0001.

pdf 0032767001

Questions
1 2 3 4 5 6 7 8 9 10
1 1 1 1 1 1 1

2 2 2 2 2 2 2
Surname, First name
3 3 3 3 3 3 3

4 4 4 4 4 4 4

5 5 5 5 5 5 5

6 6 6 6 6 6 6
Signal Analysis (WB2235)
Exam WB2235 Signaalanalyse 7 7 7 7 7 7 7

28 June 2019 13:30 - 16:30 8 8 8 8 8 8 8

9 9 9 9 9 9 9

0 0 0 0 0 0 0

• On this cover page:

– Write your name in the provided space;

– Write your ’studentnummer’ in the boxes in the top-right corner of this page;
– Color in the corresponding digit circles below your student ID to enable automatic recognition.

• This exam consists of 10 problems. The number of points that can be achieved is indicated for
each problem. The maximum number of achievable points is 35.
• Use the provided answer boxes to write down your answers. Draft paper sheets are available as
needed, but will be ignored for grading. Only answers provided on these exam paper sheets will be
graded.
• If you make a mistake in the answer spaces, stickers to cover wrong answers are available from the
invigilators as needed.
• You may answer either in English or in Dutch.
• During the exam, you are allowed to use a non-graphical calculator and the provided formula sheet.
The use of any other materials is not permitted.
• When asked to justify an answer or determine something, make sure to write down all major steps
of your argument and/or derivation. Correct answers with missing (or grossly incorrect) arguments
and/or derivations will not be given any points unless something else is stated in the problem.
• When you are asked to sketch something, it is not necessary to write down any calculations. Make
sure that all significant characteristics such as, for example, jumps and maxima are indicated on
both axes. Always make clear which quantity is indicated on which axis.
• Simplify results as much as possible and make clear what your final answer/conclusion is.

1 / 20
 0001.pdf 0032767002

Problem 1: Elementary Properties of Signals and Systems (6 points: 2+2+2)

x

3.5
3
2.5 /2

x[n]
|x[n]|

2 0
1.5
1 - /2
0.5
-
0
-2 -1 0 1 2 -2 -1 0 1 2
n n

2p 1a Consider the discrete-time signal x[n], whose average power is 6, and whose magnitude and phase
are depicted above for time points n = −2, ..., 0, 2. At n = 1, only the phase is given. The signal is
periodic with period 5. Sketch the real part of x[n] for n = −2, ..., 2. Label all axes appropriately.

2 / 20
 0001.pdf 0032767003

2p 1b Consider the discrete-time system that maps the input x [n] to the output

x [n]
y [n] = ,
1 + u [n]

where u [n] denotes the unit step. Determine if the system is time-invariant. Justify your answer.

3 / 20
 0001.pdf 0032767004

x
2.5

1.5

0.5
Im(z)
0

-0.5

-1

-1.5
Poles
-2 Zeros
-2.5
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
Re(z)
x

2p 1c A causal linear time-invariant discrete-time system has the pole-zero plot shown above. Determine if
the system is stable. Justify your answer.

4 / 20
 0001.pdf 0032767005

Problem 2: Discrete-time Fourier Transform (3 points)

3p 2 Compute the inverse discrete-time Fourier transform x [n] of

2πδ (Ω − 1) e−jΩ
X(ejΩ ) = + , |Ω| < π,
ejΩ 2 − e−jΩ
where δ (Ω) denotes the Dirac delta function. Provide all steps of your calculation.

5 / 20
 0001.pdf 0032767006

Problem 3: Bode Plots (3 points)

3p 3 Sketch the Bode plot (both magnitude and phase) of the frequency response

(jω + 1)
H(jω) =
(jω + 100)2

for 10−2 ¬ ω ¬ 104 . Hint: It is sufficient to draw the straight-line approximation in this problem.

6 / 20
 0001.pdf 0032767007

Problem 4: Laplace Transform (4 points: 3+1)

3p 4a Consider the continuous-time signal

x (t) = e−4t (u (t) − u (t − 1)) ,

where u (t) denotes the unit step. Determine the Laplace transform of x (t). Provide all steps of your
calculation. Hint: Do not forget to justify the ROC.

7 / 20
 0001.pdf 0032767008

1p 4b Consider a causal continuous-time linear time-invariant system with the rational transfer function
s−1 s+2
H(s) = .
s2 + 4 s + 1
Make a sketch of the ROC for −3 ¬ Re(s) ¬ 3 and −3 ¬ Im(s) ¬ 3. Label all axes appropriately.

8 / 20
 0001.pdf 0032767009

Problem 5: Modulation and Demodulation (3 points)

x
5(#) 2(#)
-(#) 1, ( < 3
Hilbert
% !(#)
0, ( > 3
transform
ℋ 7(
Low-pass
sin(5#) cos(5#) filter

Modulator Demodulator
x
Consider the modulation-demodulation scheme depicted above. Furthermore, assume that the input x(t)
has the Fourier transform X(jω) (
2, |ω| ¬ 1.5
X(jω) = .
0, |ω| > 1.5

3p 5 Sketch the Fourier transforms Y (jω) and Z(jω) of the corresponding signals y(t) and z(t) for
−7 ¬ ω ¬ 7. Label all axes appropriately. Hint: H(jω) is provided on the formula sheet.

9 / 20
 0001.pdf 0032767010

Problem 6: Sampling (4 points: 3+1)

x

C/D
!" ($) conversion
!& ' = !" ')

)
x
Consider the continuous-time to discrete-time converter with sampling interval T . The Fourier transform
Xc (jω) of the continuous-time input signal xc (t) is
(
π(1 − |ω|), |ω| ¬ 2
Xc (jω) = .
0, |ω| > 2

3p 6a Sketch the discrete-time Fourier transform Xd (ejΩ ) of the output xd [n] for −π ¬ Ω ¬ π that results
from the input signal xc (t) and a sampling interval T = 2π
3 . Label all axes appropriately.

10 / 20
 0001.pdf 0032767011

1p 6b For the same system and input as in Problem 6a, determine all positive values of T for which aliasing
occurs. Justify your answer using the sampling theorem.

11 / 20
 0001.pdf 0032767012

Problem 7: Discrete-time Filter Design (3 points: 2+1)

2p 7a Consider a desired discrete-time filter transfer function

−z −1
Hd (z) = , |z| > 1.
(z −1 + 1)2

A discrete-time FIR filter can be designed by windowing the corresponding impulse response hd [n]
with a Hann window with M = 3. The resulting linear time-invariant filter has an impulse reponse h[n].
Determine h[n] for n = −1, 0, 1, 2, 3, 4. Note: You may use that cos (2π/3) = −0.5 and cos (4π/3) =
−0.5.

12 / 20
 0001.pdf 0032767013

1p 7b State one advantage and one disadvantage of using a Bartlett window compared to using a
rectangular window.

13 / 20
 0001.pdf 0032767014

Problem 8: Elementary Properties of Random Processes (4 points: 1+3)

1p 8a Let X[n] be an independent and identically distributed (i.i.d.) random process where each X[n]
follows a normal distribution with zero mean and variance σ 2 = 3. We consider the random process

X[n]
Y [n] = n √ + 1.
2
Determine if Y [n] is wide-sense stationary. Justify your answer.

14 / 20
 0001.pdf 0032767015

3p 8b Consider two random processes W [n] and Z[n], where W [n] is a wide-sense stationary (WSS)
process with mean µW [n] = 1 and auto-correlation

RW W [n + m, n] = δ[m − 2] + 3δ[m] + δ[m + 2].

When W [n] is provided as input to a linear time-invariant discrete-time system with impulse response

h[m] = 0.25δ[m − 1] + 1.5δ[m] + 0.25δ[m + 1],

the output is Z[n]. Determine the cross-covariance CZW [n + m, n] for all m and n. Hint: The process
Z[n] has a mean µZ [n] = 2 and is jointly wide-sense stationary with W [n].

15 / 20
0001.pdf 0032767016

16 / 20
 0001.pdf 0032767017

Problem 9: Estimation of Spectral Densities (2 points)

Consider the following realization x[n] of an unknown random process X[n]:

n 0 1 2 3 4 5 6 7 8 9 10 11
.
x[n] 4 0 −1 0 0 −2 3 0 0 0 2 0

The random process X[n] is second-order ergodic.

2p 9 Compute the averaged periodogram with K = 4 segments for the given samples. Simplify as much
as possible.

17 / 20
0001.pdf 0032767018

18 / 20
 0001.pdf 0032767019

Problem 10: Wiener Filter (3 points)

x

Wiener
$# "! #
Filter
x
Consider a random wide-sense stationary (WSS) process X[n] with zero mean and auto-correlation

RXX [m] = 0.5δ[m − 2] + δ[m] + 0.5δ[m + 2].

The cross-correlation between X[n] and a second zero-mean random WSS process Y [n] is given by

RY X [m] = −2δ[m − 2] + 4δ[m − 1] − 3δ[m] + 4δ[m + 1] − 2δ[m + 2].

The processes X[n] and Y [n] are jointly WSS.

3p 10 Determine the impulse response h[n] of the finite impulse response (FIR) Wiener filter of length
N = 3 that estimates Y [n] from X[n] for all n ∈ Z.

19 / 20
0001.pdf 0032767020

20 / 20