EE1101 Signals and Systems Jan-May 2018

Tutorial 2 Solutions

Solution 1 rational number, x[n] is periodic.

The fundamental period would be N =
(a) Let x1 (t) and x2 (t) be periodic with periods L.C.M(1, 7, 5) = 35
T1 and T2 respectively. such that

x1 (t + T1 ) = x1 (t) Solution 2
x2 (t + T2 ) = x2 (t) 1
To prove δ(at) = δ(t)
|a|
For x(t) = x1 (t) + x2 (t) to be periodic, Let g1 (t) and g2 (t) be generalized functions. Then
the equivalence property states that
x(t + T ) = x1 (t + T ) + x2 (t + T )
Z ∞ Z ∞
=⇒ x(t + T ) = x1 (t + mT1 ) + x2 (t + kT2 ) g1 (t) = g2 (t) ⇐⇒ φ(t)g1 (t)dt = φ(t)g2 (t)dt
−∞ −∞
So,
mT1 = kT2 = T for all suitably defined
R ∞testing functions φ(t).
T1 k Consider the integral −∞ φ(t)δ(at)dt
= = rational number Let τ = at. Consider the case when a > 0 ⇒ τ =
T2 m
at = |a|t ⇒ dτ = |a|dt
In other words, the sum of two periodic
signals is periodic only if the ratio of their Z ∞
1
Z ∞
τ
respective periods can be expressed as a φ(t)δ(at)dt = φ( )δ(τ )dτ
−∞ |a| −∞ a
rational number. Then the fundamental
period is the least common multiple of T1 1 τ
= φ( )
and T2 i.e. T = L.C.M {T1 , T2 } |a| a
τ =0
1
= φ(0)
|a|
(b) (a) Let x(t) = x1 (t) + x2 (t) where x1 (t) =
2cos(10t+1) and x2 (t) = −sin(4t−1).
x1 (t) and x2 (t) are peridic with peri-
2π 2π Now, consider the case when a < 0 ⇒ τ = at =
ods given by T1 = and T2 =
10 4 −|a|t ⇒ dτ = −|a|dt
respectively.
T1 2 Z ∞
1
Z −∞
τ
Since = is a rational number, φ(t)δ(at)dt = − φ( )δ(τ )dτ
T2 5 |a| a
−∞
x(t) is periodic. Z ∞∞
1 τ
The fundamental period is T = L.C.M = φ( )δ(τ )dτ
2π 2π |a| −∞ a
{ , }=π
10 4 1 τ
= φ( )
|a| a
(b) For a discrete time signal, the same τ =0
concept as in Q1 can be applied. 1
= φ(0)
Each of the complex exponentials are |a|
2πk
periodic with periods N1 = 4π = 7
7
2πk
(for k = 2) and N2 = 2π = 5 (for Thus for any a,
5 Z ∞
N1 7 1
k = 1) respectively. Since = is a φ(t)δ(at)dt = φ(0)
N2 5 −∞ |a|

1
 R∞
But we know φ(0) = −∞ φ(t)δ(t)dt (d)
∞ Z ∞
1
Z
φ(t)δ(at)dt = φ(t)δ(t)dt
Z ∞
−∞ |a| −∞ y= sin(πt)δ(2t − 3)dt
Z ∞ −∞
1
= φ(t) δ(t)dt Let2t − 3 = u =⇒ du = 2dt
−∞ |a|
1 ∞
 
π(u + 3)
Z
Now by the equivalence property mentioned be- y= sin δ(u)du
2 −∞ 2
fore, 1

π(0 + 3)
Z ∞
1 = sin δ(u)du
δ(at) = δ(t) 2 2 −∞
|a|
1
=−
2
Solution 3
For a Dirac Delta function δ(t), the following two
(e)
properties hold true:

x(t)δ(t − t0 ) = x(t0 )δ(t − t0 ) Z ∞

Z ∞ y= e−t δ(t + 3)dt
δ(τ )dτ = 1 −∞
−∞ Let t + 3 = u
Z ∞
(a) y = e3 e−u δ(u)du
−∞
∞
= e3
Z
y(t) = x(τ )δ(t − τ )dτ
Z−∞
∞
= x(t)δ(t − τ )dτ
−∞ (f)
Z ∞
= x(t) δ(t − τ )dτ
Z−∞
∞
Z ∞
= x(t) δ(τ )dτ y= (t3 + 4)δ(1 − t)dt
−∞ −∞
Z ∞
= x(t) 3
= (1 + 4) δ(1 − t)dt
−∞
Z ∞
(b) =5 δ(t)dt
−∞
Z ∞ =5
y(t) = x(t − τ )δ(τ )dτ
−∞
Let t − τ = u
Z ∞
(g)
y(t) = x(u)δ(t − u)du
−∞
Z ∞
Z ∞
= x(t) δ(t − u)du
−∞ y= x(2 − t)δ(3 − t)dt
−∞
= x(t)
2−t=u
Z ∞
(c) y= x(u)δ(1 + u)du
−∞
Z ∞
Z ∞
y= e−jωt δ(t)dt y = x(−1) δ(1 + u)du
−∞
−∞ Z ∞
Z ∞
=e −jω∗0
δ(t)dt y = x(−1) δ(u)du
−∞
−∞
=1 = x(−1)

2
 (h) (b) x(t) can be expressed as follows.
∞  
π
Z
x−1
y= e cos (x − 5) δ(x − 3)dx
−∞ 2 [x(t) = u(t) − δ(t − 1) − δ(t − 2) − δ(t − 3)
x−1 4 x−5
Let e =e e
Z ∞  
4 x−5 π Z t Z t Z t
y=e e cos (x − 5) δ(x − 3)dx x(t)dt = u(t)dt − δ(t − 1)dt
−∞ 2
−∞ −∞ −∞
x−5=u Z t Z t
Z ∞
π
  − δ(t − 3)dt − δ(t − 3)dt
4 u −∞ −∞
y=e e cos (u) δ(u + 2)du
−∞ 2 = r(t) − u(t − 1)
 Z ∞
4 −2 π − u(t − 2) − u(t − 3)
y = e ∗ e cos (−2) δ(u + 2)du
2 −∞
Z ∞
y = −e2 δ(u)du 
−∞ 
0, if t≤0
= −e2

t, if 0≤t<1


Z t 
x(t)dt = t − 1, if 1≤t<2
−∞ 
Solution 4 




t − 2, if 2≤t<3
t − 3, if t≥3

(a) x(t) can be expressed as follows.
Rt
x(t) = u(t) − 2u(t − 1) + u(t − 3) + δ(t − 3) −∞ x(t)dt

Z t Z t Z t
2
x(t)dt = u(t)dt − 2 u(t − 1)dt
−∞ −∞ −∞ 1
Z t Z t
+ u(t − 3)dt + δ(t − 3)dt
−∞ −∞
-1 0 1 2 3 4 t
= r(t) − 2r(t − 1) + r(t − 3) + u(t − 3)
-1

Thus, -2



0, if t≤0
Z t 
t, if 0≤t≤1
x(t)dt =
−∞ 

2 − t, if 1≤t<3 Solution 5
t≥3

0, if
• Suppose y1 (t) is the output of a system to
Rt input x1 (t) and y2 (t) is the output to input
−∞ x(t)dt x2 (t).The system is linear if the response of
the system to the input ax1 (t) + bx2 (t) is
2 ay1 (t) + by2 (t) (a and b can be complex in
general).
1
• Suppose y(t) is the output of a system to
the input x(t). The system is time invari-
-1 0 1 2 3 4 t ant if the output of the system to the input
x(t − t0 ) is y(t − t0 ).
-1
• A system is causal if the output of the sys-
-2 tem at any time depends only on the present
and past values of the input and not on the
future values of input.

3
 • A system is said to be stable if all bounded • If a system is invertible then an inverse sys-
inputs to the system result in bounded out- tem exists which when cascaded with the
puts. original system gives an output equal to the
input to the first system.

Qn Linear Time invariant Causal Stable Invertible

dx
(a)y(t) = dt Yes Yes Yes No No
R 3t
(b)y(t) = −∞ x(τ )dτ Yes No No No Yes
(c)y(t) = x(t/2) Yes No No Yes Yes

x(t) − x(t − 100) t ≥ 0
(d)y(t) = Yes No Yes Yes No
0 otherwise
dy(t) 2 dx(t)
(e) dt + 3ty(t) = t dt Yes No Yes No No
(f)y(t) = ∞
P
n=−∞ x(t)δ(t − nT ) Yes No Yes No No
(g)y(t) = x(2t − 4) Yes No No Yes Yes

The detailed explanations are given below: Consider the following input to the system.

(a) 1 −1 ≤ t ≤ 1
dx(t) x(t) =
0 otherwise
y(t) =
dt
Suppose y1 (t) is the output of a system to The output of the system to the above in-
input x1 (t) and y2 (t) is the output to input put will go unbounded at -1 and +1. Hence
x2 (t). Let y(t) be the output of the system the system is unstable.
to the input ax1 (t) + bx2 (t). Then
The system is not invertible since differ-
d
y(t) = (ax1 (t) + bx2 (t)) entiating any constant will give 0 and hence
dt
different inputs give the same output.
d d
=a (x1 (t)) + b (x2 (t))
dt dt
(b)
= ay1 (t) + by2 (t) Z 3t

Hence the system is linear. y(t) = x(τ )dτ

−∞
Suppose y(t) is the output of a system to Suppose y1 (t) is the output of a system to
input x(t). Let y1 (t) be the output of the input x1 (t) and y2 (t) is the output to input
system to the input x(t − t0 ). Then x2 (t). Let y(t) be the output of the system
d to the input ax1 (t) + bx2 (t).
y1 (t) = (x(t − t0 ))
dt Z 3t
d y(t) = ax1 (τ ) + bx2 (τ )dτ
= (x(t − t0 )) −∞
d(t − t0 ) Z 3t Z 3t
= y(t − t0 ) (∵ d(t − t0 ) = dt) =a x1 (τ )dτ + b x2 (τ )dτ
−∞ −∞

Hence the system is time invariant. = ay1 (t) + by2 (t)

Hence the system is linear.

dx x(t) − x(t − dt)
y(t) = = lim Suppose y(t) is the output of a system to
dt dt→0 dt
input x(t). Let y1 (t) be the output of the
Since the output of the system depends on
system to the input x(t − t0 ).
the present and past values of the input
Z 3t
alone the system is causal.
y1 (t) = x(τ − t0 )dτ
−∞

4
 τ − t0 = u For negative values of t we find that the
output of the system depends on the future
hence dτ = du values of the input. Hence the system is
not causal.
Z 3t−t0
= x(u)du
−∞ For all bounded values of the input we get
But bounded values at the output. Hence the
Z 3t−3t0
system is stable.
y(t − t0 ) = x(τ )dτ
−∞
Hence The system is invertible since the inverse
y(t − t0 ) 6= y1 (t) of the system exists which is y(t) = x(2t).
Hence the system is time variant. (d)
For t > 0 the output of the system depends 
x(t) − x(t − 100) t≥0
on the future values of the input hence the y(t) =
0 otherwise
system is not causal.
Suppose y1 (t) is the output of a system to
input x1 (t) and y2 (t) is the output to input
Consider an input x(t) = 1 for all val-
x2 (t). Let y(t) be the output of the system
ues of t.The output of the system goes un-
to the input ax1 (t) + bx2 (t).
bounded.Hence the system is unstable.

ax1 (t) + bx2 (t) − ax1 (t − 100) − bx2 (t − 100)
y(t) =
If two different inputs give the same output 0
to the above system their values should be 
a(x1 (t) − x1 (t − 100)) + b(x2 (t) − x2 (t − 100))
the same at every t since the limit of the in- y(t) =
0
tegral depends on t. Hence the two inputs
have to be the same. Hence the system is y(t) = ay1 (t) + by2 (t)
invertible. Hence the system is linear.
(c)
y(t) = x(t/2) Suppose y(t) is the output of a system to
input x(t). Let y1 (t) be the output of the
Suppose y1 (t) is the output of a system to system to the input x(t − t0 ).
input x1 (t) and y2 (t) is the output to input 
x2 (t). Let y(t) be the output of the system x(t − t0 ) − x(t − t0 − 100) t≥0
y1 (t) =
to the input ax1 (t) + bx2 (t). 0 otherwise
But
y(t) = ax1 (t/2) + bx2 (t/2) 
x(t − t0 ) − x(t − t0 − 100) t ≥ t0
= ay1 (t) + by2 (t) y(t−t0 ) =
0 otherwise
Hence the system is linear. Hence the system is time variant.
Suppose y(t) is the output of a system to
input x(t). Let y1 (t) be the output of the The output of the system depends on the
system to the input x(t − t0 ). present and past value of the input .Hence
the system is causal.
y1 (t) = x(t/2 − t0 )

But The system is stable since bounded inputs

y(t − t0 ) = x((t − t0 )/2) to the system give bounded outputs.
= x(t/2 − t0 /2)
Consider the constant input function. The
Hence
constant can be of any value. But they all
y(t − t0 ) 6= y1 (t)
give the same zero output at all values of t.
Hence the system is time variant. Hence the system is not invertible.

5
 ∞
(e) X
dy(t) dx(t) = ax1 (t)δ(t − nT ) + bx2 (t)δ(t − nT )
+ 3ty(t) = t2 n=−∞
dt dt
= ay1 (t) + by2 (t)
Suppose y1 (t) is the output of a system to
input x1 (t) and y2 (t) is the output to input Hence the system is linear.
x2 (t). Let y(t) be the output of the system
to the input ax1 (t) + bx2 (t). Suppose y(t) is the output of a system to
  input x(t).Let y1 (t) be the output of the
dy(t) 2 d system to the input x(t − t0 ).
+ 3ty(t) = t ax1 (t) + bx2 (t)
dt dt
∞
X
dx1 (t) dx2 (t) y1 (t) = x(t − t0 )δ(t − nT )
= at2 + bt2
dt dt n=−∞

But
 
dy1 (t) dy2 (t)
=a + 3ty1 (t)) + b( + 3ty2 (t)
dt dt ∞
X
    y(t − t0 ) = x(t − t0 )δ(t − t0 − nT )
d n=−∞
= ay1 (t)+by2 (t) +3t ay1 (t)+by2 (t)
dt
Hence y(t − t0 ) 6= y1 (t) Hence the system is
Hence the system is linear. time variant.

dx(t)
Since the coefficient of contains t, the The system represents an ideal sampler.
dt The sampler output has non-zero values
system is time variant.
only at t = nT which is equal to the value of
the input x(t) at the same instant. Hence,
The system is causal since the output of
it is causal.
the system depends only on the present
value of the input.
The system is not stable since output is a
sequence of impulses and impulse function
The system is unstable. For a signal hav- takes an unbounded value.
dy(t) t dx(t)
ing small value of , y(t) = . If
dt 3 dt
suppose x(t) is a triangular wave, the out- The system is not invertible since the in-
put goes unbounded. puts can be different but if they have the
same value at nT then they give the same
The system is not invertible due to the output.
dx(t)
presence of , whose value will be same
dt (g)
for x(t) and x(t) + C where C is any con-
stant. y(t) = x(2t − 4)
Suppose y1 (t) is the output of a system to
input x1 (t) and y2 (t) is the output to input
(f) x2 (t). Let y(t) be the output of the system
∞
to the input ax1 (t) + bx2 (t).
X
y(t) = x(t)δ(t − nT )
n=−∞
y(t) = ax1 (2t − 4) + bx2 (2t − 4)
Suppose y1 (t) is the output of a system to
input x1 (t) and y2 (t) is the output to input = ay1 (t) + by2 (t)
x2 (t). Let y(t) be the output of the system Hence the system is linear.
to the input ax1 (t) + bx2 (t). Suppose y(t) is the output of a system to
input x(t).Let y1 (t) be the output of the
system to the input x(t − t0 ).
∞
X y1 (t) = x(2t − 4 − t0 )
y(t) = (ax1 (t) + bx2 (t))δ(t − nT )
n=−∞ y(t − t0 ) = x(2(t − t0 ) − 4)

6
 Hence y(t − t0 ) 6= y1 (t) Hence the system is The given system is linear.
time variant. Check for time-invariance: Delay the input
x(t) by t0 . The y1 (t) = t2 x(t − 1 − t0 ).
For all values of t > 4 the output of the Now delaying the output y(t) by t0 to get
system depends on the future value of the y(t − t0 ) = (t − t0 )2 x(t − 1 − t0 ) 6= y1 (t).
input. Hence the system is not causal. Hence the system is time variant.

The system is stable since bounded inputs (b) y[n] = x2 [n−2] : System is non-linear and
give bounded outputs. The system per- time-invariant.
forms only a time-shifting and time-scaling Check for Linearity:Let y1 [n] be the re-
operation sponse of x1 [n], y2 [n] be the response of
x2 [n] and y[n] be the response of the com-
bined input x[n] = ax1 [n] + bx2 [n] .
The system is invertible since the inverse
of the system exists which is y(t) = x( 2t +2).
y1 [n] = x21 [n − 2]
Solution 6
We have,
y2 [n] = x22 [n − 2]
y[n] = x[n]x[n − 2]

(a) The system is not memoryless. Since, the

output at any time instant n depends on the
input at the time instant n-2. y[n] = x2 [n − 2]
= (ax1 [n − 2] + bx2 [n − 2])2
(b) When x[n] = Aδ[n]
= a2 x21 [n − 2] + b2 x22 [n − 2]
y[n] = Aδ[n]Aδ[n − 2] = 0 + 2abx1 [n − 2]x2 [n − 2]
Aδ[n] is non-zero only at n = 0 and Aδ[n−2] 6= ay1 [n] + by2 [n]
is non-zero only at n = 2. When both are
multiplied, we get zero for all n values. The given system is non-linear.
Check for time-invariance: Delay the input
(c) The system is clearly non-invertible.
x[n] by n0 . Then y1 [n] = x2 [n − 2 − n0 ].
To prove non-invertibility, we need one
Now delaying the output y[n] by n0 to get
counter-example, which is part (b) itself.
y[n − n0 ] = x2 [n − 2 − n0 ] = y1 [n]. Hence
the system is time invariant.
Solution 7
(a) y(t) = t2 x(t−1): System is linear and time (c) y[n] = x[n + 1] − x[n − 1] : System is lin-
variant ear and time invariant. (Proof similar to
Check for Linearity:Let y1 (t) be the re- above)
sponse of x1 (t), y2 (t) be the response of
x2 (t) and y(t) be the response of the com- x[n] − x[−n]
bined input x(t) = ax1 (t) + bx2 (t) . (d) y[n] = Odd{x[n]} = : System
2
is linear and time variant.
y1 (t) = t2 x1 (t − 1) Check for time-invariance: Delay the
y2 (t) = t2 x2 (t − 1) input x[n] by n0 . Then y1 [n] =
x[n − n0 ] − x[−n − n0 ]
. Now delay-
2
y(t) = t2 x(t − 1) ing the output y[n] by n0 to get
x[n − n0 ] − x[−(n − n0 )]
= t2 (ax1 (t − 1) + bx2 (t − 1)) y[n − n0 ] = =
2
= at2 x1 (t − 1) + bt2 x2 (t − 1) x[n − n0 ] − x[−n + n0 ]
. Hence the system
= ay1 (t) + by2 (t) 2
is time variant.

7
Solution 8 Solution 10
Let y(t) be the output of the system for input (a)
x(t) = est . Then,  Z t 
H{est } = y(t) y(t) = A cos ωc t + ω∆ m(τ )dτ
−∞

Since the system is time-invariant, we have Given ωc = 8π, ω∆ = 2π and m(t) =

u(t + 2) − u(t − 1).
H{es(t+t0 ) } = y(t + t0 )
 Z t


for arbitrary real t0 . Since the system is linear, y(t) = A cos 8πt + 2π u(τ + 2) − u(τ − 1) dτ
we have −∞

H{es(t+t0 ) } = H{est est0 } = est0 H{est } = est0 y(t) m(t) is sketched below:

Thus, m(t)
y(t + t0 ) = est0 y(t)
1
Setting t = 0, we obtain

y(t0 ) = y(0)est0 -2 -1 0 1 2 t

Since t0 is arbitrary, by changing t0 to t, we can

rewrite the above equation as Case 1: t < −2
In this case, the signal m(t) = 0.
y(t) = y(0)est = λest

=⇒ H{est } = λest y(t) = A cos (8πt)

where λ = y(0)
Case 2: −2 < t < 1

Solution 9  Z t 
The signal x2 (t) can be represented in terms of y(t) = A cos 8πt + 2π 1 · dτ
−2
the given signal x1 (t) as follows:
= A cos (8πt + 2π(t + 2))
x2 (t) = x1 (t) − x1 (t − 2) = A cos (10πt + 4π)

Given that y1 (t) is the response to the input x1 (t), Case 3: t > 1
let y2 (t) be the response to the input x2 (t). Since
the given system is an LTI system, y2 (t) is given
as
 Z 1 
y(t) = A cos 8πt + 2π 1 · dτ
−2
y2 (t) = y1 (t) − y1 (t − 2)
= A cos (8πt + 2π(3))
y2 (t) = A cos (8πt + 6π)
2
1

0.5
y(t)

0 1 2 3 4 t 0

−0.5

−1
−2 −1 0 1
-2 ω = 8π ω = 10π ω = 8π
t→

8
(b) – Let y1 (t) be the response of the system input at previous time instants.
to m1 (t) and y2 (t) be the response of
the system to m2 (t).
– The system is causal as output at any
Z t
instant does not depend on the future
 
y1 (t) = A cos ωc t + ω∆ m1 (τ )dτ
−∞
values of input.
 Z t 
y2 (t) = A cos ωc t + ω∆ m2 (τ )dτ – Let y1 (t) be the response of the system
−∞
to m(t − t0 ).
Consider y(t) be the response of the
system to the input am1 (t) + bm2 (t).  Z t 
 y1 (t) = A cos ωc t + ω∆ m(τ − t0 )dτ
y(t) = A cos ωc t −∞

Z t 
However, y(t) delayed by t0 may be ex-
 
+ ω∆ am1 (τ ) + bm2 (τ ) dτ pressed as:
−∞
6= ay1 (t) + by2 (t)  Z t−t0 
y(t−t0 ) = A cos ωc (t − t0 ) + ω∆ m(τ )dτ
Hence the modulation system is non- −∞
linear.
– The system has memory as the out- Since y(t − t0 ) 6= y1 (t), the system is
put at any instant is dependent on the time-variant.