0% found this document useful (0 votes)
24 views16 pages

9 Digital Filters.

1. The document discusses digital filters and signals and systems concepts. 2. It provides 8 multiple choice questions related to Fourier transforms, linear time invariant systems, impulse responses, sampling, and properties of discrete time signals and systems. 3. The questions cover topics such as energy calculation of a sinusoidal signal, Fourier transform properties, convolution of sequences, frequency response of filters, and the discrete Fourier transform.

Uploaded by

goutham a
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
0% found this document useful (0 votes)
24 views16 pages

9 Digital Filters.

1. The document discusses digital filters and signals and systems concepts. 2. It provides 8 multiple choice questions related to Fourier transforms, linear time invariant systems, impulse responses, sampling, and properties of discrete time signals and systems. 3. The questions cover topics such as energy calculation of a sinusoidal signal, Fourier transform properties, convolution of sequences, frequency response of filters, and the discrete Fourier transform.

Uploaded by

goutham a
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
You are on page 1/ 16

Signals and Systems Digital Filters

UNIT DIGITAL
 FILTERS
5. A Fourier transform pair is given by
ONE MARK QUESTIONS n  j 6 f
2 Ae
  u  n  3 
FT
sin c n
1. Given x  n   , the energy of the 3 2
n 1    e  j 2 f
 3
 x  n
2
signal given by is Where u  n denotes the unit step
n 

(IN-GATE-03) sequence. The value of A is _____


(EC-GATE-14)
c
(a) (b) c 6. A discrete time signal x  n  is obtained

(c) Infinite (d) 2c by sampling an analog signal at 10 kHz.
The signal x  n  is filtered by a system
2. If the Fourier transform of x  n is
h  n 
X  e j  , then the Fourier transform of
with impulse response
0.5   n     n  1 . The cut-off
 1 x  n 
n
is (IN-GATE-04)
frequency of (IN-GATE-14)
 j X  e j  (b)  1 X  e j 
 
(a) (a) 1.25 kHz (b) 2.50 kHz

(c) X e j    (d)
d
d

X  e j   7.
(c) 4.00 kHz (d) 5.00 kHz
An LTI system with unit sample
1
n
response h  n 
3. Let x  n     u  n  , y  n   x2  n 
2 5  n  7  n  1  7  n  3 5  n  4
and Y  e j  be the Fourier transform of (EC-GATE-17)
y  n  . Then Y  e j0
 is (EC-GATE-05) (a) Low pass filter
(b) High pass filter
1 4
(a) (b) 2 (c) 4 (d) (c) Band pass filter
4 3
(d) Band stop filter
4. For an N-point FFT algorithm with
8. Consider two 16-point sequences x[n]
N  2m , which one of the following
and h[n]. Let the linear convolution of
statements is TRUE? (EC-GATE-10) x[n] and h[n] be denoted by y[n], while
(a) It is not possible to construct a z[n]denotes the 16-point inverse
signal flow graph with both input discrete Fourier transform (IDFT) of the
and output in normal order product of the 16-point DFTs of x[n]
(b) The number of butterflies in the
and h[n]. The value(s) of k for which
m th stage is N/m
z[n] = y[k] is/are [ECE-GATE-2021]
(c) In place computation requires
storage of only 2N node data (a) k = 0, 1, 2, ……, 15
(d) Computation of a butterfly requires (b) k = 0 and k = 15
only one complex multiplication (c) k = 0 (d) k = 15
113 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
(c)
TWO MARK QUESTIONS
1. The impulse response h  n of a linear
time invariant system is given as h  n  
2 2, n  1,  1

4 2, n  2,  2 . If the input (d)
0, otherwise

to the above system is the sequence
e jn /4 , then the output is
(EC-GATE-04)
(a) 4 2e jn /4 (b) 4 2e jn /4
(c) 4e j n /4 (d) 4e j n / 4 3. The Fourier transform of y  2n  will be

Statement of Linked Answer (EC-GATE-05)


Questions 02 & 03 (a) e j 2  cos 4  2 cos 2  2
A sequence x  n  has non-zero values (b)  cos 2  2cos   2
as shown in the figure. (EC-GATE-05) (c) e j  cos 2  2 cos   2
x  n (d) e j 2  cos 2  2 cos   2
2
4. A signal x  n   sin  0 n    is the
1 1
1/ 2
input to a linear time invariant system
1/ 2
having a frequency response H  e j  . If
n
1
the output of the system is Ax  n  n0  ,
2 0 1 2

2. The sequence y  n   then the most general form of H  e j 


 n  will be (EC-GATE-05)
 x   1 , for n even
 2  will be (a)  n0 0   for any arbitrary real 
0, for n odd
 (b)  n0 0  2k for any arbitrary
(a) integer k
(c) n0 0  2k for any arbitrary
integer k
(d)  n0 0 

5. Match the following and choose the


(b) correct combination. (EC-GATE-05)
Group-I
A. Continuous and aperiodic signal
B. Continuous and periodic signal
C. Discrete and aperiodic signal
D. Discrete and periodic signal

114 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
Group-II 8. The 4-point Discrete Fourier Transform
1. Fourier representation is (DFT) of a discrete time sequence {1, 0,
continuous and aperiodic 2, 3} is (EC-GATE-00)
2. Fourier representation is discrete
and aperiodic
(a)  0, 2  2 j , 2, 2  2 j 
3. Fourier representation is (b)  2, 2  2 j, 6, 2  2 j 
continuous and periodic
4. Fourier representation is discrete
(c)  6, 1  3 j , 2, 1  3 j 

and periodic (d)  6, 1  3 j , 0, 1  3 j 


Codes:
(a) A  3, B  2, C  4, D  1 9. A system with transfer function H  z 
(b) A  1, B  3, C  2, D  4 has impulse response h  n  defined as
(c) A  1, B  2, C  3, D  4
h  2   1, h  3  1 and hk   0
(d) A  2, B  1, C  4, D  3
otherwise. Consider the following
6. A 5-point sequence x  n is given as statements. (EC-GATE-09)
x  3  1, x  2  1, x  1  0, S1: H  z  is a low pass filter
x  0  5, x 1  1. Let X  e j  denote S2: H  z  is an FIR filter
the discrete time Fourier transform of Which of the following is correct?

 X  e  d  is
x  n . The value of j (a) Only S2 is true
 (b) Both S1 and S2 are false
(EC-GATE-07) (c) Both S1 and S2 are true, and S2 is a
(a) 5 (b) 10 reason for S1
(c) 16 (d) 5  j10  (d) Both S1 and S2 are true, but S2 is
not a reason for S1
7.  x  n  is a real valued periodic
10. The first six points of the 8-point DFT
sequence with a period N. x  n  and of a real valued sequence are 5, 1  j 3,
X  k  form N-point Discrete Fourier 0, 3  j 4, 0 and 3  j 4. The last two
Transform (DFT) pairs. The DFT Y  k  points of the DFT are respectively
of the sequence (EC-GATE-11)
1 N 1 (a) 0, 1  j 3 (b) 0, 1  j 3
y n 
N
 x  r  x  n  r  is
r 0 (c) 1  j 3 , 5 (d) 1  j 3 , 5
(EC-GATE-08)
11. The DFT of a vector  a b c d  is
X k 
2
(a)
the vector      . Consider the
N 1
1
(b)
N
 X r  X k  r 
r 0

product
N 1 a b c d
1
(c)
N
 X k  r  X r 
r 0 p q r
d
s   a b c d  
a b c  .
c d a b
(d) 0  
b c d a

115 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters

The DFT of the vector  p q r s  is 15. Two sequence  a, b, c  and  A, B,C  are
a scaled version of (EC-GATE-13)  A 1 1 1  a 
(a)   2 2 2  2  related as,  B   1 W3 W32  b  .
   1

C  1 W32 W34   c 


(b)       2
j
Where W3  e 3
if another sequence
(c)            
 p, q, r  is derived as,
(d)    
 p  1 1 1  1 0 0   A / 3
 q   1 W 1 W 2  0 W 2 0   B / 3
12. The N-point DFT x of a sequence    3 3  3  
x  n , 0  n  N 1 is given by  r  1 W32 W34  0 0 W34  C / 3
. Then the relationship between the
N 1 2
1 j
X k    x n e sequence  p, q, r  and  a, b, c  is
nk
N
,
N n0
(EC-GATE-15)
0  k  N  1 . Denote this relation as
X  DFT  x  . For N  4 which one of
(a)  p, q, r   b, a, c 
the following sequence satisfies DFT (b)  p, q, r   b, c, a 
(DFT(x))  _______ (EC-GATE-14) (c)  p, q , r    c , a, b 
(a) x  1 2 3 4 (d)  p, q, r    c, b, a 
(b) x  1 2 3 2  n 
16. Let x  n   1  cos   be a periodic
(c) x  1 3 2 2  8 
signal with period 16. Its DFS
(d) x  1 2 2 3
coefficients are defined by
13. Consider a discrete time periodic signal 1 15   
ak   x  n  exp   j kn  for all k.
 n 
16 n 0  8 
x  n   sin   . Let ak be the
 5  The value of the coefficient a31 is ____
complex Fourier series coefficients of (EC-GATE-15)
x  n . The coefficients ak  are non- 17. Consider two real sequence with time
origin marked by the bold value.
zero when k  Bm  1, when m is any
x1  n  1, 2,3, 0 , x2  n   1,3, 2,1 , .
integer. The value of B is _____
Let X 1  k  and X 2  k  be 4-point DFTs
(EC-GATE-14)
14. X  k  is the discrete Fourier Transform of x1  n and x2  n  respectively.

of a 6-point real sequence x  n  . If Another sequence x3  n  is derived by

X  0   9  j 0, X  2   2  j 2, taking 4-point inverse DFT of X 3  k  

X  3  3  j 0, X  5  1  j1, x  0  is X 1  k  X 2  k  . The value of x3  2 is


(IN-GATE-14) ______ (EC-GATE-15)

116 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
18. The discrete Fourier Transform (DFT) h  n  0 for n  0 and n  2. Let
of the 4-point sequence
H   be the discrete time Fourier
x  n    x  0  , x 1 , x  2 , x 3
transform (DTFT) of h  n . where  is
 3, 2,3, 4 is Xl1 k  is the DFT of the
the normalized angular frequency in
12-point sequence x1  n   {3,0,0,2,
radians. Given that H  0   0 and
X 1 8 
0,0,3,0,0,4,0,0} the value of is 0  0  , the value of 0 (in radian)
X 1 11
is equal to _____ (EC-GATE-17)
_______ (EC-GATE-16)
22. Let h  n be a length-7 discrete time
19. A continuous time speech signal xa  t 
finite impulse response filter, given by
is sampled at a rate of 8 kHz and the
h  0  4, h 1  3, h  2  2, h 3  1,
samples are subsequently grouped in
h  1  3, h  2  2, h  3  1 and
blocks, each of size N. The DFT of each
block is to be computed in real time h  n is zero for n  4. A length-3
using the radix-2 decimation in finite impulse response approximation
frequency FFT algorithm. If the g  n of h  n has to be obtained such
processor performs all operations 
H  e j    G  e j  d 
2
sequentially, and takes 20 s for that E  h, g   

computing each complex multiplication
is minimized, where H  e j  and
(including multiplications by 1 and 1 )
and the time required for G  e j  are the discrete-time. Fourier
addition/subtraction is negligible, then
transforms of h  n and g  n ,
the maximum value of N is ________
respectively. For the filter that
(EC-GATE-16)
minimizes E  h, g  , the value of
20. Consider the signal
x  n  6  n  2  3  n  1  8  n 10 g  1  g 1 , rounded off to 2
7  n  1 4  n  2 . If X  e j  is decimal places, is ____ (EC-GATE-19)
23. Consider a six-point decimation in time
discrete time Fourier transform of
 Fast Fourier Transform (FFT)
1
x  n , then  X  e j  sin 2  2  d  is algorithm, for which the signal flow
 
graph corresponding to X 1 is shown
equal to _____ (EC-GATE-16)
 j 2 
21. Let h  n be the impulse response of a in the figure. Let W6  exp    . In
 6 
discrete time linear time invariant (LTI)
the figure, what should be the values of
filter. The impulse response is given by
the coefficients a1 , a2 , a3 in terms of
1 1 1
h  0  ; h 1  ; h  2  ; and W5 so that X 1 is obtained correctly?
3 3 3

117 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters

X 0   j n   j n 
x 0 
x  n   c1 exp     c2  .
a1  2   2 
x  3 X 1
1
a2
Where c1 and c2 are real numbers. The
x 1 X 2
a3 desired three tap filter is given by
x 4 X  3
1 h  0  1, h 1  a, h  2  b and h  n  0
x  2 X 4
for n  0 or n  2 . What are the values
x 5 X 5 of the filter taps a and b if the output is
1
(a) a1  1, a2  W61 , a3  W62 y  n  0 for all n, when x  n is as
(b) a1  1, a2  W62 , a3  W61 given above? (EC-GATE-19)
(c) a1  1, a2  W62 , a3  W61 x n  n0
y n   0
(d) a1  1, a2  W6 , a3  W 6
2
h  n   1, a, b

24. It is desired to find a three-tap causal (a) a  1, b  1 (b) a  0, b  1


filter which gives zero signal as an (c) a  1, b  1 (d) a  0, b  1
output to an input of the form.

25. The input signal shown below

3 3 3 3

2 2 2 2

1 1 1 1 1 1 1 1 1 1 1 1

…… 0 0 0 0 0 0 0 0 ……

is passed through the filter with the following taps.


The number of non-zero output samples is _____. (IN-GATE-20)

ANSWER KEY
ONE MARK QUESTIONS
1 a 2 c 3 d 4 d 5 3.375
6 b 7 c 8 d
TWO MARK QUESTIONS
1 d 2 a 3 c 4 b 5 c
6 b 7 a 8 d 9 a 10 b
11 a 12 b 13 10 14 a 15 c
16 0.5 17 11 18 6 19 4096 20 8
21 2.094 22 –27 23 a 24 b 25 10

118 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters

SOLUTIONS
1. Answer: (a) 2n
1
Solution: y  n  x n    u  n 
2

2
sin  c n     n
 rect   1
n y  n    u  n 
 2c  4
Y  e j   y  n 

Now, Y e j 0    y n
n 

 n
1
Y  e      u  n 
j0

n   4 
sin  c n 
x  n  is the impulse  n
n 1 1 4
    
response o ideal DT LPF n 0  4 
1 3
1
4
With cutoff frequency  c and pass
4. Answer: (d)
band magnitude  1 as shown
Solution:
Using Parsaval’s relation
For an N-point FFT algorithm with
 
1
X  e j  d  N  2m , computation of a butterfly
2
 x  n
2
E  
2 
n  requires only one complex
multiplication and two complex
2c c
E  additions.
2 
For an N-point FFT algorithm: For
2. Answer: (c) Decimation in Time (DIT) algorithm
Solution: input is in the shuffle order and output
is in normal order.
If x  n  
FT
 X  e j 
(i) For Decimation in Frequency (DIF)
algorithm input is in the normal
e jo n x  n  
FT
X e  
j  o 
order and output is in the shuffle
order.
 1 x  n   e jn x n  X e   
n j 
(ii) There are ‘m’ stages of computation
3. Answer: (d) and in each stage the number of
butterflies is N/2.
Solution:
(iii) In place computation requires
Given: storage of only N node data.
n (iv) Computation of butterfly requires
1
x  n    u  n  one complex multiplication and
2 two complex additions.
119 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
5. Answer: 3.375 H  e j 0   1 H  e j 0   1
Solution:
 j 2  1 1
n H e    j
2  2 2
Given: x  n     u  n  3 
3
 j  1 1 1
i.e., graph will start from n  3 H e 2     , H  e j    0,
  4 4 2

X  e j    x n e  j n
H  e j   0
n 3

 n  2 f
 2   j n 3-dB cut off frequency,   
    e
n 3  3 
2 Fs
Fs 10 kHz
2 

n
f    2.5 kHz
X  e     e  j 
j
4 4
n 3  3 
7. Answer: (c)
3 2
2  2  Solution:
  e  j    e  j   ....
3  3  Given
3 h  n   5   n  7   n  1  7   n  3
2   2  
X  e    e j  1   e  j   ....
j

3   3   5  n  4
2  1 Frequency response,
  e j 
3  1   2  e  j
  H  e j   5  7e  j  7e 3 j  5e 4 j
 3
Putting   0
(Using binomial theorem)
3 H e j0   5  7  7  5  0
2 e j 3
 
3  2 Putting   
1    e  j
 3 H  e j   5  7  7  5  0
3 j 6 f
3 e System would act as band pass filter for
 
 2  1   2  e  j 2 f
  frequency range  0, 
3
8. Answer: (d)
3
3 Solution:
 Value of A     3.375
2 z(n) = x(n) * h(n)
6. Answer: (b) Let x(n) 
DFT
 X(k)
Solution: h(n) 
DFT
 H(k)
1 1 1 y(n)  x(n) & h(n) 
DFT
 X(k).H(k)
H z   z
2 2
Z(k)  Y(k)
1 1  j
H  e j    e When k  15
2 2
120 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters

TWO MARK SOLUTIONS  0 


 x  2  1 for n  0
  
1. Answer: (d)  2 
 x   1 for n  2
Solution:  2 
 4 
h  n  can be represented as  x   1 for n  4
 2 
y  n  
h  n   2 2  n  1  2 2  n  1  x  6  1 for n  6
  2 
4 2  n  2   4 2  n  2  ------ (1) 
 x  2  1  for n  2
  2 

 
Input x  n   e jn /4  e jn , where  
4  x   4  1 for n  4
  2 

From equation (1), the frequency
response is  x  1  1 for n  0

H  e j   2 2  e j  e  j   4 2  e j 2   e 2 j   x  0  2 for n  2

 4 2 cos    8 2 cos  2 y  n    x 1  1 for n  4

  x  2  0.5 for n  6
For    x  2  0.5 for n  2
4 
y n
 
H  e j / 4   4 2 cos  8 2 cos    4
4 2 2
For complex exponential input 1 1
1/ 2 1/ 2
x  n   e jn
n

Output, y  n   H  e j
 x n 2 0 2 4 6

So, answer will be (a)


n
j
y  n   4e 4 3. Answer: (c)
Solution:
2. Answer: (a)
2
Solution:
1 1
Given sequence is 1/ 2 1/ 2

x n
n
2
2 1 0 1 2
1 1
1/ 2 1/ 2 y  2n   x  n  1

n f  n   y  2n 
2 1 0 1 2

f  n  
FT
F  e j 
 n 
x 1 ; for n even
y  n    2  

0, F  e j    f n e  j n

 for n odd n 

121 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
 0.5e j  1e  j 0   2e  j  1 e 2 j  0.5 e 3 j 6. Answer: (b)

 e  j  0.5 e 2 j  e j  2  e  j  0.5e 2 j 
Solution:

 e j 2  0.5  e j 2   e j 2     e j  e j   Given, x  n   1 1 0 5 1


  

 e  j  2  cos 2  2cos 
x  n  
DTFT
 X  e j 
4. Answer: (b)

1
Solution: x  n   X  e j  e jn d 
2 
Let x  n  
DTFT
 X  e j 
Apply the area property of DTFT
Output, y  n   Ax  n  no  
x 0  5   X e  d
j

As per Time shifting property 


Y  e j   Ae  jno  X  e j 
 X  e  d   10
j


The frequency response is given by
7. Answer: (a)
Y  e j 
H  e j    Ae  jno  Solution:
X e j
 x  n  
DFT
 X K ------(1)
H  e j
   n    A   n   2 k N 1
o o
1
Given, y  n  
N
 x r  x n  r 
Now for x  n   sin  o n    r 0

… (2)
At   o
Take DFT on both sides of (2)
H  e j
  n   2k N 1
o o
1
Y K  
N
  x  r   DFT of x  n  r 
(For any arbitrary integer k) r 0

5. Answer: (c) … (3)

Solution: Use time shifting property in eq. (1)

Fourier representation of a signal can be DFT of x  n  r  WN rK X  K 


determined by following:
… (4)
Continuous  Aperiodic
Where, WN  e  j 2  / N
Discrete  Periodic
N 1
1
From above it can be derived that a Now, Y  K   X  K   x  r  WN rK
N r 0
continuous signals Fourier
representation results into aperiodic … (5)
signal and vice versa. Similarly, a N 1

discrete signals Fourier representation By definition, X  K    x  n WNnK


n0
results into periodic signal and vice
versa. … (6)
122 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters

H  e j    e  j 2   e  j 3
N 1
X   K    x  n WN nK
n 0
5
j   j 2 j 

N 1 e 2
 e  e 2

  x  n  WN nK … (7)  
n 0

( x  n  is real) 5  j 2 j 

e  e  2 j
j  2
e 2

N 1
 2j 
X   K    x  r  WN rK … (8)  
r 0
5
j  
Substituting (8) in (5)  2 je 2
sin
2
1 1
Y K   X K  X  K   X K
2

N N  H  e j   2 sin
2
If factor 1/N is used in (6)
Graph:
Y K   X K
2

 
H e j

8. Answer: (d)
2
Solution:
1 1 1 1   1   6  
1  j 1 j   0   1  3 j   0 
     
1 1 1 1  2   0  The graph shows the impulse response
    
1 j 1  j   3   1  3 j  of HPF.

6,  1  3 j, 0,  1  3 j  Hence S1 statement is false.

9. Answer: (a) Now, since h  n    n  2    n  3


Solution: contains finite length. Hence it
Given: represents an FIR filter.
10. Answer: (b)
h  2  1
Solution:
h 3  1
For a real valued sequence DFT
h  K   0, otherwise X K   X * N  K 
h n Here N  8 where n  0 to 7
1
X  K   X  8  K 
3 n
0 1 2 4 X  6   X   2  0
1
X  7   X  1  1  j 3

h  n     n  2     n  3 X  6  0

H  z   Z 2  Z 3 X 7  1  j3

123 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
11. Answer: (a) 1 1 1 1   x1 
2

   
Solution: 1 1  j 1 j   x2 

Given 4 1 1 1 1  x3 
   
a b c d 1 j 1  j   x4 
d a b c 
p q r s  a b c d  
c  4 x1   x1 
d a b
     
b c d a 1  4 x4   x4 
 
4  4 x3   x3 
Taking transpose on both sides    
 4 x2   x2 
 p  a d c b a
q  b a d c   b  Given DFT  DFT  x    x
 
r  c b a d  c
     For that to be true x4  x2
 s  d c b a  d 
Only option (b) satisfies this condition.
 p a  a
 q  b  b  13. Answer: 10
Hence,    
 r  c 
  N   
c Solution:
     
 s  d  d 
 n 
So DFT of x  n   sin  
 5 
2
 p   a    2  j
n
j
n
q     2  e 5
e 5

    DFT of  b      
r    c    2  2j
       2 
s   d    1 j n5 1  j n5
 e  e
2j 2j
Property: x  n   N  y  n   X  k  Y  k 
N 1 n
  ak e
j .k
12. Answer: (b) 5

k 0
Solution:
Here N  4  2 2 
 Here, N    5  10 
 o  
DFT  DFT  x  
1
 W40 W40 W40 W40   x1   Thus, a1   a11  a21  ..... and
  0    2j
 DFT  1 W4 W41 W42 W43   x2  

 4 W4 W42 W44 W46   x3   1
0

  a 1    a19  a29  ....


 W4
0
W43 W46 W49   x4   2j

W40 W40 W40 W40   x1  Since ak are also periodic with a period
  
1 W40 W41 W42 W43   x2  of N  10, ak are non-zero coefficients

4 4 W40 W42 W44 W46   x3  for k  Bm  1 (for any integer ‘m’),
 0  
W4 W43 W46 W49   x4  when B  10.

124 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
14. Answer: (a)  p 1 W32 W31   A
Solution:  q   1 1 1 
1   B 
  3
 r  1 W31 W32  C 
x  n  is real X  k   X   N  K   

N 6 1 W32 W31  1 1 1  a 
1 
 1 1 1  1 W3 W32   b 
1
X 1  X   6  1  X   5  1  j1 3
1 W31 W32  1 W32 W34   c 
 
X  4  2   X   6  4  X   2  2  j 2
1  W32  W31 1  W31  W31 1  W30  W30   a 
1 
j 2 nk   3 1  W31  W32 1  W32  W31   b 
1 N 1
x  n   X  K  e N 3
1  W31  W32 1  W30  W30 1  W31  W31   c 
 
N k 0
As W32  W31  1
5
1
x 0 
N
 X K 
K 0
W31  W31  1

W31  W32  1
  X  0   X 1  X  2  
  
6   X  3  X  4   X  5    1 j
2
1 3
W
 3  e 3
  j 
 2 2 
1 9  j 0  1  j1  2  j 2  3  2

 W 1  e j 3   1  j 3 
6   j 0  2  j 2  1  j1 
  3 2 2 
18  p 0 0 3 a 
 3
6  q   1 3 0 0 b 
  3  
15. Answer: (c)  r   0 3 0   c 
Solution:
 p  0 0 1   a   p   c 
Given:  q   1 0 0   b    q    a 
        
 A  1 1 1  a   r  0 1 0   c   r   b 
 B   1 W 1 W 2   b 
   3 3   16. Answer: 0.5
C  1 W32
W34   c 
Solution:
 p  1 1 1  1 0 0   A / 3  n 
 q   1 W 1 W 2  0 W 2 0   B / 3  Given, x  n   1  cos  
   3 3  3    8 
 r  1 W32 W34  0 0 W34  C / 3
N  16
 p  1 W W   A / 3
2 4
2 n j 2 n
3 3 1 j 1 
 q   1 W 1W 2 W 2W 4   B / 3 x  n   1  e 16  e 16

   3 3 3 3   2 2
 r  1 W3 W3 W3 W34  C / 3
2 2 4
By comparison
Now, 1 1
ao  1, a1  , a1 
W34  W331  W31 2 2
Now, a1  aN 1  a16 1  a15
W35  W32 and W34  W31
1 1
W36  W33  W30  1 So, a0  1, a1  , a2 to a14  0, a15 
2 2
125 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
Now DFS coefficients are periodic with 19. Answer: 4096
period N i.e., M  16 Solution:
So, a31  a1615  aN 15  a15 Given data
Sampling rate (or) sampling frequency
1
a31  a15 
2
 0.5  f s   8kHz
17. Answer: 11 Block size  N
Radix  2
Solution:
Time for computing each multiplication
x1  n   1, 2,3, 0
Tm   20 s
x2  n   1,3, 2,1

X 3 k   X1  k  .X 2  k 

 x3  n   x1  n  ( N ) x2  n

(N)  Circular convolution of N-point Ts


DFT
NTs
1 0 3 2 1 
1
2 1 0 3   3  Sampling time period Ts  
x3  n    fs
3 2 1 0   2
   1
0 3 2 1  1    125μs
8k
x3  n   9, 8, 11, 14 Time period of block size  N   NTs

  x3  0  , x3 1 , x3  2  , x3  3  N 125  


If the processor performs 1
x3  2  11
multiplication in 20 s then the no. of
18. Answer: 6 multiplication performed by the
Solution: processor in time period of NTs is
Here, x1  n   x  n / 3 1 20 s
?  N 125 s 
X 1  k    X  0 , X 1 , X  2 , X 3 , X  0 ,
No. of multiplications performed by
x 1 , x  2 , x 3 , x  0 , x 1 , processor in time period of
x  2 , x 3 N 125  
NTs  … (1)
20 
X 1 8  x  0 and X 1 11  X 3
We know,
X 1 8  12, X1 11  2 j No. of multiplications required for DIF
 FFT
X 1 8 12 N
  6  log 2N … (2)
X 1 11 2 j 2

126 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
Therefore, 21. Answer: 2.094
No. of multiplication of DIF – FFT Solution:
should be less than or equal to
1
multiplications performed by processor h  0 
in block time  NTs  3
1
N N 125   h 1 
log 2N  3
2 20 
1
125  h  2 
log 2N  3
10 
h  n  0
log  12.5
N
2
For and also given that H  o   0 and
log 2N  12
0  o  
(rounding to nearest decimal)
1 1 1 
log N  212 h n   , , 
3 3 3
N  4096
1 1
The maximum value of Block size Since, h  n    n    n  1
3 3
 N   4096
1
   n  2
20. Answer: 8 3
Solution: 1 1 1
H  e j    e  j  e2 j
1  cos  4  3 3 3
Let Y  e j   sin 2  2  
2 1 1
 e  j   e  j   e j  e  j  
1 1 j 4  j 4 3 3
  e  e 
2 4
 e j   e  j 
  cos 
If y  n is the inverse Fourier transform  2 
of Y  e j  then
1
H  e j   e j 1  2 cos 
1 1 1 3
y  n     n     n  4    n  4
2 4 4

H e jo  0
Using Parseval’s relation
 
1  2 cos o  0
1
 X  e j  Y   e j  d    x  n  y   n 
2   n  1
cos o 
As Y  e j
  Y  e 
j 2
 1
1 
o  cos 1    120
 
X    
e j
Y e j
d   2.  x n y n 2
n 

1 2
 28 8   2.094 rad
2 3
127 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs
Signals and Systems Digital Filters
22. Answer: -27 Let Fourier transform of h  n  is H  
Solution:
H    1  ae j  b e j 2
The value of 10 g  1  g 1
For input,
 10   3  3  n 
 j 
 n 
j 
x  n   c1e  2 
 c2e  2 

 30  3
For output
 27
 n 
    j n    j 
y  n   c1 H      e 2  c2 H     e  2 
23. Answer: (a)  2  2

Solution: For y  n   0

Six-point DFT of x  n    
H    0
 2 
5
X  k    x  n W6kn 
j
n 0
1  ae 2
 be j  0
X 1  x  0  .W60  x 1 w16  x  2  W62  1  aj  b  0

x  3W63  x  4  W64  x  5 W65 1  b   aj  0 … (1)

x 1  x  0  .1   x 1  x  4   W61   


H     0
 2
 x  2   x  5   W62  x  3  W63 
j
1  ae 2  be j  0
W64  W63  W61  W61 and W65 
  1  aj  b  0
  W62 .W63  W62 

24. Answer: (b)


1  b   aj  0 … (2)

Solution: From (1) and (2)


1 b  0
x  n y  n
 
h  n   1, a, b
 b  1, a  0
25. Answer: 10
Solution:
x(t)  1 1 1 1 1 1 1 1 2 2 2 2 3 3 3 3 1 1 1 1
h(n)  –1 2 –1

1 1 1 1 1 1 1 1 2 2 2 2 3 3 3 3 1 1 1 1
2 2 2 2 2 2 2 2 4 4 4 4 6 6 6 6 2 2 2 2
1 1 1 1 1 1 1 1 2 2 2 2 3 3 3 3 1 1 1 1
1 1 0 0 0 0 0 0 1 1 0 0 1 1 0 0 2 2 0 0 1 1

The number of non-zero output samples is 10.

128 Prepared by: MY (B.E.) (Hons.) BITS Pilani OHM Institute – Hyderabad GATE|ESE|PSUs

You might also like