
Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713)
1st Nyquist Criterion: Time domain
p(t): impulse response of a transmission system (infinite length)
1p(t)
� shaping function
Equally spaced zeros,
interval Tfn
=2
1
TfN
=2
1
02t0t
t0
-1
no ISI !
Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713)
1st Nyquist Criterion: Time domain
Suppose 1/T is the sample rateThe necessary and sufficient condition for p(t) to satisfy
( ) ( )( )
≠=
=0,0
0,1
n
nnTp
Is that its Fourier transform P(f) satisfy
( ) TTmfPm
=+∑∞
−∞=
Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713)
1st Nyquist Criterion: Frequency domain
( ) TTmfPm
=+∑∞
−∞=
2a Nf f= 4 Nff
0(limited bandwidth)
Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713)
Proof
( ) ( ) ( )( )
( )∑ ∫
∞
−∞=
+
−=
m
Tm
TmdffnTjfPnTp
212
2122exp π
( ) ( ) ( )∫∞
∞−= dfftjfPtp π2exp
( ) ( ) ( )∫∞
∞−= dffnTjfPnTp π2expAt t=T
Fourier Transform
( ) ( )
( ) ( )
( ) ( )∫
∫ ∑
∑ ∫
−
−
∞
−∞=
∞
−∞=−
−∞=
=
+=
+=
T
T
T
Tm
m
T
T
m
dffnTjfB
dffnTjTmfP
dffnTjTmfP
21
21
21
21
21
21
2exp
2exp
2exp
π
π
π
( ) ( )∑∞
−∞=
+=m
TmfPfB
Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713)
Proof
( ) ( )∑∞
−∞=
=n
n nfTjbfB π2exp
( ) ( )∫− −=T
Tn nfTjfBTb21
212exp π
( ) ( )∑∞
−∞=
+=m
TmfPfB
( )nTTpbn −= ( )( )
==
0nTb ( )
≠=
00 nbn
( ) TfB = ( ) TTmfPm
=+∑∞
−∞=
Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713)
� Pulse shape that satisfy this criteria is Sinc(.) function, e.g.,
� The smallest value of T for which transmission with zero ISI is possible is
)2(sinsin)(or)( WtcT
tctpthe =
=
T1=
Nyquist Criterion: Time domain
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 16
� Problems with Sinc(.) function
– It is not possible to create Sinc pulses due to
– Infinite time duration– Sharp transition band in the frequency domain
– Sinc(.) pulse shape can cause ISI in the presence of timing errors
• If the received signal is not sampled at exactly the bit instant, then ISI will occur
WT
21=
Nyquist Criterion: Time domain
1p(t)
� shaping function
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 17
Equally spaced zeros,
interval Tf s
=2
1
Tf s
=2
1
02t0t
t0
-1
no ISI !
Sample rate vs. bandwidth
� W is the channel bandwidth for P(f)
� When 1/T > 2W, there is no way, we can design a system with no ISI
P(f)
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 18
Sample rate vs. bandwidth
� When 1/T = 2W (The Nyquist Rate), rectangular function satisfy Nyquist condition
( ) ( ) ( )( )
,otherwise,0
,;sinc
sin
<
=
==WfT
fPT
t
t
Tttp
πππ
( ) ( );rect2
rect21
fTTW
f
WfP =
=
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 19
( ) ( );rect2
rect2
fTTWW
fP =
=
T
W
Sample rate vs. bandwidth
� When 1/T < 2W, numbers of choices to satisfy Nyquist condition
– Raised Cosine Filter
– Duobinary Signaling (Partial Response Signals)
– Gaussian Filter Approximation
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 20
– Gaussian Filter Approximation
� The most typical one is the raised cosine function
Raised Cosine Pulse
� The following pulse shape satisfies Nyquist’s method for zero ISI
� The Fourier Transform of this pulse shape is
2
22
2
22 41
cossinc
41
cossin)(
T
trT
tr
T
t
T
trT
tr
T
trT
tr
tp
−
=−
=
ππ
π
π
20-May-15 Muhammad Ali Jinnah University, Islamabad Advanced Digital Communications (EE5713) 21
� The Fourier Transform of this pulse shape is
� where r is the roll-off factor that determines the bandwidth
+≥
+≤≤−
−−+
−≤≤
=
T
rf
T
rf
T
r
T
rf
r
TT
T
rfT
fP
21
||,0
21
||2
1,
21
||cos12/
21
||0,
)(π
Responses for different roll-off factors (a) Frequency response. (b) Time response
� Bandwidth occupied beyond 1/2T is called the excess bandwidth (EB)
� EB is usually expressed as a %tage of the Nyquist frequency, e.g.,