Upload
thiban-raaj
View
217
Download
0
Embed Size (px)
Citation preview
8/9/2019 Solution for Q1 & 3 of Final Sem1 2010 2011
1/7
SOLUTIONS
COLLEGE OF ENGINEERING
PUTRAJAYA CAMPUS
FINAL EXAMINATION
SEMESTER I 2010/2011
PROGRAMME : Bachelor of Electrical & Electronic Engineering /
Bachelor of Electrical Power Engineering
SUBJECT CODE : EEEB 363
SUBJECT : Digital Signal Processing
DATE : October 2010
TIME : (3 Hours)
INSTRUCTIONS TO CANDIDATES:
1. This paper contains SIX (6) questions in FOUR (4) pages.
2. Answer all questions.
3.
Write all answers in the answer booklet provided.
4. Write answer to each question on a new page.
TH IS QUESTION PAPER CONSISTS OF 4 PRINTED PAGES INCLUDING TH IS
COVER PAGE.
8/9/2019 Solution for Q1 & 3 of Final Sem1 2010 2011
2/7
Digital Signal Processing EEEB363, Semester 1 2010/2011
________________________________________________________________________
Page 2 of 7
QUESTION 1 [20 marks]
(a) The impulse response of a LTI discrete-time system is given by h[n]=[1 -2 1 0]
for 30 n . If the input to this system is given by x[n]=[2 -1 -1 2] for 30 n ,
determine the output sequence of this system via linear convolution.
[6 marks]
(b) Determine even part and odd part of x[n].
[4 marks]
(c) Determine the total energy and average power of y[n].
]4 marks]
(d) What is the frequency response of the above system?
[2 marks]
(e)
Is the above system causal and stable? Explain and give your reasons.[4 marks]
ANSWER. (a)
Using Tabular Convolution to get at y[n]:-x[n] 2 -1 -1 2
h[n] 1 -2 1 0
2 -1 -1 2
-4 2 2 -4 0
2 -1 -1 2
0 0 0 0y[n] 2 -5 3 3 -5 2 0
8/9/2019 Solution for Q1 & 3 of Final Sem1 2010 2011
3/7
Digital Signal Processing EEEB363, Semester 1 2010/2011
________________________________________________________________________
Page 3 of 7
(b) Writing x[n] as train of delayed unit impulses:-
],3[2]2[]1[][2][ nnnnn x
Even part and odd part of x[n].
marks][4
1] 0.5- 0.5- 0 0.5 0.5 [-1
]3[]2[21]1[
21]3[]2[
21]1[
21
])3[2]2[]1[][2]3[2]2[]1[][2(2
1][
])[][(2
1][
1] 0.5- 0.5- 2 0.5- 0.5- [1
]3[]2[2
1]1[
2
1]3[]2[
2
1]1[
2
1][2
])3[2]2[]1[][2]3[2]2[]1[][2(2
1][
])[][(2
1][
nnnnnn
nnnnnnnnn x
n xn xn xand
nnnnnnn
nnnnnnnnn x
n xn xn x
od
od
ev
ev
(c) The total energy and average power of y[n].
marks][4
0P(finite),76ESince
intervalfiniteoverenergyis|][|Ewhere
E12
1Lim
|][|12
1LimP
-:sequenceAperiodicof PowerAverage
762)5(33)5(2|][|E
-:lengthinfiniteoverEnergyTotal
yK y,
2
,
,k
2
k
2222222
K
K n
K y
K y
K
K n
y
n
y
n y
K
n y K
n y
(d)
2 j21)H(eis systemabovetheof responsefrequencythe
,)2()1(2-(n)0] 1 2- 1[][Since
j jee
nnnh
[2 marks]
8/9/2019 Solution for Q1 & 3 of Final Sem1 2010 2011
4/7
Digital Signal Processing EEEB363, Semester 1 2010/2011
________________________________________________________________________
Page 4 of 7
(e) The above system is causal because h[n] is a right hand function.
The system is stable because summation of h[n] over an infinite duration is finite.
[4 marks]
QUESTION 2 [15 marks]
Consider the system in Figure 2.
Figure 2The input of the system is
t t t t xa 400cos3)100cos(4)(
The frequency response of the system is shown in the following figure 3.
2
2
Figure 3: Frequency Response
If the frequency of the sampling is given as f s=500 samples/second,
(a) Determine if aliasing occurs. Explain.[4 marks]
(b) Sketch the spectrums of xa(t), x[n], y[n], and y(t). Make sure you identify the
important points.
[6 marks]
(c) Write the expressions for x[n] and y(t).
[5 marks]
QUESTION 3 [15 marks]
C/D)(t x
a ][n x ][n y)(
je H D/C)(t y
( je H 1
8/9/2019 Solution for Q1 & 3 of Final Sem1 2010 2011
5/7
Digital Signal Processing EEEB363, Semester 1 2010/2011
________________________________________________________________________
Page 5 of 7
(a) Compute the DFT of a real sequences x[n]={1 1 0 1} using the radix 2, DIT-FFT
method.
[6 marks]
(b) Compute the inverse DFT of H(k)={2 1-j 0 1+j} using the flow graph of the DIT-
FFT method arrived in question 3(a) above.
[5 marks]
(c) If the input to the LTI discrete time system in question 3(b) is the real sequence x[n]
as given in question 3(a), what is the DFT of the output of this system.
[4 marks]
ANSWER. 3(a).
Flow Graph of Radix 2 , 4 Point
DIT FFT
04W 1
0
4W
x[0]=1
x[1]=1
x[2]=0
x[3]=1
1]0[0 X
1]1[0 X
2]0[1 X
0]1[1 X
0
4W 1
1
1
3]0[ X
1]1[ X
1]2[ X
1]3[ X jW 14
8/9/2019 Solution for Q1 & 3 of Final Sem1 2010 2011
6/7
Digital Signal Processing EEEB363, Semester 1 2010/2011
________________________________________________________________________
Page 6 of 7
1)(*01]1[]1[]3[
1)(*01]1[]1[]1[
11*21]0[]0[]2[
31*21]0[]0[]0[
01*11]3[]1[]1[
21*11]3[]1[]0[
11*01]2[]0[]1[
11*01]2[]0[]0[
)
2
sin()
2
cos(,1
.0,0
1,0,12
4log
2log
1
410
1
410
0
410
0
410
0
41
0
41
0
40
0
40
24/21
4
0
4
22
jW X X X
jW X X X
W X X X
W X X X
W x x X
W x x X
W x x X
W x x X
j jeeW W
r m
r N
m
j j
3(b)
], j-1 0 j1 [2=(k)H
], j+1 0 j-1 [2=H(k)
},[k]W*HX[k]
,}[k]W*H{ N
1h[n]
[k].*Xsequenceof point DFT- NcomputetoalgorithmFFTusecanWe
,W][*][
-:conjugatecomplextakingand N byside bothgMultiplyin
./...............10,W][1
][
-: bygivenIDFT
IDFT.thecomputetoalsoFFTuseCan
*
1 N
0k
kn
N
1 N
0k
*kn
N
1
0
kn*
1
0
kn-
Let
k H n Nh
IDFT EqnSynthesis N nk H N
nh
N
k
N
N
k
N
8/9/2019 Solution for Q1 & 3 of Final Sem1 2010 2011
7/7
Digital Signal Processing EEEB363, Semester 1 2010/2011
________________________________________________________________________
Page 7 of 7
Flow Graph of Radix 2 , 4 Point
DIT FFT
0
4W 1
0
4W
H*[0]=2
H*[1]=1+j
H*[2]=0
H*[3]=1-j
2]0[0 X
2]1[0 X
2]0[1 X
j X 2]1[1
0
4W 1
1
1
2]0[ X
4)(*22]1[ j j X
0]2[ X
0)1)((*22]3[ j j X
jW 14
h[n]=X*[k]/N=[2/4 4/4 0 0]=[0.5 1 0 0]
3(c) Y[k]=DFT x[n] multiply with DFT h[n] where L>=N+M-1=4+4-1=7. This isequivalent to linear convolution for y[n] in the frequency domain.
Since length of x[n] and h[n] are both equal to 4 (N=M=4), we need to pad zeros to
them before we can compute DFT of length 8 ( e.g. using FFT radix 2).
Y[k]=DFT of {1 1 0 1 0 0 0 0}.DFT of {0.5 1 0 0 0 0 0 0}
For Solution of Question 2, 4, 5, 6 refer to separate files.