EEE 391

Basics of Signals and Systems

Midterm Exam
9 November 2020, Monday
closed book and notes
no calculators
Given Time: 100 min
Instructor: Billur Barshan
Last Name : .....................................................
First Name : .....................................................
ID number : .....................................................

Signature : .....................................................

exam total points

part points received
Q1 28
Q2 25
Q3 21
Q4 26
total 100

Allocation of points:
1) 28 pts: a) 4 pts b) 5 pts c) 5 pts d) 6 pts e) 8 pts
2) 25 pts: a) 10 pts b) 15 pts
3) 21 pts
4) 26 pts: a) 8 pts b) 7 pts c) 4 pts d) 7 pts

Attention:

• Read all the questions carefully and show your work for full or partial credit.
Justify all your answers.
• Do not take the staple out or tear off any sheets.
 1) The signal x(t) is given as follows:
 π
2
x(t) = −2 + 2 sin (1.2πt + 2) − cos 15t − − 4e−j(10t+1)
2
where t has the units of seconds.

a) Is the signal real? Justify.

The signal is not real because the complex conjugate of the last term does not appear
in the expression given for x(t).

b) What is the average (d.c.) value of the signal?

1−cos(2θ)
Rewriting x(t) using the given formula sin2 θ = 2 :
[1 − cos (2.4πt + 4)]  π
x(t) = −2 + 2 − cos 15t − − 4e−j(10t+1)
2 2
 π
= −2 + [1 − cos (2.4πt + 4)] − cos 15t − − 4e−j(10t+1)
2
 π
= −1 − cos (2.4πt + 4) − cos 15t − − 4e−j(10t+1)
2
The constant part (with zero freq.) is the average (d.c.) value of x(t) which is −1.
c) Is the signal periodic or not? If periodic, what is the fundamental period? If not,
why? If periodic, identify all the harmonics that exist in the frequency spectrum.

The signal is not periodic because the three frequency components 2.4π, 15, −10 rad/s
are not harmonically related. It is not possible to calculate their greatest common divisor.

d) What is the minimum sampling frequency that should be used to sample the above
signal without aliasing any of its components? You may take the number π as 3.15.

If you compare the three radian frequencies (2.4π = 7.56, 15, −10) rad/s, the largest one
15
(in magnitude) is ωmax = 15 rad/s which corresponds to fmax = 2π Hz. The corresponding
15
Nyquist rate is 2ωmax = 30 rad/s or 2fmax = π Hz. To avoid aliasing any of the
components, the minimum sampling frequency is the same as the stated Nyquist rate.

e) If a sampling period of Ts = 0.25 sec is used, is this a case of oversampling or

undersampling for x(t)? If aliasing and/or folding occurs, indicate which components are
aliased and/or folded. Please do this by circling and labelling the terms of x(t) given
above.

The given sampling period corresponds to fs = T1s = 4 Hz. Since fmax < fs < 2fmax ,
this is a case of undersampling where aliasing with folding occurs. Aliased and folded
term is emboxed in red above.
 2) Some of the Fourier series coefficients of a real and periodic signal x(t) are given as:
1 π √
a1 = a2 = 2ej 3 a4 = j 3 − 1 where j = −1
j
It is known that the average value of the signal is –2. The Fourier series coefficients for
|k| =
6 0, 1, 2, 4 are all zero.

a) Write all the Fourier series coefficients of x(t) for −4 ≤ k ≤ 4 in polar form.

Since the signal is stated to be real, FS coefficients must exist in complex conjugate
pairs in the FS representation.
1 π 3π π 3π
a1 = = −j = e−j 2 or ej 2 −→ a−1 = a∗1 = ej 2 or e−j 2
j
π 2π π 2π
a2 = 2ej 3 or 2e−j 3 (already in polar form) −→ a−2 = a∗2 = 2e−j 3 or 2ej 3
3
√ j 5π √ −j 3π ∗
√ −j 5π √ j 3π
a4 = j − 1 = −j − 1 = −(j + 1) = 2e or 2e 4 4 −→ a−4 = a4 = 2e 4 or 2e 4
The given d.c. value of –2 corresponds to ao = −2.
Summarizing what we have for −4 ≤ k ≤ 4:
 √ −j 5π

 a −4 = 2e 4 for k = −4
= −3



 a−3 = 0 for k
π
−j



 a−2 = 2e 3 for k = −2
π
j

 a−1 = e 2 for k = −1


ak = a0 = −2 for k =0
π
−j
a1 = e 2 for k =1



 π
j
a2 = 2e 3 for k =2







 a3 = √0 for k =3
j 5π

a4 = 2e 4 for k =4


b) Express the signal x(t) in the form

∞
X
x(t) = Ao + Ak cos(ωk t + φk )
k=1

where ωk = kωo and ωo = 4 rad/s.

x(t) = a−4 e−j4ωo t + a−2 e−j2ωo t + a−1 e−jωo t + ao + a1 ejωo t + a2 ej2ωo t + a4 ej4ωo t
√ −j 5π −j4ω t π π π π √ 5π
= 2e 4 e o
+ 2e−j 3 e−j2ωo t + ej 2 e−jωo t − 2 + e−j 2 ejωo t + 2ej 3 ej2ωo t + 2ej 4 ej4ωo t
√ −j(4ω t+ 5π ) π π π
t+ π3 )
√ 5π
= 2e o 4 + 2e−j(2ωo t+ 3 ) + e−j(ωo t− 2 ) − 2 + ej(ωo t−2 ) + 2ej(2ωo + 2ej(4ωo t+ 4 )
 π  π √ 5π
= −2 + 2 cos ωo t − + 4 cos 2ωo t + + 2 2 cos 4ωo t +
2 3 4
√
 
 π   π  5π
= −2 + 2 cos 4t − + 4 cos 8t + + 2 2 cos 16t +
2 3 4
 3) Given the signal x(t) below, sketch and clearly label y(t) = − 12 x(t + 2) below.

y(t) = − 21 x(t + 2)

–7 –6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 7
t

− 12

Note that y(t) is advanced by two units (shift to the left by two units on the time axis)
and scaled in amplitude by − 12 . Amplitude scaling by a negative value means that the
signal will be flipped over the horizontal axis.
 4) A causal finite impulse response (FIR) filter produces the following output when
δ[n + 2] is given to it as input:
y[n] 2
•
1
•
··· ···
• • • • n
–3 –2 –1 0 1 2 3 4

•
–2
•
–3
a) Find the impulse response h[n] of the filter and plot it below.
δ[n + 2] −7 → h[n + 2] = y[n]
δ[n] 7−→ h[n] = y[n − 2]
From the given information, we understand that y[n] = h[n + 2]. Therefore, we need to
delay y[n] by two units to get h[n]. After shifting by two units to the right, we get:
2
h[n] •
1
•
··· ···
• • • • • • • • • • • n
–7 –6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 7

•
–2
•
–3

b) Identify and list all the filter coefficients bk as a set.

n o
bk = 0, 1, 0, 2, −2, −3

c) What is the order (M ) and what is the length (L) of the filter?

M =5 L=6

d) Write the difference equation of the filter.

General form for the 5th-order causal FIR filter is:

y[n] = bo x[n] + b1 x[n − 1] + b2 x[n − 2] + b3 x[n − 3] + b4 x[n − 4] + b5 x[n − 5]
Substituting the filter coefficients from part b):
y[n] = x[n − 1] + 2x[n − 3] − 2x[n − 4] − 3x[n − 5]