Escolar Documentos
Profissional Documentos
Cultura Documentos
PART A
1.
Define Signal.
A Signal is defined as any physical quantity that varies with time, space or any other
independent variables.
2.
Define a system.
A System is a physical device (i.e., hardware) or algorithm (i.e., software) that performs an
operation on the signal.
5.
6.
9.
19. Why the result of circular and linear convolution is not same?
Circular convolution contains same number of samples as that of x (n) and h (n), while in linear
convolution, number of samples in the result (N) are,
N=L+M-1 Where L= Number of samples in x (n) M=Number of samples in h (n)
20. What is meant by aliasing? How it can be avoided?
Nov2012
When the sampling frequency is less than twice of the highest frequency content of the signal,
then the aliasing is frequency domain takes place. In aliasing, the high frequencies of the signal mix
with lower frequencies and create distortion in frequency spectrum.
Part B
1. State and Explain sampling theorem.
It is a process of converting a continuous time signal to discrete time Signal. The continuous time
signal is sampled at regular interval.
The sampling interval is defined as time interval between two successive samples. It is also called
as sampling time.
Where,
=
Sampling Rate
Sampling Time
Sampling Theorem states that a band limited signal with highest frequency component (fm)
can be determined from its samples. If the sampling frequency is greater than or equal to twice the
maximum frequency of the signal,
2
,
fs = 2f
Rearranging,
1
=
=
If frequency lies between f to +f
Substitute equation 7 in 9
1
1
567
7 8 ,
= 1 2 3 4
+
Reconstruction of x(t):
x(t) can be represented from 10 by putting fs = 2f.
1
567
7 8
=
1 2 34
2
2
)1
n
xt = $ ; x < > e() ? e()* df
2f
2f
-
7
n 1
xt = x < > $ e()<*)> df
2f 2f
-
7
-
Since,
sin DE =
FG HE
HE
sin 2ft n
= sin C2ft n
2ft n
n
xt = x < > sin C2ft n ,
2f
-
< n <
O
M
Then,
O
M
Proof:
By definition of Z transform,
Xz = xnz
= z - + z
=
z -
O
M
Q OR S
+ z
Proof:
zxn = xnz
Put, n k = m
n=m+k
When = , V =
= , V =
V WXG4F XYV Y +
= xmz [ z ]
]
zxn k = z
Xz
O
M
O
Proof:
zxn = xnz
Put, n = m
m = n
When = , V =
= , V =
V WXG4F XYV Y +
zxn = xmz [
= xmz [
zxn = Xz
O
M
R
< >
Proof:
= xna z
O
M
Proof:
R ab R
a
c
c
Xz =
; xnz - ?
cR
cR
=
c
xnz -
cR
= xnnz -
= xnnz - z
c
1
Xz = nxnz cR
z
c
z Xz = nxnz cR
O
M
O
Then,
]
]
S
-
= Qz ] Q z -]
S
-
= X z 1 Q z -] 8
Put, n k = m
n=m+k
When = , V =
= , V =
V WXG4F XYV Y +
-
O
M
Then,
O
,
=
,
1
1
f R 2 3 R cR
2He
R
O
Proof:
Then,
b,
If O
b,
b,
b,
,
|b|
3. Check whether the following systems are static or dynamic, linear or non-linear, time variant
or invariant, causal or non-causal, stable or unstable.
Nov/Dec 2013
i.
y(n) = Cos[x(n)]
ii.
y(n) = x(-n+2)
iii. y(n) = x(2n)
iv.
y(n) = x(n). coslj (n)
Solution:
A) static: depends on present input state, dynamic:- depends on past and future input state
B) causal:- depends on present and past inputs, non-causal:- depends on future input
C) Linear:- satisfies the super position principle
+
= +
Otherwise non-linear
D) Time invariant: system do not change with time
m, Q = m Q
E) Stable: it produces a bounded output sequence for every bounded input sequence
Unstable: it produces a unbounded(infinite) output sequence for every bounded input sequence
i.
ii.
iii.
iv.
static: depends on present input state, dynamic:- depends on past and future input state
causal:- depends on present and past inputs, non-causal:- depends on future input
Linear:- satisfies the super position principle
+
= +
Otherwise non-linear
iv.
v.
m, Q m Q
Stable: it produces a bounded output sequence for every bounded input sequence
Unstable: it produces a unbounded(infinite) output sequence for every bounded input
sequence
vi.
FIR:- The system is of finite duration, IIR:- The system is of infinite duration.
MAY 2013
a) Residue Method
b) Convolution Method
Solution:
a) Residue Method:
NOV 2011
yz{|}~z =
w
}w
w u stuw
w! u }u
= w . v
,
Find the system transfer function H(z), unit sample response, magnitude Response and phase
function of the system.
Solution:I)
II)
Transfer function:
1
1 + O
2
R =
1
1 + O
4
1
1 + 2 4 5
4 5 =
1
1 + 4 4 5
Magnitude response
|| =
III)
Phase response:
l = tan
l = tan 2
1.25 + YFl
1.06 + 0.5YFl
GVGXm Y l
4 Y l
0.25 sin l
0.5 sin l
3 tan 2
3
1 + 0.25 cos l
1 + 0.5 cos l
,
R = O
,
R = O
,
ROC: |Z|<|
R =
1
1 O
two sequences are evaluated and combined to give the N point DFT. Similarly the N/2 point DFTs can be
expressed as a combination of N/4 point DFTs. This process is continued till we left with 2-point DFT.
This algorithm is called Decimation-in-time because the sequence x(n) is often splitted into smaller sub
sequences.
18. What are the differences and similarities between DIF and DIT algorithms?
Differences:
1. For DIT, the input is bit reversal while the output is in natural order, whereas for DIF, the
input is in natural order while the output is bit reversed.
2. The DIF butterfly is slightly different from the DIT butterfly, the difference being that the
complex multiplication takes place after the add-subtract operation in DIF.
Similarities: Both algorithms require same number of operations to compute the DFT. Both
algorithms can be done in place and both need to perform bit reversal at some place during the
computation.
19. What are the applications of FFT algorithms?
1. Linear filtering
2. Correlation
3. Spectrum analysis
NOV2012
Fourier transform
Gives the frequency information
for an aperiodic signal.
Continuous frequency spectrum
Part B
1. A)Find eight point DFT of the following sequence using direct method:
{1,1,1,1,1,1,0,0}
Solution:
MAY 2013
Q = 0,1, 1
j
For the given sequence N=8, by substituting k and N values in above equation, we get
X(0)=
X(1)=
X(2)=
X(3)=
X(4)=
X(5)=
X(6)=
X(7)=
6
-0.707- j1.707
1- j
0.707+j0.293
0
0.707- j0.293
1+ j
- 0.707+j1.707
MAY 2013
2. A) Compute eight point DFT of the following sequence using radix 2 Decimation in time
FFT algorithm.
MAY 2011(16 mark)/MAY 2013(8 mark)
X(n)={1,-1,-1,-1,1,1,1,-1}
Solution:Butterfly diagram should be drawn
(2 marks)
The output of each stage are given below
(each stage 2 marks)
Input of I stage
Output of I stage
Output of II stage
Output of III stage
1
2j
-1.414+j3.414
-1
2-2j
-2
-2j
1.414 - j0.586
-1
-2
-2
-2
1.414+j0.586
-1
-2
2+2j
-1
-2
-1.414 j3.414
MAY 2013
The overlap-save and overlap-add methods are used for filtering a long data sequence with an
FIR filter based on the use of DFT.
FFT algorithm can be used for computing DFT and IDFT.
(2 marks)
Comparison of Overlap-save and overlap-add method in filtering.
(4 marks)
3. Compute the FFT of the sequence = v + w p w, where N=8 using DIT
algorithm.
NOV 2012
Solution: - N=8
= v + w p
x(0)=1
x(1)=2
x(2)=5
x(3)=10
x(4)=17
x(5)=26
x(6)=37
x(7)=50
Butterfly diagram should be drawn
The output of each stage are given below
Input of I stage
Output of I stage
Output of II stage
Output of III stage
1
18
60
148
17
-16
-16+32j
-4.688+13.248j
5
42
-24
-24+32j
37
-32
-16-32j
-27.312+13.248j
2
28
88
-28
26
-24
-24+40j
-27.312-13.248j
10
60
-32
-24-32j
50
-40
-24-40j
-4.688-77.248j
4. Find DFT for {1,1,2,0,1,2,0,1} using FFT DIT butterfly algorithm.
Butterfly diagram should be drawn
The output of each stage are given below
Input of I stage
Output of I stage
1
2
1
0
2
2
0
2
1
3
2
-1
0
1
1
-1
Output of II stage
4
-2j
0
2j
4
-1+j
2
-1-j
NOV 2013
NOV 2013
1. Periodicity:
If discrete time signal is periodic, then its DFT is also periodic.
If x(n) is a discrete signal of length N, then
i.
x(n+N)=x(n)
ii.
X(k+N)=x(k)
Where x Input signal (time domain)
X DFT[x(n)] (frequency domain)
2. Linearity:
It satisfies superposition principle
Q
YG
Similarly,
Q
YG
Q
YG
By Linearity Property,
+
Q +
Q
YG
Where a,b constants
3. Symmetry Property:
If signal or sequence repeats its waveform in negative direction after N/2 number of
samples, then it is called symmetric signal or sequence.
From periodicity property,
If x(n)=x(N-n)
Then X(k)=X(N-k)
For Symmetric signal,
If
Q
YG
Then
Q
YG
V = . V
j
It is denoted by
=
Q
YG
Where m=0,1,2,..N-1
[ DGXXDYWYGY]
Q
YG
If
Then V = 4
4
YG
7. Circular Frequency Shift:
i.e. V
If
Then
Q
YG
4
Q
Q
Q V
YG
Q
YG
Q
YG
V
Q
R
=
Q
Then circular correlation is given by
V = V
j
Q
YG
Then
Q
YG
1
Q Q
YG
[ DGXXDYWYGY]
Q
YG
Then
Q
YG
1
= Q Q
j
Sj
j
Sj
1
|| = |Q|
6. Draw the flow chart for N=8 using radix 2 DIF algorithm for finding DFT coefficient. (Nov
2010)
Decimation-in-frequency (DIF) is another important radix 2 FFT algorithm. In case of
Decimation-in-Time, the input data sequence x(n) are decimated but in case of Decimation-infrequency, the data coefficients X(k) are decimated. In this algorithm, we first divide the DFT
points.
Xk = xnW-] ,
j
k = 0,1,2, , N 1
Xk =
j
xnW-]
<-\ >]
+ x 2n + 3 W
2
j
]
Xk = xnW-] + W x 2n + 3 W-]
2
j
W = 1]
Since
j
j
w
j
X2k = @xn + x 2n + 3C W-] ,
2
j
k = 0,1,2, , 2 13
2
k = 0,1,2, , 2 13
2
v
-]\
X2k + 1 = @xn x 2n + 3C W
,
2
j
k = 0,1,2, , 2 13
2
k = 0,1,2, , 2 13
2
x
k = 0,1,2, , 2 13
2
j
g n = + 2 + 3
2
Where
Where
k = 0,1,2, , 2 13
2
g n = @xn x 2n + 3C W2
Decimation in frequency stands for splitting the sequences in terms of frequency. That means we
have split output sequences into smaller subsequences. This decimation is done as follows.
First stage of decimation:
first stage of decimation as shown in fig. below.
Second stage of decimation: In the first stage of decimation we have used 4-point DFT. We
can further decimate the sequence by using 2 point DFT. The second stage of decimation is shown in
fig below.
Third stage of decimation : In the second stage of decimation we have used 2- point DFT. So
further decimation is not possible. Now we will use a butterfly structure to obtain 2-point DFT. Thus
the total flow graph of 8 point DIF-FFT is shown below.
If r=2, radix-2
= X
= 2
(1)
(2)
In radix-2,N is an even integer. Let us consider computing X(k) by separating x(n) into two
N/2 point subsequences consisting of the even-numbered points in x(n) and the odd-numbered
points in x(n).
x
-
j
j
\]
S = S
6
5< >
4
6
5
4
S = S
Where,
j
j
k = 0,1,2, , N 1
Q = x2rS
j
j
2Q + 3 = Q
2
2Q + 3 = Q
2
<S\ >
= S
k = 0,1,2, , 2 13
2
S\
2Q + 3 = 2Q + 3 + W H 2Q + 3
2
2
2
2Q + 3 Q W] Hk,
2
k = 0,1,2, , 2 13
2
8. Draw a 8 point radius 2 FFT DIT flow graphs and obtain DFT of the following sequence
x(n)={0,1,-1,0,0,2,-2,0}
Where,
l + l
2
=
l l
cos 2
cos
Q = tan @
l
l + l
C tan
2
2
Value of o = v
Substitute s=
M
M\
Given =1 db
For normalize filter F sample = 1 Khz
Calculate sk=acosQ+jbsinQ,
where Q= +
S
Calculate =-1+(1+-2)
Calculate a = p [
b = p [
] = 0.297 and
] = 1.043
.Calculate Poles
ES = +
6
S\6
, Q = 0,1,2
k=0, Ej = + =
6
k=1, E = +
6
k=2, E = +
6
S k=a cosk+jbsink
S 0=a cos0+jbsin0,
S 1=a cos1+jbsin1,
S 2=a cos2+jbsin2,
=H
3. Convert the analog filter transfer function to digital filter using Impulse Invariance methodH(S) =
{\v
{\w{\x
Ans
Using Partial Fraction
b
4. Design a digital chebyshev filter for the following specifications using bilinear transformation
MAY 2012
Given:
5. Design a digital chebeshev filter for the following specifications using Impulse in variance
transformation
Ans:
Ans:
w\v
= p.
=3.179
w\v
p p. v
p.
= p. x wp = p. v ws = p.
=1.02
s= p. v
Ha(s) =
\ \
It is a LPF s = s/ c
c = p /
1/N
=0.622 rad/sec
s = s/ 0.622
=
Using partial fraction A = 0.439 j B = -0.439 j
=
p. x
+ p. + p. x
p. x
p. x
p. x
=
v + p. + p. x
p. x p. x p. x + p. x
{|
wz|
uw
= 10j. 1 = 3
H(s) = 1/(S+1)
H(s) = 1/ (s/3.24)+1
S.S
S= S/c
c = p /
1/N
=3.24
M
M\
Nov 2011
May2013
1
1
2
3 2
3
2
2
10. State the condition for a digital filter to be causal and stable.
A digital filter is causal if its impulse response h(n) = 0 for n<0 A digital filter is stable if its
impulse response is absolutely summable,
11. What are the properties of FIR filter?
i.
FIR filter is always stable.
ii.
A realizable filter can always be obtained.
iii. FIR filter has a linear phase response.
PART B
1. Prove that an FIR filter has Linear phase if the unit sample response satisfies the condition
h(n)=h(N-1-n).Also discuss the symmetric and anti symmetric cases of FIR filter when N is
even.
NOV 2013
Case 1:- Symmetric Impulse Response with Even Length:
The frequency response of h(n) is
4
= 4 5
j
j
< >
j
j
j
5<
>
5<
>
5<
>
5
4 = 4
4
+ 4
j
j
1
5<
>
5
4 = 4
2 cos @2
3 C
2
j
1
5<
>
5
4 = 4
2 2 3 cos @2 3 lC
2
2
j
1
5<
>
5
4 = 4
cos @2 3 lC
2
j
Where
= 2 < >
4 5 = 4 5<
>
4 5
= 4 5 4 5
1
4 5 =
cos 2 3 l
2
+1
El = 2
3l
2
4 5 = 4 5
j
< >
j
j
j
j
5<
>
5<
>
5<
>
5
4 = 4
4
4
j
j
6
1
5<
> 5
5
4 2 2 3 sin @2 3 lC
4 = 4
2
2
6
1
5<
> 5
5
4 = 4
4
c cos @2 3 lC
2
4 5 = 4 5<
6
>
4 5 4 5
= 4 5 4 5
1
4 5 =
cos 2 3 l
2
El =
H
1
2
3l
2
2
2. Prove that an FIR filter has Linear phase if the unit sample response satisfies the condition
h(n)=h(N-1-n).Also discuss the symmetric and anti symmetric cases of FIR filter when N is odd.
Case 1:- Symmetric Impulse Response with Odd Length
The frequency response of h(n) is
4 5 = 4 5
j
= 4
j
5
1 5<>
+2
34
+
2
<
\
>
4 5
4 5 = 4 5 + 2
j
1 5<>
34
+ 4 5-
2
j
1
5<
>
5<
>
5<
>
5
4 = 4
4
+2
3 + 4
2
j
j
Let <
> = ,
1
5<
>
5<
>
5
4 = 4
2 4
+2
3
2
j
1
1
5<
>
5
4 = 4
2. 2
3 cos l + 2
3
2
2
1
1
5<
>
5
4 = 4
2. 2
3 cos l + 2
3
2
2
4 5 = 4 5<
>
> 3
4 5 = 4 5<
cos l
j
>
4 5
= 4 5 4 5
4 5 = cos l
j
1
El = 2
3l
2
1
3=0
2
= 4 5
j
= 4
j
5
1 5<>
+2
34
+
2
j
j
<
\
>
4 5
5<
>
5<
>
5<
>
5
4 = 4
4
4
j
j
1
5<
>
5
e 2 sin @2
4 = 4
3 C
2
j
6
1
5<
> 5
5
4 2. 2
4 = 4
3 sin l
2
> 3
4 5 = 4 5<
6
>
4 5
cos l
j
4 5 = 4
5<
>
4 5 4 5 = 4 5 4 5
4 5 = sin l
El =
H
1
2
3l
2
2
3H
| |
Hd 4el = 1 for 4 l 4
0 Y4XGF4
Find the values of h(n) for N=7. Find the realizable filter transfer function and magnitude
function of H4el
Soln. Step 1. Draw the ideal desired frequency response of bandpass filter.
Form the desired frequency response, we can find that the given response is symmetric N odd
Step 2. To find
|l| H
4
5
H 4 =
H
0 YX |l| <
4
1 for
1 for l
2
2
Hd 4el =
H
0 YX l H
2
Find the values of h(n) for N=1. Find the realizable filter transfer function and magnitude
function of H4el
From the frequency response, we can find that the given response is a symmetrical
N odd response.
Step 2. To find
In general,
1 for 0 |l|
6
Hd 4el =
H
0 YX |l| H
6
Use 10 tap filter and obtain the impulse response of the desired filter
Overflow in addition
Overflow in addition of two or more binary numbers occurs when the sum exceeds the word
size available in the digital implementation of the system.
Limit cycles
Since quantization inherent in the finite precision arithmetic operations render the system
nonlinear, in recursive system these nonlinearities often cause periodic oscillation to occur in the
output, even when input sequence is zero or some nonzero value. Such an oscillation in recursive
systems are called limit cycles.
As explained in above paragraphs finite word length affects LTI system in many ways. We
have concentrated on effects due to coefficient quantization on filter response and in that also on IIR
filters. Later we have given brief overview of effects of coefficient quantization in FIR system for the
sack of completeness.
8. Explain about limit Cycle oscillations
When a stable IIR digital filter is excited by finite input sequence that is constant the output will
ideally zero. However non linearity's due to finite precision arithmetic operation often cause periodic
oscillation to occur in the output. Such oscillations in recursive systems are called zero input limit
cycles Consider a first order IIR filter the difference equation)
y(n) =x(n) +a y(n-1)
Let us assume a=1/2 and the data register length is 3 bits plus a sign bit. If the input is
n
0
1
2
3
4
5
x(n)
0.875
0
0
0
0
0
0.875 YX = 0
x(n) =
0 Y4XGF4
y(n-1)
0.0
7/8
1/2
1/4
1/8
1/8
ay(n-1)
0.0
7/16
1/4
1/8
1/16
1/16
Q[ay(n-1]
0.000
0.110
0.010
0.001
0.001
0.001
Y(n)
7/8
1/2
1/4
1/8
1/8
1/8
Note that beyond n=4 the value of ay(n-1) is 1/16 and in binary 0.000100 an 0.000100 which
whenrounded gives 1.001 and 0.001 exhibiting oscillatory.
Dead ban d will b calculated by using y(n-1)
w
v
v
w||
UNIT V APPLICATIONS
PART A
1. Define multi rate digital signal processing?
NOV 2012
Digital signal processing system handles processing at multiple sampling rates and then it is
called multi-rate signal processing.
2. What are the two techniques of sampling rate conversion?
i. D/A conversion and resampling at required rate.
ii. Sampling rate conversion in digital domain (multi-rate processing)
NOV2012
MAY 2013
MAY2012
6. Define speech compression and decompression.
Speech analysis by a vocoder becomes the compression and synthesis by a vocoder becomes
decompression.
Vocoder extracts the spectral envelope of speech and information regarding voicing and pitch.
This data is coded and transmitted. The synthesizer generates speech from the received data.
7.
NOV 2013
NOV 2011
i.
ii.
iii.
Noise cancellation
Signal prediction
Adaptive feedback cancellation
Echo cancellation
13. List various special audio effects that can be implemented digitally.
Echo effect
Reverberation
Chorus effect
Phasing effect
Flanging
MAY 2013
PART B
1. Discuss sub band coding process in detail.
MAY 2013
Digitized speech signals may be transmitted over a limited bandwidth channel or it can be
stored. Reducing the size of the signal before transmission or storage is known as speech
compression. The signal compression can be achieved by sub band coding. This method is making
use of the non uniform distribution of signal energy in the frequency component.
Transmitter:
The signal is spit into many narrow band signals which occupy continuous frequency
bands using analysis filter bank.
Down sampling these signals gives sub band signals.
Then compress it using encoders and the compressed signal is multiplexed and transmtted.
Receiver:
The received signal is demultiplexed, decoded, up sampled and then passed through a
synthesis filter.
The output of synthesis filter bank are combined to get the original uncompressed signal.
NOV2013/2012
The noise cancellers are used to eliminate intense background noise. This configuration is
applied in mobile phones and radio communications, because in some situations these devices are
used in high-noise environments. Figure 6 shows an adaptive noise cancellation system.
The canceller employs a directional microphone to measure and estimate the instantaneous
amplitude of ambient noise r(n), and another microphone is used to take the speech signal which is
contaminated with noise d(n) + r(n). The ambient noise is processed by the adaptive filter to make it
equal to the noise contaminating the speech signal, and then is subtracted to cancel out the noise in
the desired signal. In order to be effectively the ambient noise must be highly correlated with the
noise components in the speech signal, if there is no access to the instantaneous value of the
contaminating signal, the noise cannot be cancelled out, but it can be reduced using the statistics of
the signal and the noise process.
6. Explain any one application using multirate processing of signals.
Some application of Multirate signal processing are
1. Sampling rate conversion
2. Design of phase shifters
3. Interfacing of digital systems with different sampling rate.
NOV 2010