EE202 Signals and Systems Tuesday, May.
13, 2025
KAIST, Spring 2025 Handout #
Due: Friday, May. 30, 11:59 pm
Problem Set 4
Announcement:
The grade of this problem set will be one of {0, 1, 2, 3, 4} based on rough estimate of the
performance (the amount of the problems you solved).
4: above 75%
3: 50% – 75%
2: 25% – 50%
1: 0% – 25%
Problem 4.1
Problem 4.10, 4.13, 4.15, 4.19 from Oppenheim and Willsky.
Problem 4.2
Problem 4.40, 4.45 from Oppenheim and Willsky.
Problem 4.3
Problem 5.6, 5.21(k) from Oppenheim and Willsky.
Problem 4.4 (Systems with Real Impulse Responses)
Suppose that we have an LTI system whose impulse response h(t) is real, i.e., h(t) = h∗ (t).
Show that when x(t) = A cos(ω0 t + θ) we have y(t) = A|H(jω0 )| cos(ω0 t + θ + ∠H(jω0 )),
where A, θ are real numbers.
Remark: A linear circuit system (e.g. circuit that consists of R,L,C) will have a real
impulse response as its step response s(t) will be real (physical variables such as voltage
and current should be real). Hence the above relationship is applicable to any linear
circuit system.
Problem 4.5
In
R t this problem, you will derive Fourier transform of unit step function u(t) and
−∞
x(τ )dτ = x(t) ∗ u(t). Although in this case Fourier transform does not converge,
but we can define its Fourier transform if we allow impulse function in frequency domain.
Let sgn(t) = 2u(t) − 1, which is -1 for t R< 0 and 1 for t ≥ 0. Note that the analysis equa-
R0 ∞
tion for sgn(t) becomes −∞ −e−jωt dt + 0 e−jωt dt, which does not converge. Instead, we
introduce a damping function e−σ|T | and compute Fourier transform of its damped version
sgn(t)e−σ|t| for σ > 0, then take limit, F{sgn(t)} = limσ→0 F{sgn(t)e−σ|t| }.
1
(a) Compute Fourier transform of sgn(t)e−σ|t| for σ > 0. Show that
limσ→0 F{sgn(t)e−σ|t| } = jω
2
.
(b) Using u(t) = 12 (sgn(t) + 1), show that F{u(t)} = 1
jω
+ πδ(ω).
Rt 1
(c) Using the convolution property show that F{ −∞
x(τ )dτ } = jω
X(jω) + πX(0)δ(ω).
Problem 4.6
(a) Determine the numerical value of
Z ∞ 3
sin t
A= dt.
−∞ πt
(Hint: Parseval’s relation will not be useful.)
(b) Determine the numerical value of
∞ n
X 1
B= n .
n=0
3
(c) Consider a discrete-time LTI system with impulse response
n
1
h[n] = u[n].
3
Determine the response to the input x[n] = (−1)n .
Problem 4.7
Consider a series (i.e. cascade) interconnection of two causal LTI systems S1 and S2
described by the following difference equations:
S1 : causal LTI,
1
w[n] = w[n − 1] + x[n]
2
S2 : causal LTI,
1
y[n] = y[n − 1] + w[n]
4
where S1 is of frequency response H1 (e ) and S2 is of frequency response H2 (ejω ).
jω
(a) Determine the frequency response H1 (ejω ) and the frequency response H2 (ejω ). De-
termine the frequency response H(ejω ) of the composite system.
(b) Determine Y (ejω ) when x[n] = ( 31 )n u[n]. Determine y[n] by performing inverse
Fourier transform of Y (ejω ).
