Correlation of Discrete-Time

Signals

Transmitted Signal, x(n)

Reflected Signal,
y(n) = x(n-D) + w(n)

0 T
 Cross-Correlation
• Cross-correlation

of x(n) and y(n) is a sequence, rxy(l)
rxy l    x n y n  l 
n 
l  0, 1, 2,

rxy l    x n  l  y n 
n 
l  0, 1, 2,

• Reversingthe order, ryx(l)

ryx l    y n  x n  l 
n 
l  0, 1, 2,

ryx l    y n  l  x n 
n 
l  0, 1, 2,

• => rxy l   ryx  l 

 Similarity to Convolution
• No folding (time-reversal)

rxy l   x l   y  l  ryx l   y l   x  l 

• In Matlab:
– Conv(x,fliplr(y))
 Auto-Correlation
• Correlation of a signal with itself

rxx l    x  n  x  n  l   r  l 
n 
xx l  0, 1, 2,

• Used to differentiate the presence of a

like-signal, e.g., zero or one
• Even function
 Properties
• Two sequences, x(n) and y(n), with finite
energy z  n   ax  n   by  n  l 
• Find energy of z(n)

Ez    ax  n   by  n  l 
2

n 
  
a 2
 x  n   b  y  n  l   2ab  x  n  y  n  l 
n 
2 2

n 
2

n 

 a 2 rxx  0   b 2 ryy  0   2abrxy l   0

Ex Ey
Ez  a 2 rxx  0   b 2 ryy  0   2abrxy l   0 (assume b  0)
2
a a
   rxx  0   2   rxy l   ryy  0   0
b b

Quadratric in (a/b) and positive, discriminant is non-negative:

For crosscorrelation case:
rxy l   rxx  0  ryy  0   Ex E y

For autocorrelation case:

rxx l   rxx  0   Ex

Maximum value occurs with zero lag (when signals are perfectly matched)
Often normalized to range [-1,1]:

rxy l  rxx l 
 xx l    xx l  
rxx  0  ryy  0  rxx  0 
x  n   a nu  n  ; 0  a 1


rxx l    x n  x n  l 
n 
 
 a a n n l
a l
a 2n

n l n l

1 l
 a l0
1  a2
rxx l 
 xx l   a
l
  l  
rxx  0 
 Periodic Sequences
• Power signals crosscorrelation:
M
1
r l   lim
xy
M  2M  1

n  M
x n  y n  l 

• Define auto and crosscorrelations over

one period of the signals
• If x(n) and y(n) are periodic signals with
period N: N 1
1
xyr l    x  n  y  n  l 
N
n 0

• Correlations are also periodic with period

N
y(n)=x(n)+w(n)
 LTI Systems
• Convolution, output of LTI system

y n   h n   x n    h k  x n  k 
k 

• Crosscorrelation between the output and

input signal: r l   y l   x  l   h l   x l   x l 
yx

 h l    x l   x  l 
 h l   rxx l 

• Similarly, input to output is:

rxy l   h  l   rxx l 
• Autocorrelation of output:
ryy l   y l   y  l    h l   x l    h  l   x  l 
  h l   h  l    x l   x  l 
 rhh l   rxx l 


ryy  0   rhh  0   rxx  0    r k  r k 
k 
hh xx