Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Paper DSP

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 3

National University of Sciences and Technology (N.U.S.

T)
Department of Computer Engineering, College of E & ME
Department of Computer Engineering
Final Exam (DE-34 CE)
Digital Signal Processing
Dr. Shoab A Khan
Marks: 100 Time: 180 mins
Note: One A4 handwritten sheet is allowed in the exam. Attach paper with answer sheet.
Question 1: (FFT, DFT, DFS)
(20=7,7,6)
a) Given is the butterfly structure of an 8-point FFT with intermediate values, find the
values of a, b, and c.
1 2 6 6
a
-2 0
b c
4 -2
b) 2 -6j
2 0
2
-1

Given the FSexpansion of a periodic signal ( N=4 )in complex exponents find the original
periodic signal ^
x [n] where
(− j 2Npi .0. n)
X [0]e =[5 , 5 , 5 ,5]
( −j

.1.n )
X [ 1] e
N
=[−2+1 i, 1+2 i, 2−1 i ,−1−2 i]
(− j 2Npi .2 .n)
X [ 2] e =[−5 , 5 ,−5 , 5]
(− j 2Npi .3.n )
X [ 3] e =¿
c) DFT X[k] of a 4 point signal x[n] =[1 3 2 a] is

X [ k ]= [ 10 ,−1+ 1 j;−4 ,−1−1 j ] ;

Without explicitly computing x[n], find a using one of the properties of DFT

Question2:
(25=7, 8, 10)
a) Using z-transform find H (z )and h [n ]from the following difference equation.
y ( n )=0.06 y ( n−2 ) +0.1 y ( n+1 ) + x ( n )−0.8 x (n+1)
b) Draw the signal x[n] and compute it’s DTFT

1/2
−n
x [ n ] =(−2 ) u [ n – 10 ]
1
c) For the signal Xc ( jΩ) shown in figure below, sample the signal at =200 ,then convert
T
the sampling rate of the digital signal to 3/4 using digital means appropriately using any
decimation or interpolation filter if required and then convert the signal back to
continuous domain.
a. Draw a diagram showing all the required blocks like C/D, up sampler, down
sampler, and D/C etc
b. Stepwise show the frequency spectrum at the output of these blocks.
1
Xc ( jΩ)

-2π 100 2π 100

Question3: (Convolution and Circular Convolution)


(20=10,8,2)
a) Using flip and drag method compute linear convolution of the following two sequences
x [n]=u [n]−u [n−(N +1)]
h [ n ] =a ( u [ n ] – u [ n−( N +1 ) ] ) where |a|<1
n

x[n]

h[n]
N
b) Also calculate the N-point circular convolution of the same sequences using the result of
linear convolution.
c) Who can you make the circular convolution to be same as linear convolution.

Question 4: (Filter Design and Analysis)


(35=12, 7,8,8)
a) Draw the poles of a digital IIR filter using (a) and (b) if given are the filter order and
cut-off frequency for Butterworth filter design technique
Ωc =0.4
N=4
a. Impulse Invariance assuming T=1
b. Bilinear Transform assuming T = 2;

2/2
b) Implement a second order IIR filter with given H (z )as transposed direct form II
1 −2
1– z
3
H ( z )=
1 1
1− z−7 + z−11
4 2

c) Write MATLAB Code to design an FIR filter using ‘rectangular’ window of length
N=50 with a cut-off frequency of w c =π /4 without using any MATLAB
high-level function.
d) Given the following pole-zero plot of a system ( H ( z )) having conjugate
π
poles at 0.75 e ± j 4 and a zero at π, graphically compute its magnitude
π
and phase response at w=
2

3/2

You might also like