Signals and Systems:Solution for Assignment 4

Problem 6.5. [10pt] Consider a continuous-time ideal bandpass filter whose frequency response is
(
1, ωc ≤ |ω| ≤ 3ωc
H(jω) = .
0, otherwise

If h(t) is the impulse response of this filter, determine a function g(t) such that
 
sin ωc t
h(t) = g(t).
πt

sin ωc t
Sol. Let h1 (t) = πt . Its Fourier transform is computed as
(
1 |ω| < ωc ,
H1 (jω) =
0 |ω| > ωc ,

which is the frequency response of an ideal lowpass filter with a cutoff frequency ωc .
Then, the frequency response of the ideal bandpass filter is H(jω) = H1 (j(ω − 2ωc )) + H1 (j(ω + 2ωc )).
Taking the inverse Fourier transform and using the frequency-shifting property of FT,

h(t) = ej2ωc t h1 (t) + e−j2ωc t h1 (t) = h1 (t) · 2 cos (2ωc t).

Therefore, g(t) = 2 cos (2ωc t).

1
Problem 6.27. [20pt] The output y(t) of a causal LTI system is related to the input x(t) by the differential
equation
dy(t)
+ 2y(t) = x(t). (1)
dt
(a) Determine the frequency response
Y (jω)
H(jω) =
X(jω)
of the system, and sketch its Bode plot.

Sol. Taking the Fourier Transform of both sides of Eq. (1),

jωY (jω) + 2Y (jω) = X(jω)

(2 + jω)Y (jω) = X(jω)
Y (jω) 1
∴ H(jω) = = .
X(jω) 2 + jω

To find the magnitude and phase of H(jω),

2 − jω 2 − jω
H(jω) = = ,
(2 + jω)(2 − jω) 4 + ω2
s 2  2
2 −ω 1
|H(jω)| = 2
+ 2
=√ ,
4+ω 4+ω 4 + ω2
 ω ω 
∠H(jω) = tan−1 − = − tan−1 .
2 2
The Bode plot of the frequency response is given in Fig. 1.

Figure 1: Bode plot of H(jω).

(b) If x(t) = e−t u(t), determine Y (jω), the Fourier transform of the output.

Sol. The FT of the input x(t) = e−1 u(t) is X(jω) = 1

1+jω . The FT of the output is

1
Y (jω) = H(jω)X(jω) = .
(2 + jω)(1 + jω)

2
Problem 9.7. [10pt] How many signals have a Laplace transform that may be expressed as

(s − 1)
(s + 2)(s + 3)(s2 + s + 1)

in its region of convergence?

√ √
Sol. The poles of the given Laplace transform are s0 = −2, s1 = −3, s2 = − 12 + j 23 and s2 = − 12 − j 3
2 .
Based on the locations of the poles, we may have the following signals with the regions of convergence:

• Re{s} < −3,

• −3 < Re{s} < −2,

• −2 < Re{s} < − 21 ,

• Re{s} > − 12 .
Therefore, we may find four different signals given the Laplace transform.

3
Problem 9.27. [10pt] We are given the following five facts about a real signal x(t) with Laplace transform
X(s):

1. X(s) has exactly two poles.

2. X(s) has no zeros in the finite s−plane.

3. X(s) has a pole at −1 + j.

4. e2t x(t) is not absolutely integrable.

5. X(0) = 8.

Determine X(s) and specify its region of convergence.

Sol. From the first, second, and third facts, we know that X(s) is of the form

A A
X(s) = = .
(s + a)(s + b) (s + 1 − j)(s + b)

Since x(t) is real, the poles are in conjugate pair, and thus b = −1 − j. Therefore,

A
X(s) = .
(s + 1 − j)(s + 1 + j)

From the fifth fact,

A
X(0) = = 8, ∴ A = 16.
(1 − j)(1 + j)
We finally find the Laplace Transform as
16 16
X(s) = = 2 .
(s + 1 − j)(s + 1 + j) s + 2s + 2

We now have to specify the region of convergence R of X(s). R may be either Re{s} < −1 or Re{s} > −1.
L
Using the property of shift in s-domain of LT, e2t x(t) ←
→ X(s − 2) with ROC shifted by 2 to the right from R.
Thus, the ROC of e2t x(t) may be either Re{s} < 1 or Re{s} > 1.
Since e2t x(t) is not absolutely integrable from the fourth fact, the jω-axis is not in the ROC of the LT of
e2t x(t) and thus the ROC is Re{s} > 1.
Therefore, the ROC of X(s) is specified as
16 16
X(s) = = 2 , Re{s} > −1.
(s + 1 − j)(s + 1 + j) s + 2s + 2

4
Problem 9.53. [30pt] The initial-value theorem states that, for a signal x(t) with Laplace transform X(s)
and for which x(t) = 0 for t < 0, the initial value of x(t) [i.e., x(0+)] can be obtained from X(s) through the
relation
x(0+) = lim sX(s). (2)
s→∞
First, we note that, since x(t) = 0 for t < 0, x(t) = x(t)u(t). Next, expanding x(t) as a Taylor series at t = 0+,
we obtain
tn
 
x(t) = x(0+) + x(1) (0+)t + · · · + x(n) (0+) + · · · u(t), (3)
n!
where x(n) (0+) denotes the nth derivative of x(t) evaluated at t = 0+.

(a) Determine the Laplace transform of an arbitrary term x(n) (0+)(tn /n!)u(t) on the right-hand side of
Eq. (3). (You may find it helpful to review the example of differentiation in s-domain on page 34 of the
lecture note.)

Sol. Using the property of differentiation in S-domain of Laplace transform on page 34 of the lecture
note,
tn −at L 1
e u(t) ←
→ , Re{s} > −a.
n! (s + a)n+1
Setting a = 0,
tn L 1
u(t) ←
→ n+1 , Re{s} > 0.
n! s
Using the linearity property of LT,
tn L 1
x(n) (0+) → x(n) (0+) n+1 ,
u(t) ← Re{s} > 0. (4)
n! s

(b) From your result in Item (a) and the expansion in Eq. (3), show that X(s) can be expressed as
∞
X 1
X(s) = x(n) (0+) .
sn+1
n=0

Sol. Eq. (3) can be rewritten as

+∞
X tn
x(t) = x(n) (0+) u(t).
n!
n=0

Taking Laplace transform and using Eq. (4),

+∞ (n)
X x (0+)
X(s) = . (5)
sn+1
n=0

(c) Demonstrate that Eq. (2) follows from the result of Item (b).

Sol. From Eq. (5),

+∞ (n)
X x (0+) x(1) (0+)
sX(s) = = x(0) (0+) + + ··· .
sn s
n=0

Taking limit of s → ∞,
lim sX(s) = x(0) (0+) = x(0+).
s→∞

5
Problem 9.42. [20pt] Determine whether each of the following statements is true of false. If a statement is
true, construct a convincing argument for it. If it is false, give a counterexample.

(1) The Laplace transform of t2 u(t) does not converge anywhere on the s−plane.

Sol. We know that

L 1
u(t) ←
→ , Re{s} > 0,
s
and using the property of the differentiation in s-domain of LT,

L 1
t2 u(t) ←
→ , Re{s} > 0.
s3
Thus, the statement is false.

(2) The Laplace transform of ejω0 t does not converge anywhere on the s−plane.

Sol. We first find the LT of x(t) = ejω0 t as follows,

∞
∞ ∞
e(jω0 −s)t
Z Z
jω0 t −st (jω0 −s)t
X(s) = e e dt = e dt =
−∞ −∞ jω0 − s
−∞

The integral does not converge for any value s and thus the statement is true.