EE1101 : Signals and Systems

Tutorial 3 Solutions

1. Convolution can be calculated by using,

∞
X
x[n] ∗ δ[n] = x[k]δ[n − k] = x[n]
k=−∞

So we plot x[k], h[−k] and then h[n − k] for finding convolution using graphical method.

x[k]

−3 −2 −1 0 1 2 3 4 5 k

−1

h[n − k]

n−3n−2n−1 n n+1n+2n+3 n+4 k

h[−k] is constructed first and then h[n − k] is drawn as shown in the figure. Note here that
the x − axis is k.

1
 h[−k] x[k]

−3 −2 −1 0 1 2 3 4 5 k (n = 0)
n−3n−2n−1 n n+1n+2n+3 n+4
−1

h[n − k] and x[k] are to be multiplied for different values of n and then sum of each product
is the convolution. The figure above depicts when n=0. Similarly by drawing at different
values of n and summing them we get y[n]

y[n]

−3 −2 −1 0 1 2 3 4 5 n

−2

2. (a)

x[n] = αn u[n]

h[n] = β n u[n], α 6= β

Now, y[n] = x[n] ∗ h[n]

∞
X
y[n] = αk u[k]β (n−k) u[n − k]
k=−∞
Xn
y[n] = αk β (n−k) u[k]
k=−∞

2
 if n ≥ 0,

n
X
y[n] = αk β (n−k) = β n + αβ n−1 + α2 β n−2 + · · ·
k=0
α α 2
= β n [1 + + ( ) + ···]
β β
β n+1 −α n+1
= .
β−α

For n < 0 the convolution result is 0.

(b) Let

h[n] = u[n]

x[n] = an u[−n − 1], |a| > 1

Now, y[n] = x[n] ∗ h[n]

∞
X
y[n] = ak u[−k − 1]u[n − k]
k=−∞
Xn
y[n] = ak u[−k − 1]
k=−∞

if n > −1,

−1
X
y[n] = ak = a−1 + a−2 + · · ·
k=−∞
1 1
= a−1 [1 + + + ···]
a a2
1
= .
a−1

3
 if n ≤ −1,

n
X
y[n] = ak = an + an−1 + an−2 + · · ·
k=−∞
1 1  1 
= an (1 + + 2 + · · · ) = an
a a 1 − a1
an+1
=
a−1

 an+1

, n ≤ −1
a−1
∴ y[n] =
1
n > −1

a−1 ,


3. (a) We have to prove the commutative property of convolution operator.

∞
X
x[n] ∗ y[n] = x[k]y[n − k]
k=−∞

Let l = n − k =⇒ k = n − l

∞
X
x[n] ∗ y[n] = x[n − l]y[l]
l=−∞
X∞
= y[l]x[n − l]
l=−∞

x[n] ∗ y[n] = y[n] ∗ x[n]

(b) We have to prove the distributive property of convolution operator.

∞
X
x[n] ∗ (y[n] + z[n]) = x[k](y[n − k] + z[n − k])
k=−∞
X∞ ∞
X
= x[k]y[n − k] + x[k]z[n − k]
k=−∞ k=−∞

x[n] ∗ (y[n] + z[n]) = x[n] ∗ y[n] + x[n] ∗ z[n]

4
 (c)

∞
X
x[n] ∗ δ[n − a] = x[k]δ[(n − a) − k]
k=−∞
X∞
= x[n − a]δ[(n − a) − k] (as x[n]δ[n − a] = x[a]δ[n − a])
k=−∞
∞
X
= x[n − a] δ[(n − a) − k]
k=−∞
∞
X
x[n] ∗ δ[n − a] = x[n − a] (as δ[k] = 1)
k=−∞

4. To compute the given convolutions, we first compute x[n] ∗ δ[n − a] (a is an integer) and then
use linearity and time invariance of convolution operation.

∞
X
x[n] ∗ δ[n − a] = x[k]δ[n − k − a] = x[n − a]
k=−∞

(a)

y1 [n] = x[n] ∗ (2δ[n + 1] + 2δ[n − 1]) = 2(x[n + 1] + x[n − 1])

= 2δ[n + 1] + 4δ[n] + 2δ[n − 1] + 2δ[n − 2] − 2δ[n − 4].

y1 [n]

−3 −2 −1 0 1 2 3 4 5 n

−2

(b) Here, we use commutative and associative properties of convolution operator and get,

y2 [n] = x[n + 2] ∗ h[n] = (x[n] ∗ δ[n + 2]) ∗ h[n] = x[n] ∗ h[n] ∗ δ[n + 2]

= y1 [n] ∗ δ[n + 2] = y1 [n + 2].

5
 y2 [n], y3 [n]

−5 −4 −3 −2 −1 0 1 2 3 n

−2

5. The signal y[n] is

∞
X
y[n] = x[n] ∗ h[n] = x[k]h[n − k]
k=−∞

In this case, the summation reduces to

9
X 9
X
y[n] = x[k]h[n − k] = h[n − k]
k=0 k=0

X9
y[4]= h[4 − k]
k=0
⇒ 5=h[4] + h[3] + h[2] + h[1] + h[0] + h[−1] + h[−2] + h[−3] + h[−4] + h[−5]

⇒ 5=h[4] + h[3] + h[2] + h[1] + h[0] (∵ h[n] = 0 ∀n < 0)

∴N ≥4
X9
y[14]= h[14 − k]
k=0
⇒ 0=h[14] + h[13] + h[12] + h[11] + h[10] + h[9] + h[8] + h[7] + h[6] + h[5]

As value of h[n] is either 0 or 1, in order to satisfy the above condition we need h[14] =
h[13] = h[12] = h[11] = h[10] = h[9] = h[8] = h[7] = h[6] = h[5] = 0. Therefore N = 4
R∞
6. (a) We know that y(t) = x(t − τ )h(τ )dτ . Here, h(τ ) is non-zero only in (3, 4), then the
−∞
R4
above integral becomes y(t) = x(t − τ )dτ . Further, x(t) is non-negative only in (2, 3),
3
and zero elsewhere. Eventually, x(t − τ ) will be non-zero only between t − 3 and t − 2.
y(t) will be zero t − 3 > 4 ⇒ t > 7 and t − 2 < 3 ⇒ t < 5. This implies that the above
integral is non-zero for 5 ≤ t ≤ 7. Hence, y(t) is non-zero for t ∈ (5, 7). This can be
seen in the figures shown below.

6
 h(t)

−2 −1 0 1 2 3 4 t

x(−τ )

−3 −2 −1 0 1 2 3 4 τ
−2.5

x(τ ) can be any signal which is non negative and symmetric about 2.5. Here x(−τ ) is
plotted.

x(t − τ )

2
h(τ )
1

−3 −2 −1 0 1 2 3 4 τ
t−3 t−2 t−1 t t+1 t+2 t+3 t+4

7
 x(t − τ )

2
h(τ )
1

−3 −2 −1 0 1 2 3 4 τ
t−4 t−3 t−2

x(t − τ ) is moved towards h(τ ) and we check conditions for y(t) to be non-zero by seeing
where the multiplication between x(t − τ ) and h(τ ) yields a non-zero value.
R4 t−3
R
(b) y(t) = x(t − τ )dτ = x(τ )dτ . Again, x(τ ) is non-negative only in τ ∈ (2, 3), with
3 t−4
symmetry around τ = 25 . The integral computes the complete area occupied by x(τ )
only when t = 6, as only at t = 6 the limits of the integral is 2 to 3. For other values of
t, the integration will either be equal to area of a part of x(τ ) or zero. Therefore, y(t)
will have maximum value at t = 6. It can be seen from the figure that the area will be
maximum when the x(t − τ ) is superimposed on the h(τ ). Note that signal x(t − τ ) can
be any signal, here it is been assumed to be square pulse.

x(t − τ )

2
h(τ )
1

−3 −2 −1 0 1 2 3 4 τ
t−4 t−3 t−2

t+1
R
7. Given: y(t) = sin(t − τ )x(τ )dτ .
−∞

(a) The response to a delayed input will be,

Zt+1 t−t
Z o +1
0 0 0
y1 (t) = sin(t − τ )x(τ − to )dτ = sin(t − τ − to )x(τ )dτ
−∞ −∞

8
 However, the delayed response of the system is given by,

t−t
Z o +1
y2 (t) = sin(t − to − τ )x(τ )dτ.
−∞

Since, y1 (t) = y2 (t), the given system is time-invariant.

t+1
R R∞
(b) Now, y(t) = sin(t − τ )x(τ )dτ = sin(t − τ )u(t + 1 − τ )x(τ )dτ . Hence, the impulse
−∞ −∞
response of the system is given by, h(t) = sin(t)u(t + 1).

(c) The system given is non-causal since the output depends on future values of the input.
R∞ Rt
8. (a) f (t) = u(t) ∗ u(t) = u(t − τ )u(τ )dτ = 1 dτ = t, (for t ≥ 0). Thus f (t) = tu(t).
−∞ 0

f (t)

−2 −1 0 1 2 3 t

(b)

Z∞ Z∞
f (t) = x(t) ∗ h(t) = h(τ )x(t − τ )dτ = (−e−τ + 2e−2τ )u(τ )10e−3(t−τ ) u(t − τ )dτ
−∞ −∞
Zt Zt
−τ −2τ −3(t−τ )
= 10(−e + 2e )e dτ = 10 (−e2τ −3t + 2eτ −3t )dτ = −5e−t + 20e−2t − 15e−3t .
0 0

Hence, f (t) = −5e−t + 20e−2t − 15e−3t for t ≥ 0 and zero elsewhere.

(c) (i) Given: h(t) = 2e−2t u(t) and x(t) = 1, ∀2 ≤ t ≤ 4 and zero otherwise. Let y(t) =
R∞
h(t) ∗ x(t) = h(τ )x(t − τ )d τ . Now, x(t − τ ), as a function of τ , will be 1 from
−∞
τ = t − 4 to τ = t − 2, for any given t, and zero outside this range. Consider the
following cases:
Case 1: When t − 2 < 0 ⇒ t < 2.
The product h(τ )x(t − τ ) = 0, as there is no common overlap between the
non-zero regions of these two signals. Hence, y(t) = 0, ∀t < 2.

9
 Case 2: Suppose t − 2 ≥ 0 and t − 4 < 0, i.e., 2 ≤ t < 4.
R∞ t−2
2e−2τ d τ = 1 − e−2t+4 , ∀2 ≤ t < 4.
R
Then, y(t) = h(τ )x(t − τ )d τ =
−∞ 0
Case 3: Finally, t − 4 ≥ 0, i.e., t ≥ 4.
t−2
2e−2τ d τ = −(e−2t+4 − e−2t+8 ). This value of y(t) is for the
R
Now, y(t) =
t−4
range t ≥ 4.

Hence, the signal y(t) is given by,


0, t<2






y(t) = 1 − e−2t+4 , 2≤t<4




e−2t+8 − e−2t+4 , t ≥ 4


(ii) Given: h(t) = 2e−2t u(t) and x(t) = cos(4πt) = 21 (ej4πt + e−j4πt ). Then, we obtain,

Z∞ Z∞
y(t) = x(t) ∗ h(t) = h(τ )x(t − τ )dτ = ej4πt e−j4πτ −2τ + e−j4πt ej4πτ −2τ dτ
−∞ 0
 Z∞ Z∞ 
= ej4πt e−(j4π+2)τ dτ + e−j4πt e−(2−j4π)τ dτ
0 0
 ej4πt e−j4πt  4 cos(4πt) + 8π sin(4πt)
= + = .
2 + j4π 2 − j4π 4 + 16π 2

(d) y(t) = [u(t) ∗ u(t − 2)]u(4 − t) = r(t − 2)u(4 − t).

y(t)

−1 0 1 2 3 4 t

(e) i)
 
Given: f (t) = u(−t) and g(t) = 2 u(t) − u(t − 1) . The signals look like,

10
 f (t) g(t)
2

0 t 0 1 t

R∞
Now, h(t) = f (t) ∗ g(t) = f (t − τ )g(τ ) dτ .
−∞

f (t − τ ) g(τ )
2

0 t τ 0 1 τ

Consider the following cases:

Case 1: t > 1.
In this case, there is no overlap between g(τ ) and f (t − τ ). Thus, h(t) = 0, ∀t > 1.

g(τ )
2
f (t − τ )
1

0 1 t τ

R1
Case 2: t ≤ 0. Here, we get, h(t) = 2 dτ = 2.
0

g(τ )
2
f (t − τ )
1

t 0 1 τ

R1
Case 3: 0 < t ≤ 1. Then, h(t) = 2 dτ = 2(1 − t).
t

11
 g(τ )
2
f (t − τ )
1

0 t 1 τ

The final expression for the signal h(t) is given by,


2 t<0






h(t) = 2(1 − t) 0≤t≤1




0

t>1

h(t)
2

0 0.5 1 t

ii)

The signals f (t) and g(t) are as given below,

f (t) g(t)
2

0 2 t 0 3 6 t

Case 1: For t < 3, we have the following:

12
 f (τ )
2
g(t − τ )
1

t−6 t−3 0 2 τ

Hence, h(t) will be zero for t < 3, as there is no overlap between g(t − τ ) and f (τ ).

Case 2: For 3 ≤ t ≤ 5,

f (τ )
2
g(t − τ )
1

t−6 0 t−3 2 τ

Here, h(t) will be,

t−3
(t − 3)2
Z
h(t) = τ dτ = .
0 2

Case 3: For 5 ≤ t ≤ 6,

f (τ )
2
g(t − τ )
1

t−6 0 2 t−3 τ

In this case, we get,

Z 2 Z t−3
h(t) = τ dτ + 2dτ = 2(t − 4).
0 2

Case 4: For 6 ≤ t ≤ 8,

13
 f (τ )
2
g(t − τ )
1

0 t−6 2 t−3 τ

Due to the above overlapping fashion, h(t) for 6 ≤ t ≤ 8 will be,

2 t−3
t2
Z Z
h(t) = τ dτ + 2dτ = − + 8t − 26.
t−6 2 2

Case 5: For t > 8, we obtain,

f (τ )
2
g(t − τ )
1

0 2 t−6 t−3 τ

Z t−3
h(t) = 2dτ = 6.
t−6





0 t≤3


(t−3)2

2 , 3 < t < 5,






Finally, the signal h(t) is given by, h(t) = 2(t − 4), 5≤t<6



2
 −t2 + 8t − 26, 6 ≤ t < 8,








6, t≥8

14
 9. Given, x[n] = 0, outside 0 ≤ n ≤ N − 1

y[n] = x[n] ∗ h[n]

∞
X
= x[k]h[n − k]
k=−∞
N
X −1
= x[k]h[n − k]
k=0

Now, substitute diiferent values for ’n’ and expand the sumation,

y[0] = x[0]h[0] + x[1]h[−1] + · · · + x[N − 1]h[−(N − 1)]

y[1] = x[0]h[1] + x[1]h[0] + · · · + x[N − 1]h[−N + 2]

..
.

y[N − 1] = x[0]h[N − 1] + x[1]h[N − 2] + · · · + x[N − 1]h[0]

We can write the above equations in matrix form (y=Hx) as follows,

    
y[0] h[0] h[−1] ··· h[−N + 1] x[0]
    
y[1] h[1] h[0] ··· h[−N + 2] x[1]
    
    
=
.. .. .. .. ..
  
..
.
    

 .  
  . . . 
 . 

y[N − 1] h[N − 1] h[N − 2] · · · h[0] x[N − 1]

10. (a) Given that f (t) ∗ g(t) = y(t). Hence,

Z∞
y(t) = f (τ )g(t − τ ) dτ.
−∞

Let us consider f (t − T1 ) ∗ g(t − T2 ), and by using the definition of convolution, we have

Z∞ Z∞
f (t − T1 ) ∗ g(t − T2 ) = f (τ − T1 )g(t − τ − T2 )dτ = f (τ − T1 )g(t − T2 − τ )dτ.
−∞ −∞

15
 Denote τ 0 = τ − T1 , note that the limits and derivative does not change.

Z∞ Z∞
0 0 0
f (τ 0 )g t − (T1 + T2 ) − τ 0 dτ 0



f (t − T1 ) ∗ g(t − T2 ) = f (τ )g t − T2 − (τ + T1 ) dτ =
−∞ −∞


= y t − (T1 + T2 ) [On comparing with the first equation].

(b) If u(t) ∗ u(t) = r(t), then

u(t + 1) − u(t − 2) ∗ u(t − 3) − u(t − 4)

= u(t + 1) ∗ u(t − 3) − u(t + 1) ∗ u(t − 4) + u(t − 2) ∗ u(t − 4) − u(t − 2) ∗ u(t − 3)

= r(t − 2) − r(t − 3) + r(t − 6) − r(t − 5).

The last equality is a consequence of the result obtained in (a). We now sketch r(t − 2) −
r(t − 3) + r(t − 6) − r(t − 5) in Fig. [1].

r(t − 2) − r(t − 3)

0
0 1 2 3 4 5 6 t

0 1 2 3 4 5 6 t
0

-1

r(t − 6) − r(t − 5)

r(t − 2) − r(t − 3) + r(t − 6) − r(t − 5)

0
0 1 2 3 4 5 6 t



Figure 1: Sketch of the signal u(t + 1) − u(t − 2) ∗ u(t − 3) − u(t − 4) using the distributive and
shift property of convolution.

16


One can verify the result by performing convolution of pulses u(t + 1) − u(t − 2) and


u(t − 3) − u(t − 4) , shown in Fig. [2].

u(t + 1) − u(t − 2) u(t − 3) − u(t − 4)

1
1
0
0
-2 -1 0 1 2 t 0 1 2 3 4 t



Figure 2: Signals x(t) = u(t + 1) − u(t − 2) and y(t) = u(t − 3) − u(t − 4) .

 
Hence, if x(t) = u(t + 1) − u(t − 2) and y(t) = u(t − 3) − u(t − 4), then x(t) ∗ y(t) is given
by,





0 t<2



t − 2 2 ≤ t < 3





x(t) ∗ y(t) = 1 3≤t≤5




6−t 5≤t≤6








0 t>6

11. Given: y(t) = f (t) ∗ g(t). Now, consider the following:

Z ∞ Z ∞
y(t) = f (τ )g(t − τ )dτ =⇒ y(ct) = f (τ )g(ct − τ )dτ
−∞ −∞

At the same time,

Z ∞
f (ct) ∗ g(ct) = f (cτ )g(ct − cτ )dτ
−∞

dτ 0
Case 1: Let c > 0. Then, c = |c|. Let τ 0 = cτ = |c|τ =⇒ dτ = |c| . Hence,

∞ ∞
dτ 0
Z Z
0 1 0 1
f (ct) ∗ g(ct) = f (τ )g(ct − τ ) = f (τ 0 )g(ct − τ 0 )dτ 0 = y(ct).
−∞ |c| |c| −∞ |c|

Case 2: Suppose c < 0, then c = −|c|. In which case, let τ 0 = cτ = −|c|τ =⇒ dτ 0 = − dτ

|c| .

Z −∞  dτ 0  Z ∞
0 0 1 1
f (ct) ∗ g(ct) = f (τ )g(ct − τ ) − = f (τ 0 )g(ct − τ 0 )dτ 0 = y(ct).
∞ |c| |c| −∞ |c|

1
Therefore, if y(t) = f (t) ∗ g(t), then f (ct) ∗ g(ct) = |c| y(ct), for all c 6= 0.

Fig. [3] shows f (t) and g(t).

17
 f (t) g(t)
1
1

−1 2 t 0 1 t

Figure 3: f (t)

R∞
Then, y(t) = f (τ )g(t−τ )dτ and g(t−τ ), as a function of τ , will be non-zero from τ = t−1
−∞
to τ = t, as shown in figure [4].

g(t − τ )

t−1 t τ

Figure 4: g(t − τ )

Case 1: If t < −1. Then, f (τ )g(t − τ ) = 0 in this range. Hence, y(t) = 0, ∀t < −1.

Case 2: If t ≥ −1 but t − 1 < −1, i.e., −1 ≤ t < 0.

Rt 2
Here, y(t) = (−τ + t) dτ = t2 + t + 12 .
−1
Case 3: If t < 2 but t − 1 ≥ −1, i.e., 0 ≤ t < 2.
Rt
Then, y(t) = (−τ + t) dτ = 0.5.
t−1
Case 4: If t ≥ 2 but t − 1 < 2, i.e., 2 ≤ t < 3.
R2 2
Now, y(t) = (−τ + t) dτ = − t2 + 2t − 32 .
t−1
Case 5: If t − 1 ≥ 2, f (τ )g(t − τ ) = 0 =⇒ y(t) = 0, ∀t ≥ 3.




0, t < −1


 t2

1


 2 + t + 2, −1 ≤ t < 0


y(t) = f (t) ∗ g(t) = 0.5, 0≤t<2



 2
− t2 + 2t − 32 , 2 ≤ t < 3








0, t≥3

18
 Using the result derived initially, we get,




0, t < −0.5



t2 + t + 0.25, −0.5 ≤ t < 0





1 
f (2t) ∗ g(2t) = y(2t) = 0.25, 0≤t<1
2 



−t2 + 2t − 0.75, 1 ≤ t < 1.5








0, t ≥ 1.5

12. x(t) is periodic with time period T : x(t + T ) = x(t)

For a time-invariant system h(t),

Z ∞
y(t) = x(τ )h(t − τ )dτ
Z−∞
∞
y(t + T ) = x(τ )h((t + T ) − τ )dτ
Z−∞
∞
= x(τ )h(t − (τ − T ))dτ
−∞

Let τ 0 = τ − T =⇒ dτ = dτ 0

Then,

Z ∞
y(t + T ) = x(τ 0 + T )h(t − τ 0 )dτ 0
Z−∞
∞
= x(τ 0 )h(t − τ 0 )dτ 0 (as x(t) is periodic)
−∞

y(t + T ) = y(t)

Hence, y(t) is periodic with time period T .

19