Assignment 1 Solution

ELE635 Communication Systems

• 2.1-2 Find the energies of the signal shown below. Comment on the effect on energy of sign
change, time shift, or doubling of the signal. What is the effect on the energy if the signal
is multiplied by k?

Solution: Let us denote the signal in question by g(t) and its energy by Eg . For parts (a)
and (b)
∫ 2π
1 ∫ 2π 1 ∫ 2π
Eg = sin2 tdt = dt − cos(2t) dt = π + 0 = π
0 2 0 2 0
∫ 4π
1 ∫ 4π 1 ∫ 4π
(c) Eg = sin2 tdt = dt − cos(2t) dt = π + 0 = π
2π 2 2π 2 2π
∫ 2π [ ∫ 2π ]
1 1 ∫ 2π
(d) Eg = 2
(2 sin t) dt = 4 dt − cos(2t) dt = 4[π + 0] = 4π
0 2 0 2 0
Sign change and time shift do not affect the signal energy. Doubling the signal quadruples
its energy. In the same way, we can show that the energy of kg(t) is k 2 Eg .

1
• 2.1-3 Find the average power of the signal shown below.

Solution: Both signals are periodic. The average power of φ(t) is

1∫T 2 1 ∫ π −t/2 2 1 − e−π

P = φ (t)dt = (e ) dt = .
T 0 π 0 π
The average power for wo (t) is

1 ∫ To 2 1 ∫ To
P = wo (t)dt = 1 · dt = 1.
To 0 To 0

• 2.1-8: Show that the power of a signal g(t) given by

∑
n
g(t) = Dk ejωk t ωi ̸= ωk for all i ̸= k
k=m

is (Parseval’s theorem)
∑
n
Pg = |Dk |2
k=m

Solution:
1 ∫ T /2 ∗ 1 ∫ T /2 ∑n ∑ n
Pg = lim g(t) · g (t)dt = lim Dk Dr∗ ej(ωk −ωr )t
T →∞ T −T /2 T →∞ T −T /2
k=m r=m

The integrals of the cross-product terms (when k ̸= r) are finite because the integrands are
periodic signals (made up of sinusoids). These terms, when divided by T → ∞, yields zero.
The remaining terms (k = r) yields

1 ∫ T /2 ∑n ∑n
lim |Dk |2 dt = |Dk |2
T →∞ T −T /2
k=m k=m

2
• 2.1-9: Determine the power and the rms value for each of the following signals:
(a) 10 cos(100t + π/3) (b) 10 cos(100t + π/3) + 16 sin(150t + π/5)
(c) (10 + 2 sin 3t) cos 10t (d) 10 cos 5t cos 10t
(e) 10 sin 5t cos 10t (f) ejαt cos ω0 t
Solution: (a) Power of a sinusoid of amplitude C is C 2 /2 regardless of its frequency (ω ̸= 0)
and phase. Therefore, in this case, P = (10)2 /2 = 50.
(b) Power of a sum of sinusoids is equal to the sum of the powers of the sinusoids. Therefore,
in this case, P = (10)2 /2 + (16)2 /2 = 178.
(c) (10 + 2 sin 3t) cos 10t = 10 cos 10t + sin 13t − sin 3t. Hence, P = (10)2 /2 + 1/2 + 1/2 = 51
(d) 10 cos 5t cos 10t = 5(cos 5t + cos 15t). Hence, P=25;
(e) 10 sin 5t cos 10t = 5(sin 15t − sin 5t). Hence, P=25;
(f) ejαt cos ω0 t = 1/2[ej(α+ω0 )t + ej(α−ω0 ) ]. We have P = 1/4 + 1/4 = 1/2

• 2.4-1: Simplify the following expressions:

( )
(a) tan(3t)
2t2 +1
δ(t − π/4)
( )
jω+π
(b) ω 2 +9π 2
δ(ω + π)
(c) [e−t sin(5t − π/6)]δ(t + π/15)
( )
(d) sin 0.5π(t+2)
t2 −4
δ(t − 1)
( )
cos(πt)
(e) t+2
δ(2t + 3)
( )
sin2 (kω)
(f) ω2
[δ(ω) − δ(ω + π/2)]

Solution: In these problems remember that impulse δ(x) is located at x = 0. Thus, an

impulse δ(t − τ ) is located at τ = t, and so on. Using the fact that

ϕ(t)δ(t) = ϕ(0)δ(t) and ϕ(t)δ(t − T ) = ϕ(T )δ(t − T )

we have
(a)
( )
tan(3π/4) −8
2
δ(t − π/4) = 2 δ(t − π/4)
π /8 + 1 π +8

3
(b)
( )
jω + π 1−j
2 2
δ(ω + π) = δ(ω + π)
ω + 9π 10π

(c)

[e−t sin(5t − π/6)]δ(t + π/15) = −eπ/15 δ(t + π/15)

(d)
( )
sin 0.5π(t + 2) sin 1.5π 1
δ(t − 1) = δ(t − 1) = δ(t − 1)
t −4
2 1−4 3

(e)
( )
cos(πt) cos(−3π/2)
δ(2t + 3) = δ(2t + 3) = 0
t+2 1/2

(f)
( )
sin2 (kω) 4 sin2 (kπ/2)
[δ(ω) − δ(ω + π/2)] = k 2
δ(ω) − δ(ω + π/2)
ω2 π2

where L’Hôspital’s rule was used for the first term as:

sin2 (kω) 2k sin(kω) cos(kω) k sin(kω) k 2 cos(kω)

lim = lim = lim = lim = k2
ω→0 ω2 ω→0 2ω ω→0 ω ω→0 1

4
• 2.4-4 Evaluate the following integrals:
∫∞
(a) −∞ g(−3τ + a)δ(t − τ )dτ
∫∞
(b) −∞ δ(τ )g(t − τ )dτ
∫∞
(c) −∞ δ(t + 2)e−jωt dt
∫1
(d) −∞ δ(t − 2) sin(πt) dt
∫∞
(e) −2 δ(2t + 3) e−4t dt
∫2
(f) −2 (t
3
+ 4)δ(1 − t) dt
∫∞
(g) −∞ g(2 − t)δ(3 − 0.5t) dt
∫∞
(h) −∞ cos π2 (x − 5) δ(3x − 1) dx

Solution: Using the fact that

∫ b
1
δ(at) = δ(t) and ϕ(t)δ(t − T )dt = ϕ(T )
|a| a
∫ ∞
g(τ )δ(t − τ )dτ = g(t)
−∞
∫ ∞
δ(τ )g(t − τ )dτ = g(t)
−∞

we have

(a) g(−3t + a)
(b) g(t)
(c) ej2ω
(d) 0
(e) e6 /2
(f) 5
(g) g(−4)
( )
7π
(h) cos 3

5
• 2.9-2: For each of the periodic signals below, find exponential Fourier series and sketch the
corresponding spectra.

Solutions: (from the Solution Manual using integration. can exercise to use the relationship
between Fourier Transform and Fourier Seriese)

(a) T0 = 4, f0 = 1/4. Also, D0 = 0 (by inspection),

1 ∫ 1 −j(nπ/2)t 1 ∫ 3 −j(nπ/2)t 2 nπ
Dn = e dt − e dt = sin |n| ≥ 1
4 −1 4 1 πn 2

6
(b) T0 = 10π, f0 = 1/10π, ω0 = 2π/10π = 1/5. Also, D0 = 1/5 (by inspection),
∞
∑ n
g(t) = Dn e j 5 t where
n=−∞

( ) ( )
1 ∫ π −j n t j nπ 1 nπ
Dn = e 5 dt = −2j sin = sin
10π −π 2πn 5 πn 5
(c)
∞
∑
g(t) = D0 + Dn ejnt where by inspection D0 = 0.5
n=−∞

1 ∫ 2π t −jnt j
Dn = e dt =
2π 0 2π 2πn
so that
1
|Dn | =
2πn
and 
 π, n>0
̸ 2
Dn =
 − π2 , n < 0

(d) T0 = π, f0 = 1/π, ω0 = 2 and D0 = 0.

∞
∑
g(t) = Dn ej2nt where
n=−∞

( )
1 ∫ π/4 4t −j(2n)t −j 2 πn πn
Dn = e dt = sin − cos
π −π/4 π πn πn 2 2
2π
(e) T0 = 3, ω0 = 3
. Also, D0 = 1/6 (by inspection),
∞
∑ 2πn
g(t) = Dn ej 3
t
where
n=−∞

[ ( ) ]
1 ∫ 1 −j 2πn t 3 j2πn
t e 3 dt = 2 2 e−j 3
2πn
Dn = +1 −1
3 0 4π n 3
Therefore, √
3 4π 2 n2 2πn 4πn 2πn
|Dn | = 2 2 2 + − 2 cos − sin
4π n 9 3 3 3
and ( )
−1
2πn
3
cos 2πn
3
− sin 2πn
3
̸ Dn = tan
cos 2πn
3
+ 2πn
3
sin 2πn
3
− 1

7
(f) T0 = 6, ω0 = π3 , and D0 = 0.5
∞
∑ πnt
g(t) = 0.5 + Dn e j 3

n=−∞

where
[∫ −1 ∫ 1 ∫ 2 ]
1 −j πnt −j πnt −j πnt
Dn = (t + 2)e 3 dt + e 3 dt + (−t + 2) e 3 dt
6 −2 −1 1
( )
3 nπ 2nπ
= cos − cos
π 2 n2 3 3

8
• 2.9-3: A periodic signal g(t) is expressed by the following Fourier series:
( ) ( )
2π π
g(t) = sin(2t) + cos 5t − + 2 cos 8t +
3 3
(a) Sketch the amplitude and phase spectra for the trigonometric series
(b) By inspection of spectra in part (a), sketch the exponential Fourier series spectra
(c) By inspection of spectra in part (b), write the exponential Fourier series for g(t).

Solution:

(a) For a compact trigonometric form, all terms must have cosine form and amplitudes
must be positive. For this reason, we rewrite g(t) as
( ) ( ) ( )
π π π π
g(t) = cos 2t − + cos 5t − − + 2 cos 8t +
2 6 2 3
( ) ( ) ( )
π 2π π
= cos 2t − + cos 5t − + 2 cos 8t +
2 3 3
Therefore, amplitudes C2 = 1, C5 = 1, C8 = 2 and phases θ2 = −π/2, θ5 = −2π/3, θ8 =
π/3.
(b) The exponential spectra is given as
(c) We rewrite g(t) to obtain the exponential Fourier series
[ ] [ ] [ ]
1 j (2t− π2 ) 1 j (5t− 2π3 )
+ e−j (2t− 2 ) + + e−j (5t− 3 ) + ej (8t+ 3 ) + e−j (8t+ 3 )
π 2π π π
g(t) = e e
2 2
( ) ( )
1 −j π j2t 1 −j 2π j5t ( j π ) j8t
= e 2 e + e 3 e + e 3 e +
2 2
( ) ( )
1 j π −j2t 1 j 2π −j5t ( −j π ) −j8t
e 2 e + e 3 e + e 3 e
2 2
Therefore,
1 jπ 1 π π
e 2 , D−2 = e−j 2 ̸ D−2 =
π
D−2 = ̸ D2 = −
2 2 2 2
1 j 2π 1 −j 2π 2π 2π
D−5 = e 5 , D5 = e 5 ̸ D−5 = ̸ D5 = −
2 2 3 3
−j π3 j π3 π π
D−8 = e , D8 = e ̸ D−8 = − ̸ D8 =
3 3

9
• 3.1-2 Consider the two signals shown below.

(a) Find the Fourier transforms of the signals.

(b) Find the Fourier transforms of g(−t) for the two signals.
(c) Compare the difference between the answers in (a) and (b).

Solution:
(a)
∫ 1 ∫ 2
−j2πf t 4 − 2e−j2πf − 2e−j4πf
Ga (f ) = 4e dt + 2e−j2πf t dt =
0 1 j2πf



 − τt , −τ ≤ t ≤ 0


gb (t) =

t
τ
, 0≤t≤τ


 0, otherwise

∫ 0 ∫ τ
t t −j2πf t 2
Gb (f ) = − e−j2πf t dt + e dt = [cos 2πf τ + 2πf τ sin 2πf τ − 1]
−τ τ 0 τ τ (2πf )2

(b)


 2, −2 ≤ t ≤ −1


ga (−t) =

4, −1 ≤ t ≤ 0


 0, otherwise

∫ −1 ∫ 0
4 − 2ej2πf − 2ej4πf
F[ga (−t)] = 2e−j2πf t dt + 4e−j2πf t dt = −
−2 −1 −j2πf

Since gb (−t) = gb (t), we have F[gb (−t)] = Gb (f ).

10
 (c)
We can observe that

F[ga (−t)] = Ga (−f ) and F[gb (−t)] = Gb (−f )

which is consistent with Fourier Transform property.

• 3.1-5

(a) Find the inverse Fourier transforms of the spectra shown below.
(b) Confirm the result by verifying the properties of Fourier transform.

Solution
(a)
∫ 1/2
ej2πf t 1/2 2
ga (t) = cos(πf ) ej2πf t df = (j2t cos πf + sin πf )−1/2 = cos(πt)
−1/2 π(1 − 4t )2 π(1 − 4t2 )

∫ 0 ( ) ∫ B( )
−f f
gb (t) = ej2πf t df + ej2πf t df
−B B 0 B
[ ] [ ]
(j2πf − 1)ej2πf t 0 (j2πf − 1)ej2πf t B
= |−B − |0
4Bπ 2 t2 4Bπ 2 t2
(−j2πBt − 1)e−j2πBt − (j2πBt − 1)ej2πBt − 2
=
4Bπ 2 t2
2 cos(2πBt) + 4πB sin(2πBt) − 2
=
4Bπ 2 t2

(b) Both ga (t) and gb (t) are real and even function of t and Ga (f ) and Gb (f ) are real and
even function of f , which are consistent with the properties of Fourier Transform.

11
• 3.1-7 Find the inverse Fourier transforms of G(f ) and G(3f ) for both spectra shown below.
Comment on the difference between the inverse Fourier transforms of G(f ) and G(3f ).

Solution
(a) Because 
 3f 2 , −B ≤ f ≤ B
G(f ) = 
0, otherwise
is an even function of f ,
∫ ∞ ∫ ∞
j2πf t
g(t) = G(f )e df = 2 3f 2 cos(2πf t) df
−∞ 0
[ ]f =B
2πf t cos(2πf t) + (2π 2 f 2 t2 − 1) sin(2πf t)
= 6
4π 3 t3 f =0
[ ]
2πBt cos(2πBt) + (2π B t − 1) sin(2πBt) 2 2 2
= 6
4π 3 t3

In comparison 
 27f 2 , −B/3 ≤ f ≤ B/3
G(3f ) =
 0, otherwise
Therefore,
∫ ∞ ∫ B/3
−1
F [G(3f )] = G(3f )e j2πf t
df = 2 27f 2 cos(2πf t) df
−∞ 0
[ ]f =B/3
2πf t cos(2πf t) + (2π 2 f 2 t2 − 1) sin(2πf t)
= 54
4π 3 t3 f =0
[ ]
36πBt cos(2πBt/3) + (12π B t − 54) sin(2πBt/3)
2 2 2
=
4π 3 t3

It is verified that ( )
1 t
F −1 [G(3f )] = g
3 3

12
(b) Because 

 1, −2 ≤ f ≤ −1




 2, −1 ≤ f ≤ 1
G(f ) =

 1, 1 ≤ f ≤ 2





0, otherwise
is an even function of f ,

∫ ∞ ∫ 1 ∫ 2
g(t) = G(f )ej2πf t df = 2 2 cos(2πf t) df + 2 cos(2πf t) df
−∞ 0 1
[ ]f =1 [ ]f =2
4 sin(2πf t) 2 sin(2πf t)
= +
2πt f =0
2πt f =1
sin(2πt) + sin(4πt)
=
πt

In comparison 

 1, −2/3 ≤ f ≤ −1/3




 2, −1/3 ≤ f ≤ 1/3
G(3f ) = 


 1, 1/3 ≤ f ≤ 2/3



0, otherwise
therefore,
∫ ∞ ∫ 1/3 ∫ 2/3
−1
F [G(3f )] = G(f )e j2πf t
df = 2 2 cos(2πf t) df + 2 cos(2πf t) df
−∞ 0 1/3
[ ]f =1/3 [ ]f =2/3
4 sin(2πf t) 2 sin(2πf t)
= +
2πt f =0
2πt f =1/3
sin(2πt/3) + sin(4πt/3)
=
πt
Once again it is verified that ( )
−1 1 t
F [G(3f )] = g
3 3

13
• 3.3-3 The Fourier transform of the triangular pulse g(t) in Fig. (a) below is given by
1
G(f ) = (ej2πf − j2πf ej2πf − 1)
(2πf )2
Using this information, and the time-shifting and time-scaling properties, to find the Fourier
Transforms of the signals g1 (t) to g5 (t) as shown in Figs. (b) to (f).

Solution:
Fig.(b) g1 (t) = g(−t) Then
1
G1 (f ) = G(−f ) = 2
[e−j2πf + j2πf e−j2πf − 1]
(2πf )

Fig.(c) g2 (t) = g(t − 1) + g1 (t − 1). Therefore

G2 (f ) = [G(f ) + G1 (f )]e−j2πf = [G(f ) + G(−f )]e−j2πf

e−j2πf j2πf e−j2πf −j2πf
= [e − j2πf e j2πf
− 1] + [e + j2πf e−j2πf − 1]
(2πf )2 (2πf )2
e−j2πf
= (2 cos 2πf − 2 + 4πf sin 2πf )
(2πf )2

Fig.(d) g3 (t) = g(t − 1) + g1 (t + 1). Therefore

G3 (f ) = G(f )e−j2πf + G(−f )ej2πf

e−j2πf j2πf ej2πf −j2πf
= [e − j2πf e j2πf
− 1] + [e + j2πf e−j2πf − 1]
(2πf )2 (2πf )2

14
 1
= 2
(2 − e−j2πf − ej2πf )
(2πf )
1
= [1 − cos(2πf )]
2(πf )2

Fig.(e) g4 (t) = g(t − 1/2) + g1 (t + 1/2). Therefore

G4 (f ) = G(f )e−jπf + G1 (f )ejπf

e−jπf j2πf ejπf
= [e − j2πf e j2πf
− 1] + [e−j2πf + j2πf e−j2πf − 1]
(2πf )2 (2πf )2
1 [ −jπf
] −j ( jπf −jπf
) 1
= 2
−j2πf e jπf
+ j2πf e = e − e = sin(πf )
(2πf ) 2πf πf

Fig.(f) g5 (t) can be obtained in three steps: (i) time-expanding g(t) by a factor 2; (ii) then
delaying it by 2 seconds; (iii) and multiplying it by 1.5 (we may interchange the sequence
for steps (i) and (ii)]. The first step (time expansion by a factor 2) yields
( )
t 1
g ⇐⇒ 2G(2f ) = 2
(ej4πf − j4πf ej4πf − 1)
2 2(2πf )

Second step of time delay of 2 secs, yields,

( )
t−2 1 1
g ⇐⇒ 2
(ej4πf − j4πf ej4πf − 1)e−j4πf = (1 − j4πf − e−j4πf )
2 2(2πf ) 2(2πf )2

The third step of multiplying the resulting signal by 1.5 yields

( )
t−2 3
g5 (t) = 1.5g ⇐⇒ (1 − j4πf − e−j4πf )
2 (4πf )2

15
• 3.3-5 Use the time-shifting property to show that if g(t) ⇐⇒ G(f ), then

g(t + T ) + g(t − T ) ⇐⇒ 2G(f ) cos 2πf T

Using this result, find the Fourier transforms of the signals shown below.

Solution:
From the time-shifting property,

g(t + T ) + g(t − T ) ⇐⇒ G(f )ej2πf T + G(f )e−j2πf T = 2G(f ) cos 2πf T

( )
t
(a) Let T = 3. Then g(t) = rect 2
, and G(f ) = 2sinc(2πf ). Then

g(t + 3) + g(t − 3) ⇐⇒ 4sinc(2πf ) cos(6πf )

( )
t
(b) Again, let T = 3 and take g(t) = ∆ 2
, and G(f ) = sinc2 (πf ), then

g(t + 3) + g(t − 3) ⇐⇒ 2sinc2 (πf ) cos(6πf )

16
• 3.3-7 The signals below are modulated signals with carrier cos(10t). Find the Fourier
transforms of these signals using the appropriate properties of the Fourier transform. Sketch
the amplitude and phase spectra for (a) and (b)

Solution:
( )
t
(a) We take g(t) = ∆ 2π
, we have
( )
t π[ ]
∆ cos 10t ⇐⇒ sinc2 (π 2 f − 5π) + sinc2 (π 2 f + 5π)
2π 2

2π
∆(t/2π)
1
t ω
−π π −4 −2 2 4

17
(b) The signal g(t) here is the same as part (a) delayed by 2π. From the time-shifting
property, its Fourier Transform is the same as in part (a) multiplied by e−j4π f . Therefore,
2

2[ ]
sinc2 (π 2 f − 5π) + sinc2 (π 2 f + 5π) e−j4π f
2
G(f ) =
π

The Fourier Transform in this case is the same as that in part (a) multiplied by e−j4π f . This
2

multiplying factor represents a linear phase spectrum −4π 2 f . Thus we have an amplitude
spectrum as in part (a) and a linear phase spectrum ̸ G(ω) = −4π 2 f .
( )
t
Note: In the foregoing solution, we first multiplied the triangle pulse ∆ 2π
by cos 10t and
then delayed the result by 2π. This means the signal can be expressed as
( )
t − 2π
∆ cos 10(t − 2π)
2π
We could have interchanged the operation in this particular case; that is, we could have
( )
t
first delayed the triangle pulse ∆ 2π
by 2π and then multiplied the result by cos 10t. In
this alternate procedure, the signal would be expressed as
( )
t − 2π
∆ cos 10t
2π
This interchange of operation is permissible here only because the sinusoid cos 10t executes
an integral number of cycles in the interval 2π. Thus both expressions are equivalent, since
cos 10(t − 2π) = cos 10t.
(c) In this case the signal is identical to that in (b), except that the basic pulse is the gate
pulse rect(t/2π) instead of a triangle pulse ∆(t/2π. We can have
π[ ]
sinc(2π 2 f − 10π) + sinc(2π 2 f + 10π) e−j4π f
2
G(f ) =
2

18
• 3.3-9 The process of recovering a signal g(t) from the modulated signal g(t) cos(2πf0 t + θ0 )
is called demodulation. Show that the signal g(t) cos(2πf0 t + θ0 ) can be demodulated by
multiplying it with 2 cos(2πf0 t + θ0 ) and passing the product through a lowpass filter of
bandwidth B Hz.
Solution: The modulated signal is g(t) cos(2πf0 t + θ0 ). Multiplying by 2 cos(2πf0 t + θ0 )
yields

2g(t) cos2 (2πf0 t + θ0 ) = g(t)[1 + cos(4πf0 t + 2θ0 )] = g(t) + g(t) cos(4πf0 t + 2θ0 )

Observe that, in the frequency domain, the resulting signal contains the original signal G(f )
and two copies of the G(f ) moved to the center frequencies of ±2f0 ,i.t., 0.5G(f ± 2f0 ). If
the bandwidth of the original signal is B < f0 , then the modulated copy will not extend
further than f0 from its center frequency and a low-pass filter from −f0 to f0 will pass only
g(t) and filter out the modulated copy.

19