Scribd
Upload
5 views5 pages

Convolution Sum Computation in DSP

The document provides examples and explanations of computing the convolution sum for finite and infinite length sequences. It shows how to set up a table to evaluate the convolution sum for two finite sequences expressed in sequence explicit representation. It also evaluates the convolution sum for two infinite sequences algebraically to arrive at a closed form solution.

Uploaded by

Kanav Mahajan
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
5 views5 pages

Convolution Sum Computation in DSP

The document provides examples and explanations of computing the convolution sum for finite and infinite length sequences. It shows how to set up a table to evaluate the convolution sum for two finite sequences expressed in sequence explicit representation. It also evaluates the convolution sum for two infinite sequences algebraically to arrive at a closed form solution.

Uploaded by

Kanav Mahajan
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

ECE 5650/4650 Modern DSP Notes !

Example: Consider two finite length se-


quences
x(n) N =n -n +1 h(n) N =n -n +1
x 2 1 h 4 3
n n
n1 n2 n3 n4
!
∞

∑
y[n]
! = x[k]h[n − k]
k=−∞
– The sum must be evaluated for each value of n
x(k)

k
n1 n2

h(n-k)

k
n-n4 n-n3
h(n-k)

k
n-n4 n-n3
h(n-k)

k
n-n4 n-n3
!
– The first nonzero value occurs when n! − n3 = n1 or
! = n1 + n3
n
– The last nonzero value occurs when n! − n4 = n2 or
n
! = n2 + n4

Additional Notes on Convolution Sum Computation


ECE 5650/4650 Modern DSP Notes !2

– Summary: y(n) is nonzero for at most


!n1 + n3 ≤ n ≤ n2 + n4 ⇒ Ny = Nx + Nh − 1
• Example: Two finite duration sequences given by
!x[n] = h[n] = u[n] − u[n − N ], N > 0
x(n) = h(n)

...
... ...
n
-1 0 1 2 N-1
!
– From the previous example we know that y[n]
! will be at
most nonzero on the interval [0,
! 2(N − 1)]
– There are 4 cases to consider:

x(k)
...
... ...
k
-1 0 1 2 N-1

h(n-k)

...
... ...
k
n-(N-1) n
!
– Case 1: !n < 0 which implies no overlap, so y[n]
! =0
– Case 2: !0 ≤ n ≤ N − 1
n

∑
y[n]
! = (1)(1) = n + 1

k=0

Additional Notes on Convolution Sum Computation


ECE 5650/4650 Modern DSP Notes !3

– Case3: !N − 1 ≤ n ≤ 2N − 2
N−1
(1)(1) = (N − 1) − (n(N − 1)) + 1
∑
!y[n] =
k=n−(N−1)
– Case 4: n! > 2N − 2 which implies no overlap, so
y[n]
! =0
– In summary:
n + 1, 0≤n≤ N−1
!y[n] = 2N − 1 − n, N − 1 ≤ n ≤ 2N − 2
0, otherwise
N
y(n)
... ...
... ...
n
-1 0 1 N-1 2N-1
!

Additional Notes on Convolution Sum Computation


ECE 5650/4650 Modern DSP Notes !4

Example: Two finite duration sequences


in sequence explicit representation:
h[n]
! = {1,2,1, − 1}, x[n] = {1,2,3,1}
↑ ↑

– In the above notation the arrows indicate where ! n = 0


– We need to evaluate the convolution sum for ! −1 ≤ n ≤ 5
– To evaluate construct the following table:
x[k]
! vs k
!h[n − k] 0 0 1 2 3 1 0 0 !y[n]
n = -1 -1 1 2 1 0 0 0 0 0 0 1
n=0 0 -1 1 2 1 0 0 0 0 0 4
n=1 0 0 -1 1 2 1 0 0 0 0 8
n=2 0 0 0 -1 1 2 1 0 0 0 8
n=3 0 0 0 0 -1 1 2 1 0 0 3
n=4 0 0 0 0 0 -1 1 2 1 0 -2
n=5 0 0 0 0 0 0 -1 1 2 1 -1

– The final output is thus


!y[n] = {…0,1,4,8,8,3, − 2, − 1,0,0,…}
↑
– Is this reasonable? The output should start at (-1 + 0) = -
1 and should stop at (3 + 2) = 5, which is indeed the case

Additional Notes on Convolution Sum Computation


ECE 5650/4650 Modern DSP Notes !5

Example: Two infinite duration se-


quences,
!x[n] = u[n], h[n] = a nu[n]
– By direct substitution into the convolution sum formula
we have
∞
a ku[k]u[n − k]
∑
y[n]
! =
k=−∞
– The term u(k) sets the lower sum limit to zero while the
term u(n - k) sets the upper sum limit to n, hence
0, n < 0 1 − a n+1
{ k=0
!y[n] = n k = u[n]
∑ a , n ≥ 0 1−a

1 y(n)
1-a
...
... 1 ...
n
-1 0 1 2
!

Additional Notes on Convolution Sum Computation

You might also like