Scribd
Upload
30 views4 pages

End Examination November 2023: Modern Signal Processing (AVDGII)

Uploaded by

Abhishek K
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
30 views4 pages

End Examination November 2023: Modern Signal Processing (AVDGII)

Uploaded by

Abhishek K
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/ 4

Indian Institute of Space Science and rrcchnology

Modern Signal Processing (AVDGII)


Deportment of Avionics
M.Tech: Digital Signal Processing/ VLSI and Microsystems/ RF and Microwave
Engineering
End Examination November 2023
Date: 29th November 2023; Marks: 100; Time: 9.30 am to 12.30 arn

l. (i) A stable LTI system is given by H(z) ) Factorize H (z) as


(1-0.5z-
a combination of
c.(a)H (z) minimum phase and all pass.
(b) H (z) minimum phase and linear phase system. [5 Marks]
1+0.9z¯1
(ii) A causal LTI system has a system function H (z) —
1-1.5z-

(a) Write the differenceequation with satisfying input output relation and plot
pole zero diagram and indicate ROC of the system function.
(b) State whether thee following are true or false with justification (1) The system
is stable (2) The system has a stable and causal inverse. [5 Marks]
(iii) A stable system H (z) has pole zero diagram shown in Figure.1. It can be
represented as the cascade of stable minimum phase system Hmin(z) and a stable
all-pass Hap(z). Determine 11min(z)and Hap(z) and draw their corresponding
pole zero plot. [5 Marks]

fin

1 3

Figure 1: 1

2. (i) Consider a uniform DFT analysis filter bank with two polyphase components
Eo 11+ 15z-l and El ==10+ 14z¯l . Find the expression for Tlk(z) for k 0, 1.
[4 Marks]
(ii) Let Jlo(z) 1 (IZ¯I + bZ¯2 + (IZ¯3 -+-Z and o(—z). Find the
polyphase components Eo(z) and El (z). Sketch the QNIF analysis bank in the
form of a uniform DFT bank. [5 Marks]

3. (i) Design a single stage and two stage interpolator to meet the followingspeci-
fications:
Passband: 0
11%', F 90; Transition band:
Input Snjnplingrate:
10 (5 = [5 Marks]
(10 shown in Figure.2, Let be
(ii) Consider the sequence :r(n) which X (c)w) as =
rate of to Fg
obtained at sa.mplingrate of Fg. Convert the sampling
spectrum of the output signal. [5
using multirate building blocks and plot the
rksl
varying and linear or time
(iii) In a multi-rate system the upsampler is a time
ariant and non linear system. Justify your answer with proof. [3 Marks]
¯l
(iv) Given an interpolation filter H (z) = 0.25+0.4z
filter for L=4. [3
0.6z-5 . Draw the block diagram for interpolation polyphase
marks]

X(ejO)

Figure 2: 2

4. (i) Use Haar wavelet family to expand the signal shown in Figure.3. (a) Use only
scaling function —k) (b) Use scaling functions and wavelet function 4(t),
V(2t —k) and V'(t —k) (c) What is the relation between scaling functions and
velet functions [4 Marks]
ii How is discrete wavelets obtained from dyadic sampling? [3 Marks]
lii) Use Haar waveletfamily to expand the signal x(t) —sin(2Tt) for 0 t 1
with scaling function and wavelets V(2t —k). [4 Marks]

0.6 ' 08 0.9

Figure 3: 3

5. (i) An AR(2) process is defined by the differenceequation x(n) = -- l)


0.6T(n 2) + w(n) where w(n) is a white noise process with variance Use
Yule walkerequation to solve for the valuesof the autocorrelation Ixx(l),
and 7$$(2). [5 marks]
(ii) DeteriJ1incthe ilnpulse response of the filter which generates a random process

Page 2
having a spectrtllii of the forin = by filtering unit variance
white noise. [5 Mark81
6. (i) adaptive filter is shown in Figure.4, The system C(z) is characterized by
the s.vst,emfunction C(z) =
1+0.92 Determine the optimum coefficientsof the
adaptive FIR filter 13(z)= bo+ biz-I that minimize MSE.The additive noise is
Xvithvariance o 2 0.1. [Assume :r(n) is white with variance [5 Marks]

Adapuve
FIR
filter

Figure 4: 4

(ii) Estimate a signal d(n) from the noisy observation d(n) + v(n), where
v(n) is unit variance white noise that is uncorrelated with d(n). The signal d(n)
is AR(I) process given by d(n) —0.8d(n —1) + w(n), w(n) is white noise with
variance a 2 — 0.36 and rd(k) — (0.8)lkl. Determine the optimal causal Wiener
filter. [5 Marks]
(iii) Consider three step predictor using first order filter given by W(z) = w(0) +
w(l)z ¯l . If the value of autocorrelation rx(k) for kags k — 0 to k = 4 are
[1.0,o, 0.1, -0.2, -0.9] T, solve the Wiener Hopf equation and find optimum
three step predictor. [3 Marks]
(iv) For high order filters the computational requirements of BLS—islargeom-
pared to LMS. For what reasons do you prefer RLS. What is the cost function of
RLS and how do you minimizeit? 4 mar s
7. (i) Determine the frequency and power of a single real sinusoid in white noise.

The signal and noise correlation function is given by [5 marks]


o Iml> 2
(ii) The signal y(n) consists of complex exponentials in white noise. Its autocor-
2 -1
relation matrix is given by j 2 —j . Use MUSIC algorithm to determine
2
the frequencies of the exponentials and their power levels. [6 Marks]
8. (i) Bartlett method is used to estimate the power spectral density of a process
sequence N = 1000. What is the minimum length M required to have a frequency
resolution 0.005. Why is it not desired to increase the length beyond the mini-
mum length? What are the differencesbetween parametric and non parametric
methods. [4 Marks]
(ii) What are the changesmade in Bartlett jnethod to form the Welchmethod of

Page 3
12

(iv) What, is the Inean and variance of a averaged modified periodogram• (3

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy