EE1101: Signals and Systems JAN — MAY 2020

Tutorial 10 Solutions

Solution 1 (d) Poles at s = -2, +j, +j, -j, -j

Zeros at s = -0.5
For a causal system, the ROC is towards the right The system is not BIBO stable, as there are
of the rightmost pole in the pole-zero plot. Also for poles are at +j and -j.
a system to be stable, the ROC should include the
imaginary axis.
=(s)
(a) Poles at s = -5, -3
Zero at s = 2
The system is BIBO stable, as all the poles are
×
×j
in the left half plane.
× ◦
-2 -0.5 <(s)
=(s) ×−j
×

× × ◦
-5 -3 2 <(s)

Solution 2
(b) Poles at s = -3, -2, -2 Stability: An LTI system is stable iff the ROC of its
Zeros at s = -1 system function H(s) includes the jω−axis.
The system is BIBO stable, as all the poles are
in the left half plane. Causality:

• If an LTI system is causal (with a right sided

=(s) impulse response function h(t) = 0 for t < 0),
then the ROC of its transfer function H(s) is a
right sided plane.

× ×× ◦ • When H(s) is rational, then the system is causal

-3 -2 -1 <(s)
if and only if its ROC is the right half plane to
the right of the rightmost pole, and the order of
numerator is no greater than that of the denom-
inator.
(c) Poles at s = -2, 0
Zeros at s = -0.25 + j0.66, -0.25 - j0.66 a) Non-causal as the ROC is not to the right to
The system is not BIBO stable, as there is a the rightmost pole s = −2, BIBO unstable as
pole at origin. the ROC does not include the jω−axis.

=(s)
=(s) ROC
× ◦ ×
-3 -2.5 -2 <(s)
◦ j0.66

× × <(s)
-2 -0.25
◦ −j0.66 b) Non-causal as the ROC is not to the right to
the rightmost pole s = 3, BIBO unstable as the
ROC does not include the jω−axis.

1
 =(s) We know the following:
ROC
× ◦ × 1
2 2.5 3 <(s) L(u(t)) =
s
L(f (t)e−at ) = F (s + a)
dF (s)
L(tf (t)) = −
ds
c) Causal as the ROC is to the right to the right-
most pole s = −1, BIBO stable as ROC in-
where L(f (t) = F (s)
cludes the jω−axis.
Therefore , the impulse response h(t) is
=(s)
t2 −2t
h(t) = te−2t u(t) + e u(t)
× ◦ × 2
-2 -1.5 -1 <(s)
ROC
(b) The Laplace transform of the output y(t) of the
system to the input x(t) can be expressed as

d) Non-causal as the ROC is not to the right to Y (s) = H(s)X(s)

the rightmost pole s = −1, BIBO unstable as
the ROC does not include the jω−axis. 10
Given x(t) = 10u(t) =⇒ X(s) =
s
=(s) 10(s + 3)
ROC =⇒ Y (s) =
s(s + 2)3
× ◦ ×
-2 -1.5 -1 <(s)
Using the final value theorem, the final value of
y(t) would be

lim y(t) = lim sY (s)

NOTE: Here we describe an example of a right sided t→∞ s→0

ROC for a H(s) which is not causal.

10(s + 3)
=⇒ lim y(t) = lim = 3.75
Consider H(s) = es
with the ROC as Re(s) > −1. t→∞ s→0 (s + 2)3
s+1
1
We know that L(e−t u(t)) = . Also, the es causes To verify this, let us compute the signal y(t) and
s+1
a time shift giving h(t) = e−(t+1) u(t + 1). This is not then evaluating at the limit t tends to ∞. First
causal as h(t) 6= 0 when −1 < t < 0. However, the we need to evaluate partical fractions of Y (s).
ROC is a right half plane.

Only if H(s) is a rational function with the order 10(s + 3) A Bs2 + Cs + D

Y (s) = 3
= +
of denominator greater than the order of the numera- s(s + 2) s (s + 2)3
tor, will the ROC being a right-half plane imply that =⇒ 10s + 30 = A(s + 2)3 + (Bs3 + Cs2 + Ds)
the system is causal. =⇒ 10s + 30 = (A + B)s3 + (6A + C)s2
+(12A + D)s + 8A

Solution 3
30
(a) Applying partial fractions to H(s), Solving we get A = 8 ,B = − 30
8 ,C =
− 180
8 , D = −15

Bs2 + Cs + D 30 s2 + 6s + 5
s+3 A B C Now 3
= − =
H(s) = = + +  2 (s + 2) 8 (s + 2)3 
(s + 2) 3 s + 2 (s + 2) 2 (s + 2)3 30 s + 4s + 4 2s + 4 1
− + − =
=⇒ s + 3 = A(s2 + 4s + 4) + B(s + 2) + C 8  (s + 2)3 (s + 2)3 (s + 2)3
30 1 2 1
− + 2
−
8 s + 2 (s + 2) (s + 2)3
Solving we get A = 0, B = 1 and C = 1 
30 30 1 2
Alternatively we can observe that H(s) = Thus, Y (s) = − + −
s+3 s+2+1 1 1  8s 8 s + 2 (s + 2)2
3
= 3
= 2
+ . 1
(s + 2) (s + 2) (s + 2) (s + 2)3
(s + 2)3

2
 30 30 60
Thus, y(t) = u(t) − e−2t u(t) − te−2t u(t) + Since the system is causal, the ROC will be the half
8 8 8 plane right of the right most pole of H(s). Since the
30 t2 −2t poles of H(s) are −1 ± j, ROC is Re(s) > −1.
e u(t)
8 2 Z ∞
30 X(s) = x(t)e−st dt
lim y(t) = = 3.75 −∞
t→∞ 8 Z 0 Z ∞
As = et e−st dt + e−t e−st dt
−∞ 0
∞
lim e−t = 0, lim te−t = 0, lim t2 e−t = 0
Z 0 Z
t→∞ t→∞ t→∞ = et−st dt + e−t−st dt
−∞ 0
1 1
Solution 4 = + ; Re{s} < 1 ∩ Re{s} > −1
1−s 1+s
−2
(a) = ROC : −1 < Re{s} < 1
(s − 1)(s + 1)
s2 Y (s) − sY (s) − 2Y (s) = X(s)
1 1 Y (s) = H(s)X(s)
H(s) = 2 =
s −s−2 (s − 2)(s + 1) s+1
= 2 X(s)
s + 2s + 2
−2
={s} =
(s − 1)(s2 + 2s + 2)
Taking partial fraction decomposition,
× × <{s}
−1 2 −2 A Bs + C
2
= + 2
(s − 1)(s + 2s + 2) s − 1 s + 2s + 2
−2 = A(s2 + 2s + 2) + (Bs + C)(s − 1)
(b) The possible ROC’s for the above system are = (A + B)s2 + (2A − B + C)s + (2A − C)

I. Re{s} < −1 Solving for A,B,C we get

II. −1 < Re{s} < 2 0 = A + B =⇒ B = −A

III. Re{s} > 2 0 = 2A − B + C =⇒ C = −3A
−2 = 2A − C =⇒ 5A = −2
1 1 1
H(s) = = − −2 2 6
(s − 2)(s + 1) 3(s − 2) 3(s + 1) A= , B= , C= .
5 5 5
(a) System is stable =⇒ The ROC should Hence
contain the jω axis. −2 2s + 6
Y (s) = + 2
∴ ROC : −1 < Re{s} < 2 5(s − 1) 5(s + 2s + 2)
1 2t 1 −t
h(t) = − 3 e u(−t) − 3 e u(t) −2 2(s + 1) 4
= + +
5(s − 1) 5((s + 1) + 1)) 5((s + 1)2 + 1))
2
(b) System is causal =⇒ The ROC should be
to the right of the rightmost pole ROC : −1 < Re{s} < 1
∴ ROC : Re{s} > 2 Notice that the pole at s = −1 in the ROC of Y(s)
h(t) = 31 e2t u(t) − 31 e−t u(t) = 31 (e2t − got cancelled with the zero at the same position, and
e−t )u(t) the left limit of ROC at s = −1 came from the poles
at −1 ± j.
(c) System is neither causal nor stable =⇒
We know the following:
ROC : Re{s} < −1
h(t) = − 13 e2t u(−t) + 31 e−t u(−t) = 13 (e−t − s+a
L(e−at cos ωt · u(t)) =
2t
e )u(−t) (s + a)2 + ω 2
ω
L(e−at sin ωt · u(t)) =
(s + a)2 + ω 2
Solution 5 Therefore, in time domain:
Given that x(t) = e−|t| ,−∞ < t < ∞ 2 t 2 −t 4
y(t) = e u(−t) + e cos tu(t) + e−t sin tu(t)
Since the system is LTI, the output is x(t) ∗ h(t) or in 5 5 5
s-domain, by L.T properties 2 2 −t
= et u(−t) + e (cos t + 2 sin t)u(t)
5 5
Y (s) = X(s)H(s)

3
Solution 6 Solution 7
Taking unilateral Laplace transform of the given dif-
Given that H(s) is a rational function, we may take
a(s) ferential equation we get,
it to be H(s) = , where a(s) and b(s) are polyno-
b(s) s3 Y (s) − s2 y(0− ) − sy 0 (0− ) − y 00 (0− )
mials in s.
+6s2 Y (s) − 6sy(0− ) − 6y 0 (0− )
+11sY (s) − 11y(0− ) + 6Y (s) = X(s)
(a) The response is absolutely integrable for the sig-
1 (a) For zero state response all the initial conditions
nal u(t), whose Laplace transform is . Therefore, are assumed to be zero. Laplace transform of
s
H(s) x(t) is given by,
has no poles at Re(s) ≥ 0. Therefore, s = 0
s
must be a root of a(s), which cancels the s in the 1
X(s) = , Re(s) > −4
denominator. Take a(s) = sa1 (s). s+4
Using above equation we get,

1 1
(b) The response to tu(t) (Laplace transform 2 ) is Y (s)(s3 + 6s2 + 11s + 6) =
s s+4
not absolutely integrable. This implies, there can-
not be a repeated root for a(s) at s = 0. Therefore,
1
Y (s) =
(s3+ 6s2
+ 11s + 6)(s + 4)
(c) If a signal with a rational Laplace transform is of 1
=
finite duration, then its denominator is a constant (s + 1)(s + 2)(s + 3)(s + 4)
polynomial. The Laplace transform of the signal 1/6 1/2 1/2 1/6
d2 h dh = − + −
+2 + 2h(t) is (s2 + 2s + 2)H(s). This s+1 s+2 s+3 s+4
dt2 dt
is given to be of finite duration in time domain. Taking inverse unilateral Laplace Transform we
1 get,
Hence, b(s) = (s2 +2s+2) for some constant K.
K
1 −t 1 1 1
y(t) = e u(t) − e−2t u(t) + e−3t u(t) − e−4t u(t)
6 2 2 6
(d) The number of zeros at infinity is deg(b(s)) −
deg(a(s)) = 1. Since deg(b(s)) = 2, deg(a(s)) = 1. (b) For zero input response, we assume that X(s) =
Therefore a(s) = K1 s. 0, Substituting the initial conditions in the main
s equation 1 we get,
Therefore H(s) = KK1 2 .
s + 2s + 2
s2 + 5s + 6 1
Y (s) = =
s3 + 6s2 + 11s + 6 s+1
(e) Using the fact that H(1) = 0.2, we find that
KK1 = 1. Taking the unilateral Laplace transform of above
equation we get,
The required impulse response is
y(t) = e−t u(t)
s
H(s) = (c) The total response is the sum of zero state re-
s2 + 2s + 2
sponse and the zero input response.
The poles of the system are −1 ± j. Since the system 7 −t 1 1 1
is causal, the region of convergence is Re(s) > −1. y(t) = e u(t) − e−2t u(t) + e−3t u(t) − e−4t u(t)
6 2 2 6