Chapter 11: Discrete-Time Fourier Series and Transform

Problem 11.1:
2π 2π π
(i) x[k ] = k , for 0 ≤ k ≤ 5 and x[k + 6] = x[k ] . Ω0 = = = .
K0 6 3
5
1 1 5 − jnΩ0 k
Dn =
K0
∑ x[k ]e− jnΩ0k =
k =0
∑ ke
6 k =0
1 − jnΩ0
=
6
e ( )
+ 2e − j 2 nΩ0 + 3e − j 3nΩ0 + 4e − j 4 nΩ0 + 5e − j 5 nΩ0

1
= [ cos(nΩ0 ) + 2 cos(2nΩ 0 ) + 3cos(3nΩ 0 ) + 4 cos(4nΩ0 ) + 5cos(5nΩ0 )]
6
1
− j [sin(nΩ0 ) + 2sin(2nΩ0 ) + 3sin(3nΩ 0 ) + 4sin(4nΩ0 ) + 5sin(5nΩ 0 ) ]
6
1 5 r − ( M + 1) r M +1 + Mr M + 2 , we
∑ ( )
k M
Expressing Dn =
6 k =0
k e − jnΩ0 and using the series sum formula ∑ kr k =
k =0 (1 − r ) 2
can also represent Dn in a compact form as follows:

1 5 1 e − jnΩ0 − 6e − j 6 nΩ0 + 5e − j 7 nΩ0 π

Dn = ∑
6 k =0
k ( e )
− jnΩ0 k
=
6 (1 − e − jnΩ0 2
)
, with Ω0 = .
3
The magnitude and phase spectra for −10 ≤ n ≤ 10 are shown below.

⎧⎪ 1 (0 ≤ k ≤ 2)
(ii) x[k ] = ⎨0.5 (3 ≤ k ≤ 5) and x[k + 9] = x[k ]
⎪⎩ 0 (6 ≤ k ≤ 8)
2π 2π
Ω0 = = .
K0 9
2 Chapter 11

8
1 1 8 − jnΩ0 k
Dn =
K0
∑ x[k ]e− jnΩ0k =
k =0
∑ ke
9 k =0
1
=
9
(
1 + e − jnΩ0 + e − j 2 nΩ0 + 0.5e − j 3nΩ0 + 0.5e − j 4 nΩ0 + 0.5e − j 5 nΩ0 )
1
( )
= ⎡⎣1 + e − jnΩ0 + e − j 2 nΩ0 + 0.5e − j 3nΩ0 1 + e − jnΩ0 + e − j 2 nΩ0 ⎤⎦
9
1
( )(
= 1 + 0.5e − j 3nΩ0 1 + e − jnΩ0 + e − j 2 nΩ0
9
)
The magnitude and phase spectra for −10 ≤ n ≤ 10 are shown below.

(iii) As solved in Example 11.5, the DTFS coefficients are obtained as (see Eq. 11.19):

⎧− j 32 e j 4
π
for n = 1
⎪⎪ − jπ
Dn = ⎨ j 32 e 4 for n = −1 , Dn = Dn + 7
⎪
⎪⎩ 0 elsewhere.

The magnitude and phase spectra for −7 ≤ n ≤ 12 are shown below.

 Solutions 3

(iv) Because 2Ωπ = 52ππ/3 = 65 = rational , x[k ] is a periodic function with a fundamental period
2π 6
Ko = ⋅ m = ⋅ m = 6 (for m=5). Therefore, Ω0 = 2π = 2π = π .
5π / 3 5 K0 6 3
The function x[k ] can be expressed in terms of complex exponential functions as follows.
⎛ 5π π ⎞ jπ 5π K 0 −1 jnΩ0 k
j⎜ k+ ⎟
∑De
j k
x[k ] = 2e ⎝ 3 4⎠
= 2e 4
⋅e 3
= n
n =0
From the above equation, we can state that
⎧ jπ
⎪ 4 n=5 and Dn = Dn + 6 (since Dn is periodic with period 6).
Dn = ⎨2e
⎪⎩ 0 0≤n≤4

n=5 ⎧⎪π / 4 n=5 .

Note that Dn = ⎧⎪⎨2 and Dn = ⎨
⎪⎩0 0≤n≤4 ⎪⎩ 0 0≤n≤4
The amplitude and phase spectrums for −6 ≤ n ≤ 11 are shown below.

∞
(v) x[k ] = ∑ δ [k − 5m]
m =−∞

2π 2π
Ω0 = = .
K0 5
4
1 1 4 1
Dn =
K0
∑ x[k ]e− jnΩ0k =
k =0
∑
5 k =0
δ [k ]e− jnΩ0k =
5
The magnitude and phase spectra for −5 ≤ n ≤ 10 are shown below.
4 Chapter 11

⎛ 10π k ⎞ ⎛ 2π k ⎞ ⎛ 2π k ⎞ ⎛ 2π k ⎞ ⎛ 2π k ⎞ ⎛ 2π k ⎞
(vi) x[k ] = cos ⎜ ⎟ cos ⎜ ⎟ = cos ⎜ 4π k − ⎟ cos ⎜ ⎟ = cos ⎜ ⎟ cos ⎜ ⎟
⎝ 3 ⎠ ⎝ 5 ⎠ ⎝ 3 ⎠ ⎝ 5 ⎠ ⎝ 3 ⎠ ⎝ 5 ⎠
1 ⎛ 16π k ⎞ 1 ⎛ 4π k ⎞
= cos ⎜ ⎟ + cos ⎜ ⎟
2 ⎝ 15 ⎠ 2 ⎝ 15 ⎠
period =15 period =15
period =15

2π 2π
Ω0 = = .
K 0 15

1 ⎛ 16π k ⎞ 1 ⎛ 4π k ⎞ 1 ⎛ 14π k ⎞ 1 ⎛ 4π k ⎞
x[k ] = cos ⎜ ⎟ + cos ⎜ ⎟ = cos ⎜ 2π k − ⎟ + cos ⎜ ⎟
2 ⎝ 15 ⎠ 2 ⎝ 15 ⎠ 2 ⎝ 15 ⎠ 2 ⎝ 15 ⎠
1 ⎛ 14π k ⎞ 1 ⎛ 4π k ⎞ 1 1
= cos ⎜ ⎟ + cos ⎜ ⎟ = cos ( 7Ω0 k ) + cos ( 2Ω0 k )
2 ⎝ 15 ⎠ 2 ⎝ 15 ⎠ 2 2
1
= ⎡⎣e j 7 Ω0 k + e − j 7 Ω0 k + e j 2 Ω0 k + e− j 2 Ω0 k ⎤⎦
4
7
= ∑De
n =−7
n
jnΩ0 k

where Dn = ⎧⎪⎨1/ 4 n = ±2, ±7

⎪⎩ 0 otherwise

The magnitude and phase spectra for −8 ≤ n ≤ 14 are shown below.

 Solutions 5

⎧⎪ 1 k =0
(vii) x[k ] = cos ( 2π k 3) = ⎨ and x[k +3]=x[k ] .
⎪⎩0.5 k = 1,2
x[k ] is a periodic function with fundamental period of 3.
2π 2π
Ω0 = = .
K0 3
2
1
Dn =
K0
∑ x[k ]e
k =0
− jnΩ0 k

1
=
3
(
1 + 0.5e − jnΩ0 + 0.5e − j 2 nΩ0 )
⎧⎪2 / 3 n=0
=⎨
⎪⎩1/ 6 n = 1,2

The magnitude and phase spectra for −5 ≤ n ≤ 7 are shown below. ▌

6 Chapter 11

% MATLAB program for plotting Fourier spectra in problem 11.1

% part(i)
omega = pi/3 ;
%n = [-10:10] ;
n = [-4:4] ;

Dn = (1/6)*(exp(-j*n*omega)+2* exp(-j*2*n*omega)+3* exp(-j*3*n*omega)+4*

exp(-j*4*n*omega)+5* exp(-j*5*n*omega));

mag_Dn = abs(Dn);
phase_Dn = angle(Dn)
%pha_Dn = unwrap(phase_Dn)

stem(n, mag_Dn, 'filled'),grid

xlabel('n');
ylabel('Magnitude Spectrum')
axis([-11 11 0 3])
print -dtiff plot.tiff

stem(n, phase_Dn, 'filled'),grid

xlabel('n');
ylabel('Phase Spectrum')
%axis([-11 11 -pi pi])
print -dtiff plot.tiff

%
% part(ii)
omega = 2*pi/9 ;
%n = [-10:10] ;
n = [-4:4] ;

Dn = (1/9)*(1+0.5*exp(-j*3*n*omega)).*(1+exp(-j*n*omega)+exp(-
j*2*n*omega));

mag_Dn = abs(Dn);
phase_Dn = angle(Dn)

stem(n, mag_Dn, 'filled'),grid

xlabel('n');
ylabel('Magnitude Spectrum')
axis([-11 11 0 0.6])
print -dtiff plot.tiff

stem(n, phase_Dn, 'filled'),grid

xlabel('n');
ylabel('Phase Spectrum')
axis([-11 11 -pi pi])
print -dtiff plot.tiff

% part(iii)
%omega = 2*pi/7 ;
n = [-7:14] ;

g=-j*1.5*exp(j*pi/4);
h=j*1.5*exp(-j*pi/4);
Dn=[0 g 0 0 0 0 h 0 g 0 0 0 0 h 0 g 0 0 0 0 h 0]

mag_Dn = abs(Dn);
phase_Dn = angle(Dn) ;
 Solutions 7

stem(n, mag_Dn, 'filled'),grid

xlabel('n');
ylabel('Magnitude Spectrum')
axis([-7 12 -0.2 2.2])
print -dtiff plot.tiff

stem(n, phase_Dn, 'filled'),grid

xlabel('n');
ylabel('Phase Spectrum')
axis([-7 12 -pi pi])
print -dtiff plot.tiff

% part(iv)
omega = pi/3 ;
n = [-6:11] ;

mag_Dn = [0 0 0 0 0 2 0 0 0 0 0 2 0 0 0 0 0 2] ;
phase_Dn = [0 0 0 0 0 pi/4 0 0 0 0 0 pi/4 0 0 0 0 0
pi/4] ;

stem(n, mag_Dn, 'filled'),grid

xlabel('n');
ylabel('Magnitude Spectrum')
axis([-7 12 -0.2 2.2])
print -dtiff plot.tiff

stem(n, phase_Dn, 'filled'),grid

xlabel('n');
ylabel('Phase Spectrum')
axis([-7 12 -pi pi])
print -dtiff plot.tiff

% part(v)
n = [-5:10] ;

mag_Dn = (1/5)*ones(1,16) ;
phase_Dn = zeros(1,16) ;

stem(n, mag_Dn, 'filled'),grid

xlabel('n');
ylabel('Magnitude Spectrum')
axis([-6 11 -0.2 0.3])
print -dtiff plot.tiff

stem(n, phase_Dn, 'filled'),grid

xlabel('n');
ylabel('Phase Spectrum')
axis([-6 11 -pi pi])
print -dtiff plot.tiff

% part(vi)
n = [-8:14] ;

mag_Dn = [0 0.25 0 0 0 0 0.25 0 0 0 0.25 0 0 0 0 0.25 0 0 0

0 0 0.25 0] ;
phase_Dn = zeros(1,23) ;

stem(n, mag_Dn, 'filled'),grid

xlabel('n');
ylabel('Magnitude Spectrum')
axis([-9 15 -0.05 0.3])
print -dtiff plot.tiff
8 Chapter 11

stem(n, phase_Dn, 'filled'),grid

xlabel('n');
ylabel('Phase Spectrum')
axis([-9 15 -1 1])
print -dtiff plot.tiff

% part(vii)
n = [-5:7] ;

a=2/3; b=1/6;
mag_Dn = [b b a b b a b b a b b a b] ;
phase_Dn = zeros(1,13) ;

stem(n, mag_Dn, 'filled'),grid

xlabel('n');
ylabel('Magnitude Spectrum')
axis([-6 8 -0.05 0.75])
print -dtiff plot.tiff

stem(n, phase_Dn, 'filled'),grid

xlabel('n');
ylabel('Phase Spectrum')
axis([-6 8 -1 1])
print -dtiff plot.tiff
 Solutions 9

Problem 11.2

⎧1 ( 0 ≤ k ≤ 2)
⎪
(i) Dn = ⎨0.5 ( 3 ≤ k ≤ 5) and D n +9 = Dn
⎪0 ( 6 ≤ k ≤ 8)
⎩
2π
x[k ] is periodic with period 9. Therefore, Ω0 = .
9
8 2 5
x[k ] = ∑ Dn e jnΩ0k = ∑ e + 0.5∑ e
jn ( 29π ) k jn ( 29π ) k

n =0 n =0 n =3

+ 0.5 ⎡ e ⎤
j ( 29π ) k j ( 49π ) k j ( 69π ) k j ( 89π ) k j ( 109π ) k
= 1+ e +e +e +e
⎣ ⎦
j ( 29π ) k j ( 49π ) k j ( 69π ) k
⎡1 + e j ( 9 ) k + e j ( 9 ) k ⎤
2π 4π
= 1+ e +e + 0.5e
⎣ ⎦
(
= 1 + 0.5e
j ( 69π ) k
)(1 + e j ( 29π ) k
+e
j ( 49π ) k
)
⎧1 − j 0.5 ( n = −1)
⎪
⎪ 1 ( n = 0)
(ii) Dn = ⎨ and D n + 7 = Dn
⎪1 + j 0.5 ( n = 1)
⎪ 0
⎩ ( 2 ≤ n ≤ 5)
2π
x[k ] is periodic with period 7. Therefore, Ω0 = .
7
5
x[k ] = ∑De
n =−1
n
jnΩ0 k
= (1 − j 0.5)e − jΩ0k + (1 + j 0.5)e jΩ0k

= ( e − jΩ0k + e jΩ0k ) + j 0.5 ( e jΩ0k − e − jΩ0k )

= 2 cos(Ω0k ) − sin(Ω0k ) = 5 cos(Ω0k − tan −1 (0.5))
≈ 5 cos( 27π k − 0.4636)

Substituting, different values of k, we obtain one period of x[k ] ( 0 ≤ k ≤ 6 ) as follows:

x[k] = [2.0000, 2.0288, 0.5298, -1.3681, -2.2358, -1.4199, 0.4653]

(iii) Dn = 1 + 34 sin ( π8n ) , (0 ≤ n ≤ 6) and Dn + 7 = Dn

2π
x[k ] is periodic with period 7. Therefore, Ω0 = .
7
6 6
x[k ] = ∑ Dn e = ∑ (1 + 43 sin ( π8n ) ) e
jnΩ0 k jn ( 27π ) k

n =0 n =0
10 Chapter 11

6 6
For k = 0 , x[k ] = ∑ (1 +
n =0
3
4 sin ( πn
8 ) ) = 7 + ∑ sin ( π8n ) ≈ 7 + 43 * 4.6447 = 10.4835 .
3
4
n =0
For k ≠ 0 , x[k ] is obtained as follows:
6 6 6
x[k ] = ∑ (1 + 43 sin ( π8n ) ) e = ∑e + 43 ∑ sin ( π8n ) e
jn ( 27π ) k jn ( 27π ) k jn ( 27π ) k

n =0 n =0 n =0
=0
j ( 27π ) k .7 j 2π k
1− e 6
⎤ e jn ( 7 ) k = 1 − e 2π +
6
+ 43 ⋅ 21j ∑ ⎡ e ∑ ⎡⎣e ⎤
jn ( π8 ) − jn ( π8 ) 2π
jnπ ( 27k + 81 ) jnπ ( 27k − 81 )
= −e 3
−e
1− e
j ( 27π ) k
n =0
⎣ ⎦ 1− e 7
j ( )k j8
n =0
⎦
⎡ 1 − e jπ ( 7 + 8 ).7 1 − e jπ ( 7 − 8 ).7 ⎤
2k 1 2k 1
⎡ 1 − e j 7π / 8 1 − e − j 7π / 8 ⎤
= j8 ⎢
3
jπ ( 27k + 18 )
− jπ ( 2 k − 1 ) ⎥
= 3
j8 ⎢ jπ ( + )
2 k 1 − jπ ( 2 k − 1 ) ⎥
⎣⎢ 1 − e 1 − e 7 8 ⎦⎥ ⎣1 − e 7 8 1 − e 7 8 ⎦
⎡1.9239 − j 0.3827 1.9239 + j 0.3827 ⎤ 0.1435 + j 0.7215 0.1435 − j 0.7215
≈ 3
j8 ⎢ jπ ( 27k + 18 )
− jπ ( 27k − 18 ) ⎥≈− jπ ( 2 k + 1 )
− jπ ( 2 k − 1 )
⎣ 1− e 1− e ⎦ 1− e 7 8 1− e 7 8
( 0.1435 +
j 0.7215) ⎡1 − e 7 8 ⎤ + ( 0.1435 − j 0.7215) ⎡1 − e 7 8 ⎤
jπ ( − ) jπ ( + )
2k 1 2k 1

≈− ⎣ ⎦ ⎣ ⎦
⎡1 − e jπ ( 7 + 8 ) ⎤ ⎡1 − e jπ ( 7 − 8 ) ⎤
2k 1 2k 1

⎣ ⎦⎣ ⎦
0.2870 − ( 0.4087 + j 0.6117 ) e j 2π k / 7
− ( 0.4087 + j 0.6117 ) e j 2π k / 7
≈−
1 − 1.8478e j 2π k / 7 + e j 4π k / 7
0.2870 − 0.8174e j 2π k / 7
=−
1 − 1.8478e j 2π k / 7 + e j 4π k / 7

Substituting, different values of k, we obtain one period of x[k ] ( 0 ≤ k ≤ 6 ) as follows:

x[k] = [10.4835, -1.0626-j0.3735, -0.3844-j0.1220, -0.2948-j0.0341,
-0.2948+j0.0341, -0.3844+j0.1220, -1.0626+j0.3735].

(iv) Dn = ( −1) , (0 ≤ n ≤ 7) and Dn +8 = Dn

n

2π π
x[k ] is periodic with period 8. Therefore, Ω0 = = ..
8 4
1 − ( − e jπ k / 4 )
8
7 7
x[k ] = ∑ Dn e jnΩ0 k
= ∑ ( −e jπ k / 4 n
) = [ k ≠ 4]
n =0 n =0 1 + e jπ k / 4
1 − e j 2π k 0
= jπ k / 4
= j 2π k
= 0.
1+ e 1+ e 8
7 7 7

∑ ( −e jπ 4 / 4 ) = ∑ ( −e jπ ) = ∑ (1)n = 8.
n n
For k = 4 , x[4] =
n =0 n =0 n =0

Combining, the above result, we obtain,

⎧0 k = 0 − 3,5 − 7
x[k ] = ⎨ , and x[k + 8] = x[k ].
⎩8 k =4
 Solutions 11

(v) Dn = e jnπ / 4 , (0 ≤ n ≤ 7) and Dn +8 = Dn

2π π
x[k ] is periodic with period 8. Therefore, Ω0 = = .
8 4
7 7 7
1 − e j 8(π / 4)( k +1)
x[k ] = ∑ Dn e jnΩ0k = ∑ e jnπ / 4e jn (π / 4) k = ∑ e jn (π / 4)( k +1) = [ k ≠ 7]
n =0 n =0 n =0 1 − e j (π / 4)( k +1)
1 − e j 2π ( k +1) 0
= j (π / 4)( k +1)
= j ( π / 4)( k +1)
= 0.
1− e 1− e
7 7 7
For k = 7 , x[7] = ∑ e jn (π / 4)(7+1) = ∑ e j 2nπ = ∑1 = 8.
n =0 n =0 n =0

Combining, the above result, we obtain,

⎧0 0 ≤ k ≤ 6
x[k ] = ⎨ , and x[k + 8] = x[k ]. ▌
⎩8 k =7

% Problem 11.2, MATLAB Calculation and Verification

% Part (ii)
k = 0:6
x2 = sqrt(5)*cos(2*pi*k/7-0.4636);
x2 = [2.0000, 2.0288, 0.5298, -1.3681, -2.2358, -1.4199, 0.4653]

% Part (iii)
x3_1= zeros(1,7);
x3_2= zeros(1,7);
x3_2(1)= 10.4835;
for k = 0:6
n = 0:6;
a =1+(3/4)*sin(n*pi/8);
b = exp(j*n*2*pi*k/7);
x3_1(k+1) = sum(a.*b);
end
x3_1
%
for k = 1:6
a = exp(j*2*pi*k/7);
b = 0.2870 - 0.8174*a;
c = 1-1.8478*a + a.*a;
x3_2(k+1) = -b./c ;
end
x3_2
x3 = 10.4835, -1.0626-j0.3735, -0.3844-j0.1220, -0.2948-j0.0341,
-0.2948+j0.0341, -0.3844+j0.1220, -1.0626+j0.3735

% Part (iv)
x4= zeros(1,8);
for k = 0:7
n = 0:7;
b = (-exp(j*pi*k/4)).^n ;
x4(k+1) = sum(b);
end
% x4 = [0 0 0 0 8 0 0 0] ;
12 Chapter 11

% Part (v)
x5= zeros(1,8);
for k = 0:7
n = 0:7;
b = exp(j*n*pi*(k+1)/4) ;
x5(k+1) = sum(b);
end
x5
% x5 = [0 0 0 0 0 0 0 8] ;

Problem 11.3

∞ ∞ ∞ M
(i) ∑
k =−∞
x[k ] = ∑
k =−∞
2= ∑ 2 = 2 + lim 2∑ 2 = 2 + lim 4M = ∞ Î The DTFT does not exist.
k =−∞
M →∞
k =1
M →∞

∞ 2
(ii) ∑
k =−∞
x[k ] = ∑ ( 3 − k ) = 1 + 2 + 3 + 2 + 1 = 9 < ∞ Î The DTFT exists
k =−2

∞ ∞ ∞ ∞ ∞
1/ 3 3
(iii) ∑
k =−∞
x[k ] = ∑
k =−∞
k3
−k
= 2∑ k 3
k =1
−k
= 2∑ k 3− k = 2 × ∑ k (1/ 3) k = 2 ×
k =1 k =1 (1 − 1/ 3) 2
= <∞
2
= even

Therefore, the DTFT exists.

∞ ∞ ∞
(iv) ∑
k =−∞
x[k ] = ∑
k =−∞
α k cos (ω0k ) u[k ] = ∑ α k cos (ω0k )
k =0

∞
< ∑ α k cos (ω0k ) [ Using Schwartz's inequality]
k =0
∞
≤ ∑α ⎡∵ cos (ω0k ) ≤ 1⎤
k
⎣ ⎦
k =0

1
= <∞ ⎡⎣∵ α < 1⎤⎦
1− α

Therefore, the DTFT exists.

∞ ∞ ∞ ∞
(v) ∑
k =−∞
x[k ] = ∑
k =−∞
α k sin (ω0k + φ ) u[k ] = ∑ α k sin (ω0k + φ ) < ∑ α k sin (ω0k + φ )
k =0 k =0
≤1
∞
≤ ∑α
k

k =0

1
= <∞ ⎡⎣∵ α < 1⎤⎦
1− α

Therefore, the DTFT exists.

 Solutions 13

∞ ∞
sin ( π5k ) sin ( π7k ) ∞
(vi) ∑
k =−∞
x[k ] = ∑
k =−∞ π 2k 2
= x[0] + 2 ∑
k =1
x[k ]

= even

sin ( π5k ) sin ( π7k ) sin ( π5k ) sin ( π7k ) 1

Note that x[0] = = = . Therefore,
π k2 2
k =0
35. πk
5 ⋅ πk
7 k =0
35

∞ ∞ ∞
sin ( π5k ) sin ( π7k ) ∞ sin ( π5k ) sin ( π7k )
∑
k =−∞
x[k ] = 1
35 + 2∑ x[k ] =
k =1
1
35 + 2∑
k =1 π 2k 2
≤ 1
35 + 2∑
k =1 π 2k 2
∞
1
≤ 1
+ 2∑ ⎡∵ sin ( π5k ) , sin ( π7k ) ≤ 1⎤
⎣ ⎦
k =1 π k
35 2 2

∞
1
= 1
35 + π22 ∑ 2
<∞
k =1 k

∞
(vii) x[k ] = ∑ δ [k − 5m − 3] is a periodic function with period 5. The sum in any one period,
m =−∞
∞

∑ x[k ] = 1 . Therefore ∑
<5> k =−∞
x[k ] is infinite, and hence the DTFT does not exist.

(viii) x[k ] is a periodic function with period 7. The sum in any one period, ∑ x[k ] = 9 . Therefore
<7>
∞

∑
k =−∞
x[k ] is infinite, and hence the DTFT does not exist.

∞ ∞ ∞

∑ ∑ ∑ 1 = ∞ Î The DTFT does not exist.

0
(ix) x[ k ] = e j (0.2π k + 45 ) =
k =−∞ k =−∞ k =−∞
=1
−k j (0.2π k + 450 )
(x) x[ k ] = k 3 u[ k ] + e . Although, the aperiodic component decays very quickly, the periodic
= aperiodic = periodic
∞
component oscillates forever. Therefore ∑
k =−∞
x[k ] is infinite, and hence the DTFT does not exist. ▌

Problem 11.4

(a)
(i) x[k ] = 2 is a periodic signal with period K 0 = 1 . Therefore, we first need to calculate the DTFS
coefficients, and then use Eq. (11.36a) to obtain the DTFT.

From Table 11.2, we obtain Dn = 2 . Using, Eq. (11.36a), the DTFT is obtained as
14 Chapter 11

∞ ∞
X ( Ω ) = 2π ∑
n =−∞
(
Dnδ Ω − 2Kπ0n = 4π ) ∑ δ ( Ω − 2π n )
n =−∞

∞ 2 2
(ii) X (Ω ) = ∑
k =−∞ = even
x[k ] ⋅ e − jΩk = ∑
k =−2
x[k ] ⋅ cos(Ωk ) − j ∑ x[k ] ⋅ sin(Ωk )
k =−2
= even function = odd function

=0
2 2
= x[0] + 2∑ x[k ] ⋅ cos(Ωk ) = 3 + 2∑ (3 − k ) ⋅ cos(Ωk )
k =1 k =1

= 3 + 4 cos(Ω) + 2 cos(2Ω)

∞ ∞ −1 ∞
(iii) X (Ω) = ∑ x[k ] ⋅ e
k =−∞
− jΩk
= ∑ k3
k =−∞
−k
e − jΩk
= ∑ k3 e
k =−∞
k − jΩk
+ ∑ k 3− k e− jΩk
k =1

∞ ∞ ∞ ∞
= − ∑ k 3− k e jΩk + ∑ k 3− k e − jΩk = ∑ k ( 13 e − jΩ ) − ∑ k ( 13 e jΩ )
k k

k =1 k =1 k =1 k =1
− jΩ
1
e 1
e jΩ
⎡ ∞ k r ⎤
= − ⎢∵ ∑ kr = (1 − r )2 ⎥
3 3

(1 − e − jΩ )
(1 − e )
2 jΩ 2
1
3
1
3 ⎣ k =1 ⎦
1
3 e (1 − e ) − e (1 − e )
− jΩ 1
3
jΩ 2 1
3
jΩ 1
3
− jΩ 2

=
(1 − e ) (1 − e ) − jΩ 2 jΩ 2
1 1
3 3

1
3 e (1 − e + e ) − e (1 − e
− jΩ 2
3
jΩ 1
9
j 2Ω 1
3
jΩ 2
3
− jΩ
+ 91 e − j 2Ω )
=
(1 − e − e + ) − jΩ jΩ 2
1 1 1
3 3 9

=
8
( e − e ) = − j sin(Ω) =
27
− jΩ jΩ
16
27 − 16
27 j sin(Ω)

( 109 − 23 cos(Ω) ) ( 109 − 23 cos(Ω) ) 81 − 27 cos(Ω) + 9 cos (Ω)

2 2 100 40 4 2

− 16
27 j sin(Ω) − 16
27 j sin(Ω)
= =
81 − 27 cos(Ω) + 9 [1 + cos(2Ω) ] 81 − 27 cos(Ω) + 9 cos(2Ω)
100 40 2 118 40 2

∞ ∞ ∞
(iv) X (Ω) = ∑
k =−∞
x[k ] ⋅ e− jΩk = ∑ α k cos (ω0 k ) u[k ]e− jΩk = ∑α k cos(ω0 k )e− jΩk
k =−∞ k =0
∞ ∞
= 1
2 ∑α
k =0
k
[e jω0 k + e− jω0k ]e− jΩk = 1
2 ∑α
k =0
k
⎡⎣e − j (Ω+ω0 )k + e − j (Ω −ω0 )k ⎤⎦
∞ ∞
1/ 2 1/ 2
∑ (α e ) (
+ 12 ∑ α e − j (Ω −ω0 ) )
k k
− j (Ω+ω0 )
= 1
= − j (Ω+ω0 )
+
2
k =0 k =0 1−αe 1 − α e− j (Ω −ω0 )

=
1 − α2 e ( − j (Ω+ω0 )
+e − j (Ω −ω0 )
) =
1 − α e − jΩ cos(ω0 )
(1 − α e − j (Ω+ω0 )
)(1 − α e − j (Ω −ω0 )
) (1 − α e − j (Ω+ω0 )
)(1 − α e − j (Ω −ω0 )
)
 Solutions 15

∞ ∞ ∞
(v) X (Ω) = ∑ x[k ] ⋅ e
k =−∞
− jΩk
= ∑α
k =−∞
k
sin (ω0 k + +φ ) u[k ]e− jΩk = ∑ α k sin (ω0 k + φ ) e− jΩk
k =0
∞ ∞
= 1
2j ∑α
k =0
k
[e j (ω0 k +φ ) − e− ( jω0 k +φ ) ]e− jΩk = 1
2j ∑α
k =0
k
⎡⎣e− j (Ω+ω0 )k e jφ − e − j (Ω −ω0 )k e− jφ ⎤⎦

∞ ∞ 1
e jφ 1
e − jφ
∑ (α e ) ∑ (α e )
k k
jφ − j (Ω+ω0 ) − jφ − j (Ω −ω0 ) 2j 2j
= 1
e − 1
e = −
2j
k =0
2j
k =0 1 − α e− j (Ω+ω0 ) 1 − α e − j (Ω −ω0 )
sin(φ ) + 2αj ⎡⎣e− j (Ω+ω0 +φ ) − e − j (Ω −ω0 −φ ) ⎤⎦ sin(φ ) − 2αj e− jΩ ⎡⎣e j (ω0 +φ ) − e − j (ω0 +φ ) ⎤⎦
= =
(1 − α e − j (Ω+ω0 )
)(1 − α e − j (Ω −ω0 )
) (1 − α e − j (Ω+ω0 )
)(1 − α e − j (Ω −ω0 )
)
sin(φ ) − α e − jΩ sin(ω0 + φ )
=
(1 − α e − j (Ω+ω0 )
)(1 − α e − j (Ω −ω0 )
)
sin(Wk ) DTFT ⎪⎧1 Ω ≤ W
(vi) Using the DTFT pair ←⎯⎯→ ⎨ , we obtain the following:
πk ⎪⎩0 W < Ω ≤ π

sin( π5 k ) DTFT ⎪⎧1 Ω ≤ π5

; sin( π7 k ) DTFT ⎪⎧1 Ω ≤ π7
←⎯⎯→ X 1 (Ω) = ⎨ ←⎯⎯→ X 2 (Ω) = ⎨
πk ⎪⎩0
π
5 < Ω ≤π πk ⎪⎩0
π
7 < Ω ≤π

Using the frequency convolution property of the DTFT, we obtain

π
⎡ ⎧ sin( π5 k ) ⎫ ⎧ sin( π7 k ) ⎫⎤ 1
X (Ω) = ℑ ⎢ ⎨
⎣ ⎩ π k
⎬× ⎨
⎭ ⎩ π k
⎬⎥ =
⎭⎦ 2π
∫π X (θ ) X
−
1 2 (Ω − θ )dθ

Using the graphical convolution method, it can easily be shown that

⎧ Ω+ 1235π − 1235π ≤ Ω< − 235π

⎪ 12π
π
⎪ 35 − 235π ≤ Ω< 235π
∫ X 1 (θ ) X 2 (Ω − θ )dθ = ⎨
⎪Ω − 35 ≤ Ω< 1235π
12π 2π
−π 35
⎪
⎩ 0 otherwise
Therefore, the DTFT X (Ω) is obtained as:

⎧ 21π ( Ω+ 1235π ) − 1235π ≤ Ω< − 235π

⎪
⎪ 1
− 235π ≤ Ω< 235π
with X (Ω) = X (Ω+2π )
7
X (Ω) Ω ≤π =⎨
⎪ 21π ( Ω − 35 ) 35 ≤ Ω< 35
12π 2π 12π

⎪
⎩ 0 otherwise
The DTFT X (Ω) is plotted below.
16 Chapter 11

% MATLAB program for plotting Fourier spectra in problem 11.4(vi)

X=zeros(1,length([-pi:0.01:pi]));
k = 1:10000 ;

ind=1;
c=2*pi*pi;
for w = -pi:0.01:pi
X(ind)=1/35;
w1=w+2*pi/35; w2=w-2*pi/35; w3=w+12*pi/35; w4=w-12*pi/35;
for k = 1:10000
X(ind)=X(ind)+(cos(w1*k)+cos(w2*k)-cos(w3*k)-cos(w4*k))/(c*k*k);
end
ind = ind+1;
end
%
mag_X = abs(X);
phase_X = angle(X)
%pha_X = unwrap(phase_X)

Omega = -pi:0.01:pi
plot(Omega, mag_X),grid
xlabel('Omega');
ylabel('Magnitude Spectrum')
%axis([-pi pi 0 3])
print -dtiff plot.tiff

plot(Omega, phase_X),grid
xlabel('Omega');
ylabel('Phase Spectrum')
%axis([-pi pi -pi pi])
print -dtiff plot.tiff

∞
(vii) x[k ] = ∑ δ [k − 5m − 3] is a periodic function with period 5. Therefore, we first need to calculate
m =−∞
the DTFS coefficients, and then use Eq. (11.36a) to obtain the DTFT.
2π 2π
Ω0 = = .
K0 5
 Solutions 17

1 4 1 4 1 1 − j 6 nπ
Dn = ∑
K 0 k =0
x[ k ]e − jnΩ0k
= ∑
5 k =0
δ [k − 3]e − jnΩ0k = e − j 3nΩ0 = e 5
5 5

Using, Eq. (11.36a), the DTFT is obtained as

∞ ∞
1 − j 56 nπ 2π ∞

( )
− j 6 nπ

X ( Ω ) = 2π ∑
n =−∞
Dnδ Ω − 2Kπ0n = 2π ∑
n =−∞ 5
e δ ( Ω − 2π n ) =
5
∑e
n =−∞
5
δ ( Ω − 2π n )

(viii) x[k ] = ⎧⎨3 − k k ≤3

and x[k + 7] = x[k ]
⎩ 0 otherwise

x[k ] is a periodic (and even) function. Fundamental period = 7. Therefore, Ω 0 = 2π . The DTFS
7
coefficients are calculated as follows.

1 3 1 3 j 4 1 3
Dr = ∑ x[k ] ⋅ e − jrΩ0k = ∑ x[k ] ⋅ cos( rΩ0k ) − ∑ x[k ] ⋅ sin( rΩ0k ) = ∑ ( 3 − k ) ⋅ cos( rΩ0k )
7 k =−3 7 k =−3 7 k =−4 7 k =−3
= even function =odd function
=0

1⎡ 3
⎤ 3 2
= ⎢ 3 + 2∑ ( 3 − k ) ⋅ cos( rΩ0k ) ⎥ = + [2 cos( rΩ0 ) + cos(2rΩ0 )]
7⎣ k =1 ⎦ 7 7
3 4 ⎛ 2π r ⎞ 2 ⎛ 4π r ⎞
= + cos ⎜ ⎟ + cos ⎜ ⎟
7 7 ⎝ 7 ⎠ 7 ⎝ 7 ⎠
1⎡ ⎛ 2π r ⎞ ⎛ 4π r ⎞ ⎤
= ⎢ 3 + 4 cos ⎜ ⎟ + 2 cos ⎜ ⎟⎥
7⎣ ⎝ 7 ⎠ ⎝ 7 ⎠⎦

with Ω 0 = 2π .
7
The DTFT is then given by the following expression.
∞
⎛ ∞
⎧3 2π r ⎞ 2 ⎛ 4π r ⎞ ⎫
∑ D δ ( Ω − ) = 2π ∑ ⎨⎩ 7 + 7 cos ⎜⎝
4
X ( Ω ) = 2π ⎟ + cos ⎜ ⎟⎬δ (Ω − 7 )
2π r 2π r
r K0
r =−∞ r =−∞ 7 ⎠ 7 ⎝ 7 ⎠⎭
0
(ix) x[k ] = e j (0.2π k + 45 )

Ω = 0.2π , and K0 = 2π = 2π = 10 . In other words, x[ k ] is a periodic signal with periodicity 10, and
Ω 0.2π
2π 2π
Ω0 = = = 0.2π .
K 0 10
1 + j j 0.2π k 9
= ∑ Dr e jrΩ0 k
0
x[ k ] = e j (0.2π k + 45 ) = e jπ / 4 e j 0.2π k = e
2 r =0

⎧1 + j
⎪ r =1
Comparing the two left-most expressions we obtain Dr ,0≤ r ≤ 9 =⎨ 2 .
⎪ 0
⎩ otherwise
18 Chapter 11

⎧1 + j
n = 1 ± 10m 1 + j ∞
As Dn = Dn +10 , Dn = ⎪⎨ 2 = ∑ δ [n − 10m]
⎪ 0 2 m =−∞
⎩ otherwise
The DTFT is therefore given by the following expression.
⎧1 + j
∞
⎪ n = 1 + 10m
X ( Ω ) = 2π ∑ Dnδ Ω − K02 nπ
( where Dn = ⎨ 2)
n =−∞ ⎪⎩ 0 otherwise
∞
= 2π (1 + j ) ∑ δ ( Ω − 210π − 2π m )
m =−∞

0
(x) x[ k ] = k 3− k u[k ] + e j (0.2π k + 45 )

X (Ω) = ℑ{k 3− k u[k ]} + ℑ e j (0.2π k +45 ) { 0

} [applying the linearity property]

1
e − jΩ ∞
= 3
+ 2π (1 + j ) ∑ δ (Ω − 210π − 2π m )
(1 − e − jΩ )
2
1 m =−∞
3

(b)
(i) x[k ] = k , for 0 ≤ k ≤ 5 and x[k + 6] = x[k ]

1 e − jnπ / 3 − 6e − j 6 nπ / 3 + 5e − j 7 nπ / 3
Dn =
6 (1 − e − jnπ / 3 ) 2
∞
Using, Eq. (11.36a), the DTFT is obtained as, X ( Ω ) = 2π ∑ D δ (Ω − π ) .
n =−∞
n
2 n
6

⎧⎪ 1 (0 ≤ k ≤ 2)
(ii) x[k ] = ⎨0.5 (3 ≤ k ≤ 5) and x[k + 9] = x[k ]
⎪⎩ 0 (6 ≤ k ≤ 8)
2π 2π
Ω0 = = .
K0 9
1
Dn =
9
(1 + 0.5e − j 6 nπ / 9 )(1 + e − j 2 nπ / 9 + e − j 4 nπ / 9 )
∞
Using, Eq. (11.36a), the DTFT is obtained as, X ( Ω ) = 2π ∑ D δ (Ω − π ) .
n =−∞
n
2 n
9

(iii) x[k ] = 3 sin ( 27π k + π

4
)
 Solutions 19

⎧− j 32 e j 4
π
for n = 1
⎪⎪ − jπ
Dn = ⎨ j 32 e 4 for n = −1 , Dn = Dn + 7
⎪
⎪⎩ 0 elsewhere.
∞
Using, Eq. (11.36a), the DTFT is obtained as, X ( Ω ) = 2π ∑ D δ (Ω − π ) .
n =−∞
n
2 n
7

(iv) x[k ] = 2e
j (53π k + π4 )

⎧ jπ
⎪ 4 n=5 and Dn = Dn + 6
Dn = ⎨2e
⎪⎩ 0 0≤n≤4
∞
Using, Eq. (11.36a), the DTFT is obtained as, X ( Ω ) = 2π ∑ D δ (Ω − π ) .
n =−∞
n
2 n
6

∞
(v) x[k ] = ∑ δ(k − 5m)
m = −∞
1
Dn =
5
∞
2π ∞
Using, Eq. (11.36a), the DTFT is obtained as, X ( Ω ) = 2π ∑
n =−∞
Dnδ ( Ω − 2π6 n ) =
5
∑ δ (Ω − π ) .
n =−∞
2 n
5

(vi) x[k ] = cos(10πk 3) cos(2πk 5)

⎧⎪1/ 4 n = ±2, ±7 , D = D
Dn = ⎨ n n +15
⎪⎩ 0 otherwise
∞
Using, Eq. (11.36a), the DTFT is obtained as, X ( Ω ) = 2π ∑ D δ (Ω − π ) .
n =−∞
n
2 n
15

(vii) x[k ] = cos(2πk 3) .

⎧⎪2 / 3 n=0 , D = D
Dn = ⎨ n +3
n = 1,2
n
⎪⎩1/ 6
Using, Eq. (11.36a), the DTFT is obtained as,
∞
X ( Ω ) = 2π ∑ D δ (Ω − π )
n =−∞
n
2 n
3

4π ∞
π ∞
⎧ ⎛ 2π ⎞ ⎛ 4π ⎞ ⎫
=
3
∑ δ ( Ω − 2π m ) +
m =−∞
∑ ⎨⎩δ ⎜⎝ Ω − 2π m −
3 m =−∞ 3
⎟ + δ ⎜ Ω − 2π m −
⎠ ⎝
⎟⎬
3 ⎠⎭
▌
20 Chapter 11

Extras – Spectrum Plots for DTFTs in Problem 11.4

11.4(a)(ii) X ( Ω ) = 3 + 4cos(Ω) + 2 cos(2Ω)

X (Ω ) = 3 + 4 cos(Ω) + 2 cos(2Ω)

⎧⎪ 0 3 + 4 cos(Ω) + 2 cos(2Ω) ≥ 0
X (Ω ) = ⎨
⎪⎩π 3 + 4 cos(Ω) + 2 cos(2Ω) < 0

∞
⎧3 4 ⎛ 2π r ⎞ 2 ⎛ 4π r ⎞ ⎫
X ( Ω ) = 2π ∑ ⎨ 7 + 7 cos ⎜⎝ ⎟ + cos ⎜ ⎟⎬ δ (Ω − 7 )
2π r
11.4(a)(viii)
r =−∞ ⎩ 7 ⎠ 7 ⎝ 7 ⎠⎭

2π ∞
⎛ 2π r ⎞ ⎛ 4π r ⎞
X (Ω ) = ∑ 3 + 4cos ⎜⎝ ⎟ + 2 cos ⎜ ⎟ δ (Ω − 7 )
2π r
7 r =−∞ 7 ⎠ ⎝ 7 ⎠

⎧0 Ω≠ 2π r
7
⎪⎪
X (Ω ) = ⎨ 0 3 + 4 cos(Ω) + 2 cos(2Ω) ≥ 0, Ω= 2π7 r
⎪
⎪⎩π 3 + 4 cos(Ω) + 2 cos(2Ω) < 0, Ω= 2π7 r
 Solutions 21

∞
(ix) X ( Ω ) = 2π (1 + j ) ∑ δ ( Ω − 210π − 2π m )
m =−∞

∞
X ( Ω ) = 2π ∑ δ ( Ω − π − 2π m )
m =−∞
5

⎧π
⎪ Ω = π5 − 2π m, m ∈
X (Ω) = ⎨ 2
⎪⎩ 0 otherwise

The DTFT spectrum for one period is shown below.

Amplitude 2π
Spectrum

−π 0 π
5
π Ω

Phase
Spectrum π /2

−π 0 π
5 π Ω
22 Chapter 11

1
e − jΩ ∞
(x) X (Ω) = 3
+ 2π (1 + j ) ∑ δ (Ω − 210π − 2π m)
(1 − e − jΩ )
2
1 m =−∞
3

Note that there are two components of X (Ω) . At Ω = − 2π m , the Diract delta function will have
2π
10
infinite value, and the first component can be ignored. At other values of Ω , the second component will
have zero magnitude, and value of X (Ω) will be equal to the first component.

1
e − jΩ 1
e − jΩ 1
3
Note that 3
= 3
= 3
= .
(1 − − jΩ 2
) (1 − 13 cos(Ω) + j 13 sin(Ω) ) (1 − 13 cos(Ω) ) 10 − 6cos(Ω)
2 2
1
3 e + 19 sin (Ω)
2

⎧2πδ ( Ω − 210π − 2π m ) Ω = 210π − 2π m

⎪
Therefore, X ( Ω ) ≈ ⎨ 3
⎪ 10 − 6cos(Ω) otherwise
⎩

⎡ 1 − jΩ ⎤
X (Ω ) = ⎢ 3e ⎥ + ⎡ 2π (1 + j )δ ( Ω − 210π − 2π m ) ⎤
⎣ ⎦
(
⎢ 1 − 1 e − jΩ
) ⎥
2

⎣ 3 ⎦
⎧π −1 ⎛ sin(Ω ) ⎞
⎪ − Ω − 2 tan ⎜ ⎟ Ω = 10 − 2π m, m ∈
2π

⎪2 ⎝ 3 − cos(Ω ) ⎠
=⎨
⎪ −1 ⎛ sin(Ω ) ⎞
⎪ −Ω − 2 tan ⎜ 3 − cos(Ω ) ⎟ otherwise
⎩ ⎝ ⎠

The DTFT spectrum for one period is shown below.

2π

−π
π
5
π

π
−π 5
π
 Solutions 23

Problem 11.5:
∞ ∞
X 1 (Ω) = ∑
k =−∞
x1[k ] ⋅ e − jΩk and X 2 (Ω) = ∑ x [k ] ⋅ e
k =−∞
2
− jΩk

∞ ∞ ∞
(i) X (Ω) = ∑ x[k ] ⋅ e
k =−∞
− jΩk
= ∑ (−1)
k =−∞
k
x1[k ]e − jΩk
= ∑ x [k ] ( − e )
k =−∞
1
− jΩ k

∞
= ∑ x [k ] ( e
k =−∞
1
− j (Ω −π ) k
) = X 1 (Ω − π )

(ii) x2 [k ] ←⎯⎯
DTFT
→ X 2 (Ω)
Applying the time shifting property, we obtain, h[k ] = x2 [k − 4] ←⎯⎯
DTFT
→ X 2 ( Ω) e − j 4 Ω = H ( Ω ) .
Using the DTFT pair kx[k ] ←⎯⎯
DTFT
→ j dXd(ΩΩ ) , we obtain the following two DTFT pairs:
⎧⎪ ⎫⎪ d { X 2 ( Ω ) e− j 4 Ω } dX 2 ( Ω )
ℑ ⎨kh[k ]⎬ = j dHd Ω( Ω ) = j dΩ = j ⎡⎣ e − j 4 Ω dΩ − j 4 X 2 (Ω )e − j 4 Ω ⎤⎦
⎪⎩ = g [ k ] ⎪⎭
= ⎡⎣ j 4 X 2 (Ω ) + dXd2Ω( Ω ) ⎤⎦ e − j 4 Ω = G (Ω )

ℑ{kg[k ]} = ℑ{k 2h[k ]} = j dGd(ΩΩ ) = j

{ dX ( Ω )
d ⎡ j 4 X 2 ( Ω ) + d2Ω ⎤ e− j 4 Ω
⎣ ⎦ }
dΩ

= j ⎡ j 4e − j 4 Ω dX 2 ( Ω )
+ 16 X 2 (Ω )e − j 4 Ω + e − j 4 Ω d 2 X 2 (Ω )
− j 4e − j 4 Ω dX 2 ( Ω ) ⎤
⎣ dΩ d 2Ω dΩ
⎦
= ⎡16 jX 2 (Ω ) + j d ⎤ e− j 4Ω = j ⎡ d + 16 X 2 (Ω ) ⎤ e − j 4 Ω
2 2
X 2 (Ω ) X 2 (Ω )
⎣ 2
d Ω ⎦ ⎣ d 2Ω ⎦

Using the above results, the DTFT of the given signal is now obtained as follows:

ℑ{(k − 5)2 x2 [k − 4]} = ℑ{(k − 5)2 h[k ]} = ℑ{(k 2 − 10k + 25)h[k ]}

= ℑ{k 2h[k ]} − 10ℑ{kh[k ]} + 25ℑ{h[k ]}

= j ⎡d + 16 X 2 (Ω ) ⎤ e − j 4 Ω − 10 ⎡⎣ j 4 X 2 (Ω ) + dXd2Ω( Ω ) ⎤⎦ e − j 4 Ω + 25 X 2 (Ω )e − j 4 Ω
2
X2 (Ω )
⎣ d 2Ω ⎦
=⎡jd + j16 X 2 (Ω ) − j 40 X 2 (Ω ) − 10 dXd2Ω( Ω ) + 25 X 2 (Ω ) ⎤ e − j 4 Ω
2
X2 (Ω )
⎣ d 2Ω ⎦
=⎡jd − 10 dXd2Ω( Ω ) + ( 25 − j 24 ) X 2 (Ω ) ⎤ e − j 4 Ω
2
X2 (Ω )
⎣ d 2Ω ⎦

(iii) x1[k ] ←⎯⎯

DTFT
→ X 1 (Ω)
Applying the time shifting property, we obtain, h[k ] = x1[k − 3] ←⎯⎯
DTFT
→ X 1 ( Ω ) e − j 3Ω = H ( Ω ) .
Applying the time inversion property, we obtain, g[k ] = x1[3 − k ] ←⎯⎯ DTFT
→ H ( −Ω) = X 1 ( −Ω)e j 3Ω = G (Ω) .
Applying the time multiplication property, we obtain,
dG (Ω) d { X 1 ( −Ω)e j 3Ω } ⎡ dX 1 ( −Ω) ⎤
p[k ] = kg[k ] = kx1[3 − k ] ←⎯⎯→ j
DTFT
= j =⎢j − 3 X 1 ( −Ω) ⎥ e j 3Ω = P(Ω) .
dΩ dΩ ⎣ dΩ ⎦
24 Chapter 11

Now applying the frequency shifting property, we obtain

⎡ dX ( −(Ω + 4)) ⎤
ℑ{ke − j 4 k x1[3 − k ]} = ℑ{e − j 4 k p[k ]} = P(Ω + 4) = ⎢ j 1 − 3 X 1 ( −(Ω + 4)) ⎥ e j 3( Ω + 4)
⎣ dΩ ⎦

⎧ ∞ ⎫ ⎧ ∞ ⎫ ⎧ ∞ ⎫
(iv) ℑ ⎨ ∑ ⎡⎣ x1 [k − 4m ] + x2 [k − 6m ]⎤⎦ ⎬ = ⎨ ∑ x1 [k − 4m ]⎬ + ℑ ⎨ ∑ x2 [k − 6m ]⎬
⎩ m=−∞ ⎭ ⎩ m=−∞ ⎭ ⎩ m=−∞ ⎭
∞ ∞
= ∑ ℑ{x1 [k − 4m]} + ∑ ℑ{x [k − 6m]} 2
m =−∞ m =−∞
∞ ∞
= ∑
m =−∞
X 1 (Ω )e − j 4 mΩ + ∑X
m =−∞
2 (Ω )e − j 6mΩ
∞ ∞
= X 1 (Ω ) ∑ e − j 4 mΩ + X 2 (Ω ) ∑ e − j 6mΩ
m =−∞ m =−∞

In order to calculate the infinite series sum, we will take help of the DTFT pair
∞ ∞ ∞
ℑ{1} = ∑e
k =−∞
− jΩk
= ∑e
m =−∞
− jΩm
= 2π ∑ δ ( Ω − 2mπ )
m =−∞

Substituting the Ω=4Ω′ in the above expression, we obtain the following identity:
∞ ∞
2π ∞
⎛ 2mπ ⎞ π ∞ ⎛ ′ mπ ⎞
∑
m =−∞
e − j 4Ω′m = 2π ∑ δ ( 4Ω′ − 2mπ ) =
m =−∞ 4
∑ δ ⎜ Ω′ −
m =−∞ ⎝ 4
⎟ = ∑ δ ⎜Ω −
⎠ 2 m=−∞ ⎝
⎟
2 ⎠
∞ ∞
π ∞
⎛ mπ ⎞
Similarly, we obtain, ∑
m =−∞
e − j 6Ω′m = 2π ∑ δ ( 6Ω′ − 2mπ ) =
m =−∞
∑ δ ⎜⎝ Ω′ −
3 m =−∞ 3
⎟.
⎠
Using the above results, we obtain:
⎧ ∞ ⎫ ∞ ∞
ℑ ⎨ ∑ ⎡⎣ x1 [k − 4m ] + x2 [k − 6m ]⎤⎦ ⎬ = X 1 (Ω ) ∑ e − j 4 mΩ + X 2 (Ω ) ∑ e − j 6 mΩ
⎩ m=−∞ ⎭ m =−∞ m =−∞
∞ ∞
= π2 X 1 (Ω ) ∑ δ ( Ω − m2π ) + π3 X 2 (Ω ) ∑ δ ( Ω − m3π )
m =−∞ m =−∞
∞ ∞
= π2 ∑ X (Ω −
m =−∞
1
mπ
2 )δ ( Ω − m2π ) + π3 ∑X
m =−∞
2 (Ω − m3π )δ ( Ω − m3π ) .

(v) Using the time-shifting and time-inversion properties, we

Applying the time shifting property, we obtain, h[k ] = x1[k − 5] ←⎯⎯
DTFT
→ X 2 ( Ω ) e − j 5Ω = H ( Ω ) .
Applying the time inversion property, we obtain,
g[k ] = x1[5 − k ] = h[ −k ] ←⎯⎯
DTFT
→ H ( −Ω) = X 1 ( −Ω)e j 5Ω = G (Ω) .

Similarly, we obtain the pair, p[k ] = x2 [7 − k ] ←⎯⎯

DTFT
→ X 2 ( −Ω)e j 7 Ω = P(Ω)
Applying the frequency convolution property, the DTFT of the given signal is obtained as:
 Solutions 25

ℑ{x1[5 − k ]x2 [7 − k ]} = ℑ{g[k ] p[k ]} = 1

2π ∫ G (θ ) P(Ω − θ )dθ
(2π )

= 1
2π
(2 )
∫π X 1 ( −θ )e j 5θ X 2 (θ − Ω )e j 7( Ω −θ ) dθ

= 1
2π e j 7Ω ∫π
(2 )
X 1 ( −θ ) X 2 (θ − Ω )e − j 2θ dθ

Problem 11.6

4 e − jΩ 4 e − jΩ A B
(i) X ( Ω ) = = = +
− jΩ
1 − 5e + 6e − j 2Ω
1 − 2e − jΩ
(1 − 3e − jΩ
)(
1 − 2e − jΩ
)
1 − 3e− jΩ

− jΩ
where A = 4e − jΩ 4/2 4 e − jΩ 4/3
= = −4 , B = = =4
1 − 3e 2 e− jΩ =1
1 − 1.5 1 − 2 e − jΩ 3 e− jΩ =1
1 − 23

Substituting the values of A and B, we get

4 4
X (Ω) = − − jΩ
+
1 − 2e 1 − 3e− jΩ
Using the DTFT pair −α k u[ − k − 1] ←⎯⎯
DTFT
→ 1 α > 1 , the sequence x[k ] is obtained as follows.
1−α e− jΩ

x[ k ] = 4 × 2k u[ − k − 1] − 4 × 3k u[ −k − 1] = 4(2 k − 3k )u[ − k − 1]

2e − j 2 Ω A B C
(ii) X ( Ω ) = = + +
(1 − 4e ) (1 − 2e ) (1 − 4e ) 1 − 2e
− jΩ − jΩ
− jΩ 2 − jΩ 1 − 4e − j Ω 2

− j 2Ω
where B = 2e − jΩ 1/ 8 1 2e − j 2 Ω 1/ 2 1
= = ,C= = = .
1 − 2e 1 − 1/ 2 4 (1 − 4e− jΩ ) (1 − 2)
2 2
e − jΩ =1/ 4
2
e − jΩ =1/ 2

Substituting B = 1 and C = 1 in X (Ω) expression and comparing the numerators, we get A = − 3 .

4 2 4
Substituting the values of A and B and C, we get
−3/ 4 1/ 4 1/ 2
X (Ω) = + +
(1 − 4e ) 1 − 2e
− jΩ − jΩ
1 − 4e − j Ω 2

Using the DTFT pairs

−α k u[ − k − 1] ←⎯⎯
DTFT
→ 1 α > 1 , and −( k + 1)α k u[ − k − 2] ←⎯⎯
DTFT
→ 1 α > 1,
1−α e− jΩ (1−α e )
− jΩ 2

the sequence x[k ] is obtained as follows:

26 Chapter 11

x[k ] = ( 43 4 k − 12 2 k ) u[ − k − 1] − 14 ( k + 1)4k u[ − k − 2]
= − 161 δ [k + 1] + { 43 4 k − 21 2 k − 14 ( k + 1)4k } u[ −k − 2]
= − 161 δ [k + 1] − 21 {2 k − (1 − k / 2)4 k } u[ − k − 2].

(iii) X ( Ω ) = 8sin ( 7Ω ) cos ( 9Ω ) = 4sin (16Ω ) − 4 sin ( 2Ω ) = − j 2 ⎡⎣ e j16Ω + e j 2 Ω − e − j 2 Ω − e − j16 Ω ⎤⎦

Using the DTFT pair δ [k − k0 ] ←⎯⎯

DTFT
→ e − jΩk , the sequence x[k ] is obtained as:
0

x[k ] = − j 2 {δ [k + 16] + δ [k + 2] − δ [k − 2] − δ [k − 16]}

4e − j 4 Ω 4e − j 4 Ω −4 e − j 5 Ω −4e − j 5Ω
(iv) X ( Ω ) = = = =
10 − 6cos Ω 10 − 3( e jΩ + e − jΩ ) 3 − 10e − jΩ + 3e − j 2Ω (3 − e − jΩ )(1 − 3e − jΩ )
⎡ −1/ 8 3/ 8 ⎤ 1 e − j 5Ω 3 e − j 5Ω
= −4e − j 5Ω ⎢ + = ⋅ − ⋅
⎣3− e
− jΩ
1 − 3e − jΩ ⎥⎦ 6 1 − (1/ 3)e − jΩ 2 1 − 3e − jΩ
Using the DTFT pairs p k u[k ] ←⎯⎯
DTFT
→ 1 p < 1 , −α k u[ − k − 1] ←⎯⎯
DTFT
→ 1 α > 1 and the
1− pe− jΩ 1−α e− jΩ
time shifting property of the Fourier transform, the sequence x[k ] is obtained as follows.

x[k ] = { 16 ( 13 ) k ′ u[k ′] + 23 (3) k ′ u[ − k ′ − 1]} ′= = 61 ( 13 ) k −5 u[k − 5] + 23 (3) k −5 u[ − k + 4]

k k −5

π π π
1 1 j
(v) x[k ] = ∫π X ( Ω ) e jk Ω d Ω = ∫π X ( Ω ) cos(Ωk ) d Ω + ∫π X ( Ω ) sin(Ωk ) d Ω
2π −
2π −
2π −
= even = even =odd
=0
π 3π / 4 3π / 4
2 1 1 ⎡ sin(Ωk ) ⎤
=
2π ∫ X ( Ω ) cos(Ωk )d Ω = π π∫
0 /4
cos(Ωk )d Ω =
π ⎢⎣ k ⎥⎦0.25π
1
= [sin(3π k / 4) − sin(π k / 4)]
πk
Alternative Solution:
X ( Ω ) = X1 ( Ω ) − X 2 ( Ω )

⎧1 Ω < 0.75π Ω < 0.25π

where X 1 ( Ω ) = ⎪⎨ and X ( Ω ) = ⎧⎪⎨1 .
0.75π ≤ Ω < π
2
⎪⎩0 ⎪⎩0 0.25π ≤ Ω < π

Therefore,
1
x[k ] = ℑ−1 { X 1 ( Ω ) − X 2 ( Ω )} = ℑ−1 { X 1 ( Ω )} − ℑ−1 { X 2 ( Ω )} = [sin(3π k / 4) − sin(π k / 4)] ▌
πk
 Solutions 27

Problem 11.7:

(a) Assuming x[k ] is a real-valued function, the Hermitian symmetry property can be proved as follows.
*
⎡ ∞ ⎤ ∞
X ( Ω ) = ⎡⎣ X ( Ω )⎤⎦ = ⎢ ∑ x[k ]e − jΩk ⎥ = ∑ ⎡⎣ x[k ]e − jΩk ⎤⎦
∗ * *

⎣ k =−∞ ⎦ k =−∞
∞
= ∑ x[k ] ⎡⎣e
k =−∞
− jΩk *
⎤⎦ [∵ x[k ] is real-valued function ]
∞
= ∑ x[k ]e
k =−∞
jΩk

= X ( −Ω )
(b)

4 e jΩ
(i) X ( −Ω ) = ..
1 − 5e jΩ + 6e j 2 Ω

4 e jΩ
X * (Ω) = .
1 − 5e jΩ + 6e j 2 Ω

Since X * ( Ω ) = X ( −Ω ) , the DT sequence is real-valued.

(ii) X ( −Ω ) = 2e j 2 Ω ; X * (Ω) = 2e j 2 Ω
(1 − 4e ) (1 − 2e )
jΩ 2 jΩ
(1 − 4e ) (1 − 2e )
jΩ 2 jΩ

Since X * ( Ω ) = X ( −Ω ) , the DT sequence is real-valued.

(iii) X ( −Ω ) = 8sin ( −7Ω ) cos ( −9Ω ) = − 8sin ( 7Ω ) cos ( 9Ω ) ; X * ( Ω ) = 8sin ( 7Ω ) cos ( 9Ω ) .

Since X * ( Ω ) ≠ X ( −Ω ) , the DT sequence is not real-valued.

4e j 4 Ω 4e j 4 Ω 4e j 4 Ω
(iv) X ( −Ω ) = = ; X * (Ω) = .
10 − 6 cos( −Ω) 10 − 6 cos(Ω) 10 − 6cos(Ω)

Since X * ( Ω ) = X ( −Ω ) , the DT sequence is real-valued.

(v) X ( −Ω ) = X ( Ω ) = X * ( Ω ) , and therefore the DT sequence is real valued. ▌

28 Chapter 11

Problem 11.8:

dX (Ω) d ⎡ ∞ − j Ωk ⎤
∞ ∞

dΩ
= ⎢ ∑
d Ω ⎣ k =−∞
x[ k ]e ⎥ = ∑
d
Ω
{ x[ k ]e − j Ωk
} = ∑ x[k ]
d
Ω
{e− jΩk }
⎦ k =−∞ d k =−∞ d
∞ ∞
= ∑
k =−∞
x[k ]( − jk )e − jΩk = ∑ (− jkx[k ])e
k =−∞
− j Ωk

= ℑ{− jkx[k ]}

DTFT dX
In other words, − jkx [k ] ←⎯ ⎯→ . ▌
dΩ

Problem 11.9
∞ ∞
⎛ ∞
⎞ ∞ ∞
ℑ{x[k ] ∗ h[k ]} = ∑ {x[k ] ∗ h[k ]}e− jΩk = ∑ ⎜ ∑
k =−∞ k =−∞ ⎝ m =−∞
x[m ]h[k − m] ⎟ e − jΩk = ∑ ∑ x[m ]h[k − m]e − jΩk
⎠ k =−∞ m =−∞
∞ ∞ ∞ ∞
= ∑∑
m =−∞ k =−∞
x[m ]h[k − m]e − jΩk = ∑
m =−∞
x[m ] ∑ h[k − m ]e − jΩk
k =−∞
=ℑ{h [ k − m ])
∞ ∞
= ∑ x[m]H (Ω)e
m =−∞
− j Ωm
= H (Ω) ∑ x[m]e
m =−∞
− j Ωm

= X (Ω)

= X (Ω) H (Ω)
▌

Problem 11.10
∞ ∞
ℑ{x[k − k0 ]} = ∑ x[k − k ]e
k =−∞
0
− j Ωk
= ∑ x[k − k ]e
k =−∞
0
− jΩ ( k − k0 ) − jΩk0
e
∞
= e − jΩ k 0 ∑ x[m]e
m =−∞
− j Ωm
[substituting k − k0 = m ]
= e − jΩ k 0 X ( Ω )
▌

Problem 11.11:

(i) H (Ω) = 1 k1 k2 k3
= + +
(1 − 0.3e − jΩ
)(1 − 0.5e )(1 − 0.7e ) 1 − 0.3e 1 − 0.5e 1 − 0.7e− jΩ
− jΩ − jΩ − jΩ − jΩ

where the partial fraction coefficients are calculated as

1 1 9
k1 = = =
(1 − 0.5e )(1 − 0.7e )
− jΩ − jΩ
0.3e− jΩ =1
(1 − 0.5
0.3 )(1 − 0.3 ) 8
0.7
 Solutions 29

1 1 25
k2 = = =−
(1 − 0.3e )(1 − 0.7e )
− jΩ − jΩ
0.5 e− jΩ =1
(1 − 0.5
0.3
)(1 − 0.7
0.5 ) 4

1 1 49
k3 = = =
(1 − 0.3e )(1 − 0.5e )
− jΩ − jΩ
0.7 e − jΩ =1
(1 − 0.3
0.7 )(1 − 0.7 ) 8
0.5

In other words, H (Ω) = 9 /8 −25 / 4 49 / 8

− jΩ
+ − jΩ
+
1 − 0.3e 1 − 0.5e 1 − 0.7e − jΩ

Calculating the inverse DTFT of each partial fractions, h[k ] is obtained as

h[k ] = ⎡⎣ 98 (0.3) k − 254 (0.5)k + 498 (0.7) k ⎤⎦ u[k ]

(ii) H (Ω) = 1 1
=
(1 − 0.3e − jΩ
)(1 − 0.5e )(1 − 0.7e ) (1 − 0.8e + 0.15e− j 2Ω )(1 − 0.7e− jΩ )
− jΩ − jΩ − jΩ

1 Y (Ω)
= − jΩ − j 2Ω − j 3Ω
=
1 − 1.5e + 0.71e − 0.105e X (Ω )

Or, (1 − 1.5e − jΩ + 0.71e − j 2 Ω − 0.105e − j 3Ω ) Y (Ω) = X (Ω)

Calculating the inverse DTFT of both sides, the input-output relationship is obtained as
y[k ] − 1.5 y[k + 1] + 0.71 y[k + 2] − 0.105 y[k + 3] = x[k ]

1
(iii) X (Ω) Ω ≤π = πδ (Ω) +
1 − e − jΩ
1
H (Ω) =
(1 − 0.3e − jΩ
)(1 − 0.5e− jΩ )(1 − 0.7e− jΩ )
⎡ 1 ⎤ 1
Y (Ω) Ω ≤π = X (Ω) H (Ω) = ⎢πδ (Ω) + ×
⎦ ( )( e − jΩ )(1 − 0.7e − jΩ )
− jΩ ⎥
⎣ 1 − e 1 − 0.3e − jΩ
1 − 0.5
πδ (Ω) 1
= +
(1 − 0.3e )(1 − 0.5e )(1 − 0.7e ) (1 − 0.3e )(1 − 0.5e )(1 − 0.7e )(1 − e )
− jΩ − jΩ − jΩ − jΩ − jΩ − jΩ − jΩ

= A( Ω ) =B (Ω)

Using the properties of continuous impulse function δ (⋅) , A(Ω) can be simplified as:
πδ (Ω)
A(Ω) = = 200
πδ (Ω)
(1 − 0.3 )(1 − 0.5 )(1 − 0.7 ) 21

The function B(Ω) can be decomposed into four partial fractions as follows.
1
B (Ω) =
(1 − 0.3e )(1 − 0.5e )(1 − 0.7e− jΩ )(1 − e− jΩ )
− jΩ − jΩ

k1 k2 k3 k4
= − jΩ
+ − jΩ
+ − jΩ
+
1 − 0.3e 1 − 0.5e 1 − 0.7e 1 − e − jΩ
where the partial fraction coefficients are calculated as
1 1 27
k1 = = =−
(1 − 0.5e − jΩ
)(1 − 0.7e− jΩ )(1 − e− jΩ ) 0.3e− jΩ =1
(1 − 0.5
0.3 )(1 − 0.3 )(1 − 0.3 )
0.7 1
56
30 Chapter 11

1 1 25
k2 = = =
(1 − 0.3e − jΩ
)(1 − 0.7e− jΩ )(1 − e− jΩ ) 0.5 e− jΩ =1
(1 − 0.5
0.3
)(1 − 0.7
0.5 )(1 − 1
0.5 ) 4

1 1 343
k3 = = =−
(1 − 0.3e − jΩ
)(1 − 0.5e− jΩ )(1 − e− jΩ ) 0.7 e− jΩ =1
(1 − 0.3
0.7 )(1 − 0.7 )(1 − 0.7 )
0.5 1
24

1 1 200
k4 = = =
(1 − 0.3e − jΩ
)(1 − 0.5e− jΩ )(1 − 0.7e− jΩ ) e− jΩ =1
(1 − 0.3)(1 − 0.5)(1 − 0.7) 21

−27 / 56 25 / 4 −343 / 24 200 / 21

Y (Ω) Ω ≤π = 200
πδ (Ω) + − jΩ
+ − jΩ
+ +
21
1 − 0.3e 1 − 0.5e 1 − 0.7e − jΩ 1 − e − jΩ
⎡ 1 ⎤ −27 / 56 25 / 4 −343 / 24
21 ⎢πδ (Ω) +
= 200 − jΩ ⎥
+ − jΩ
+ − jΩ
+
⎣ 1− e ⎦ 1 − 0.3e 1 − 0.5e 1 − 0.7e − jΩ

y[k ] = ⎡⎣ 200 k 25 k 343 k

⎤
21 − 56 (0.3) + 4 (0.5) − 24 (0.7) ⎦ u[ k ]
27

⎧( k + 1)u[k ] b =1
(iv) Note that {u[k ]} ∗{bk u[k ]} = ⎪⎨
⎩⎪ 1−b ⎡⎣1 − b ⎤⎦ u[k ] b ≠ 1
1 k +1

Therefore,
y[k ] = x[k ] ∗ h[k ]
{
= u[k ]* ⎡⎣ 98 (0.3) k − 254 (0.5)k + 498 (0.7) k ⎤⎦ u[k ] }
= 9
8 {u[k ]* (0.3) u[k ]} − {u[k ]* (0.5) u[k ]} + {u[k ]* (0.7) u[k ]}
k 25
4
k 49
8
k

⎧ 9 1 − (0.3) k +1 25 1 − (0.5) k +1 49 1 − (0.7) k +1 ⎫

=⎨ ⋅ − ⋅ + ⋅ ⎬ u[k ]
⎩ 8 1 − 0.3 4 1 − 0.5 8 1 − 0.7 ⎭
⎧ 45 25 245 ⎫
= ⎨ (1 − (0.3) k +1 ) − (1 − (0.5) k +1 ) + (1 − (0.7) k +1 ) ⎬ u[k ]
⎩ 28 2 12 ⎭
⎧⎛ 45 25 245 ⎞ 45 k +1 25 k +1 245 ⎫
= ⎨⎜ − + ⎟ − (0.3) + (0.5) − (0.7)k +1 ⎬ u[k ]
⎩⎝ 28 2 12 ⎠ 28 2 12 ⎭
⎧ 200 45 × 0.3 k 25 × 0.5 k 245 × 0.7 k ⎫
=⎨ − 0.3 + 0.5 − 0.7 ⎬ u[k ]
⎩ 21 28 2 12 ⎭
= ⎡⎣ 200 k 25 k 343 k
⎤
21 − 56 (0.3) + 4 (0.5) − 24 (0.7) ⎦ u[k ]
27

Note that the results obtained in parts (iii) and (iv) are identical. ▌
 Solutions 31

Problem 11.12:

(i) y[k ] + y[k − 1] + 14 y[ k − 2] = x[k ] − x[k − 2]

Calculating the DTFT of both sides and rearranging the terms (see section 11.6), the transfer function can
be expressed as
1 − e − j 2Ω 1 − e − j 2Ω 1 1
H (Ω) = = = − ⋅ e − j 2Ω .
(1 + 2 e ) (1 + 0.5e ) (1 + 0.5e )
− jΩ − j 2Ω
1 + e + 0.5e 1 − j Ω 2 − j Ω 2 − jΩ 2

(ii) Rearranging the terms, the transfer function H ( Ω ) can be expressed as

1 − e− j 2Ω 1 1
H (Ω) = = − ⋅ e − j 2Ω .
(1 + 1
2 e )
− jΩ 2
(1 + 0.5e ) (1 + 0.5e )
− jΩ 2 − jΩ 2

Using Entry 7 of Table 11.2, we get the following DTFT pair

1
( k + 1) ( −0.5)k u[k ] ↔
(1 + 0.5e ) − jΩ 2

Applying the DTFT time shifting property (see Eq. 11.42) on the above DTFT pair, we get
1
( k ′ + 1) ( −0.5)k ′ u[k ′] k ′=k −2 ↔ ⋅ e− j 2Ω
(1 + 0.5e ) − jΩ 2

1
Or, ( k − 1) ( −0.5)k −2 u[k − 2] ↔ ⋅ e− j 2Ω
(1 + 0.5e ) − jΩ 2

By calculating the inverse DTFT of H ( Ω ) , the impulse response h[k] is given by

h[k ] = ( k + 1) ( −0.5) k u[k ] − ( k − 1) ( −0.5) k −2 u[k − 2]

= −4δ [k ] − (3k − 5)( −0.5) k u[k ]

(iii) y[k ] = y[k ] ∗ h[k ] = {−4δ [k ] − (3k − 5)( −0.5)k u[k ]} ∗ {0.5k u[k ]}

= −4 × 0.5k u[k ] − {(3k − 5)( −0.5) k u[k ]} ∗ {0.5k u[k ]}

= −4 × 0.5k u[k ] − 3{k ( −0.5) k u[k ]} ∗ {0.5k u[k ]} + 5{( −0.5) k u[k ]} ∗ {0.5k u[k ]}

Using the following two known convolution results

⎧⎪ k ( k +1) k
a u[k ] a=b
( ka u[k ]) ∗ ( b u[k ]) = ⎨ a ⎡ka k +1 − (2k + 1)a k b + bk +1 ⎤ u[k ] a ≠ b.
k k

⎪⎩ ( a −b )2 ⎣ ⎦

⎪⎧ ( k + 1)a u[k ] a=b

k

( a u[k ]) ∗ ( b u[k ]) = ⎨ 1 a k +1 − bk +1 u[k ] a ≠ b,

k k

⎪⎩ a −b ( )
32 Chapter 11

the sequence y[k ] is obtained as follows

y[k ] = −4 × 0.5k u[k ] − 3{k ( −0.5) k u[k ]} ∗ {0.5k u[k ]} + 5{( −0.5)k u[k ]} ∗ {0.5k u[k ]}
= −4 × 0.5k u[k ] + 1.5 ⎡⎣ k ( −0.5) k +1 − 0.5(k + 1)( −0.5)k + 0.5k +1 ⎤⎦ u[k ] + 5 ⎡⎣0.5k +1 − ( −0.5)k +1 ⎤⎦ u[k ]

{
= −4 × 0.5k + ⎡⎣1.5k ( −0.5) k +1 − 0.75( k + 1)( −0.5) k + 1.5 × 0.5k +1 ⎤⎦ + ⎡⎣5 × 0.5k +1 − 5 × ( −0.5)k +1 ⎤⎦ u[k ] }
= {( −4 + 0.75 + 2.5)0.5k + ( −0.75k − 0.75k − 0.75 + 2.5)( −0.5)k } u[k ]
= {−0.75 × 0.5k + (1.75 − 1.5k + )( −0.5)k } u[k ]

− j 2Ω
(iv) X ( Ω ) = 1
; H (Ω) = 1 − e
(1 + 12 e− jΩ )
− Ω
1 − 12 e j 2

1 − e− j 2Ω A B C
Y (Ω) = X (Ω) H (Ω) = = + +
(1 − )(1 + ) (1 + 2 e− jΩ )
− jΩ 1 − jΩ
1
e − jΩ 1
e − jΩ 2 1− e1
2 1+ 2 e 1
2
2 2

1 − e− j 2Ω 3 1 − e− j 2Ω 3
A= =− , C= =− .
(1 + 1
2 e )
− jΩ 2
e − jΩ = 2
4 (1 − 12 e− jΩ ) e − jΩ =−2
2

Substituting A = − 3 and C = − 3 in X (Ω) expression and comparing the numerators, we get B = 13 .

4 2 4
Substituting the values of A and B and C, we get
−3/ 4 13/ 4 3/ 2
X (Ω) = + − .
1 − jΩ
1− 2 e 1 − jΩ
1+ 2 e (1 + 2 e )
2
1 − jΩ

Using the DTFT pairs in Table 11.2, the sequence y[k ] is obtained as follows.

y[k ] = {− 43 ( 12 ) k + 134 ( − 12 ) k − 23 ( k + 1)( − 12 )k } u[k ] = {− 43 ( 12 ) k + [ 74 − 1.5k )] ( − 12 ) k } u[k ] .

Note that the results obtained in parts (iii) and (iv) are identical. ▌
 Solutions 33

% Problem 11.12
%The MATLAB code to plot the magnitude frequency response in Problem 11.12.
omega = [-pi:0.01:pi] ;
H = (1-exp(-j*2*omega))./(1+ exp(-j*omega) + 0.5*exp(-j*2*omega));
abs_H = abs(H);

plot(omega, abs_H), grid

xlabel('Frequency (in rad/s)') % Label of X-axis
ylabel('Magnitude Spectrum') % Label of Y-axis %
axis([-4 4 0 5])
print -dtiff plot.tiff
34 Chapter 11

Problem 11.13:

−k
(i) x[k ] = u[ k ] and h[k ] = 4

To solve this problem, we will use the following DTFT pairs

1
p k u[k ] ↔ p <1
1 − pe − jΩ
1
p − k u[ − k ] ↔ p <1
1 − pe jΩ
1
− p k u[ −k − 1] ↔ p >1
1 − pe − jΩ
The impulse response can be expressed as

h[k ] = 4 − k = 4 − k u[k ] + 4 k u[− k − 1] = 0.25k u[k ] + 0.25− k u[ − k ] − δ [k ]

The transfer function H ( Ω ) can be expressed as:

1 1 1 −4e − jΩ
H (Ω) = + − 1 = + −1
1 − 0.25e − jΩ 1 − 0.25e jΩ 1 − 0.25e − jΩ 1 − 4e − jΩ
−3.75e − jΩ
=
(1 − 0.25e− jΩ )(1 − 4e− jΩ )
On the other hand,
1
X (Ω) Ω ≤π = πδ (Ω) +
1 − e − jΩ
Therefore, Y ( Ω ) can be expressed as

⎡ 1 ⎤ −3.75e − jΩ
Y (Ω) Ω ≤π = X (Ω) H (Ω) = ⎢πδ (Ω) + ×
⎣ 1 − e ⎦ (1 − 0.25e − jΩ )(1 − 4e − jΩ )
− jΩ ⎥

−3.75e − jΩ −3.75e − jΩ
= πδ (Ω) × +
(1 − 0.25e− jΩ )(1 − 4e− jΩ ) (1 − e− jΩ )(1 − 0.25e− jΩ )(1 − 4e− jΩ )
−3.75 −3.75e − jΩ
= πδ (Ω) × +
(1 − 0.25)(1 − 4) (1 − e − jΩ )(1 − 0.25e − jΩ )(1 − 4e − jΩ )
k1 k2 k3
= 53 πδ (Ω) + + +
1 − e − jΩ 1 − 0.25e − jΩ 1 − 4e − jΩ

The partial fraction coefficients can be calculated as k1 = 5 , k2 = − 1 , k3 = − 4 .

3 3 3
5⎡ 1 ⎤ 1/ 3 4/3
Y (Ω) Ω ≤π = ⎢πδ (Ω) + − −
3⎣ 1 − e − jΩ ⎥⎦ 1 − 0.25e − jΩ 1 − 4e − jΩ

Using the DTFT pairs mentioned above, y[k ] can be expressed as

 Solutions 35

y[k ] = 53 u[k ] − 13 × 0.25k u[k ] + 43 × 4 k u[ − k − 1] = ( 53 − 13 × 0.25k ) u[k ] + 43 × 0.25− k u[− k − 1]

⎧ 53 − 13 × 0.25k k ≥0
=⎨ 4 −k
⎩ 3 × 0.25 k<0

(ii) x[k ] = 2 − k u[k ] and h[k ] = 2k u[ − k − 1]

Using the DTFT pairs

1
p k u[k ] ↔ p <1
1 − pe − jΩ
1
− p k u[ −k − 1] ↔ p >1
1 − pe − jΩ
the DTFT X ( Ω ) and H ( Ω ) are obtained as:

1 1
X (Ω) = − jΩ
; H (Ω) = − .
1 − 0.5e 1 − 2 e − jΩ

Therefore, the DTFT Y ( Ω ) can be expressed as

1 k1 k2 .
Y ( Ω ) = X (Ω ) H ( Ω) = − = +
(1 − 0.5e − jΩ
)(1 − 2 e − jΩ
) 1 − 0.5 e − jΩ
1 − 2 e − jΩ

The partial fraction coefficients can be calculated as k1 = 1 , k2 = − 4 .

3 3
Using the DTFT pairs mentioned above, y[k ] can be calculated as
⎡ −1/ 3 −4 / 3 ⎤ 1 − k
y [ k ] = ℑ −1 ⎢ + = 3 × 2 u[k ] − 43 × 2 k u[ − k − 1] .
⎣1 − e
− jΩ
1 − 2e − jΩ ⎥⎦

⎧1 0 ≤ k ≤ 8
(iii) x[k ] = u[k ] − u[k − 9] = ⎨ and h[k ] = 3k u[ − k + 4]
⎩0 otherwise
∞ 8
1 − e− j 9Ω
X ( Ω ) = ∑ x[k ]e = ∑1 ⋅ e
− jΩ k − jΩk
=
k =−∞ k =0 1 − e − jΩ
1
Using the DTFT pair, − p k u[ −k − 1] ↔ p > 1 , we get
1 − pe − jΩ
1
−3k u[ −k − 1] ↔
1 − 3e − jΩ
e − j 5Ω
Or, −3( k −5) u[ −( k − 5) − 1] ↔
1 − 3e − jΩ
e − j 5Ω
Or, 3 u[ − k + 4] ↔ −3
k 5

1 − 3e − jΩ
36 Chapter 11

e − j 5Ω
In other words, H (Ω) = −35
1 − 3e − jΩ
Using the convolution property of DTFT, Y ( Ω ) can be expressed as

1 − e− j 9Ω e − j 5Ω 5 − j 5Ω 1 − e − j 9Ω
Y (Ω) = X (Ω) H (Ω) = −35 × = −3 e
1 − e − jΩ 1 − 3e − jΩ (1 − e− jΩ )(1 − 3e− jΩ )
1
= 35 ⎡⎣ e − j14 Ω − e − j 5Ω ⎤⎦
(1 − e − jΩ
)(1 − 3e− jΩ )
Noting that 1 ⎡
= 0.5e jΩ ⎢
1
−
1 ⎤, Y ( Ω ) can be expressed as
(1 − e − jΩ
)(1 − 3e − jΩ
) ⎣ 1 − 3e − jΩ
1 − e − jΩ ⎥⎦
⎡ 1 1 ⎤
Y (Ω) = 35 ⎡⎣ e − j14 Ω − e − j 5Ω ⎤⎦ × 0.5e jΩ ⎢ −
⎣1 − 3e
− jΩ
1 − e − jΩ ⎥⎦
⎡ 1 1 ⎤
= 0.5 × 35 ⎡⎣ e − j13Ω − e − j 4 Ω ⎤⎦ ⎢ −
⎣1 − 3e
− jΩ
1 − e − jΩ ⎥⎦
⎧ ⎫
− j 9Ω
⎪ − j13Ω 1 − j 4Ω 1 − e ⎪
= 0.5 × 3 ⎨ ⎡⎣ e
5
−e − j 4Ω
⎤⎦ − jΩ
+e − jΩ ⎬
⎪ 1 − 3e 1− e ⎪
⎩ =Q ( Ω ) = X (Ω) ⎭
1
Using the DTFT pair, q[k ] = −3k u[ −k − 1] ↔ = Q (Ω) , and the linearity and time shifting
1 − 3e − jΩ
property of the DTFT, we obtain
y[k ] = 0.5 × 35 {q[k − 13] − q[k − 4] + x[k − 4]}
= 0.5 × 35 {−3k × 3−13 u[ − k + 12] + 3k × 3−4 u[ −k + 3] + u[k − 4] − u[k − 13]}
= 0.5 × 35 {0.5 × 3k +1 u[ − k + 3] − 0.5 × 3k −8 u[ −k + 12] + u[k − 4] − u[k − 13]}

(iv) x[k ] = k 5 − k u[k ] and h[ k ] = 5 k u[− k ]

X ( Ω ) = ℑ{k 5− k u[k ]} = ℑ{k 0.2k u[k ]} = ℑ{( k + 1)0.2k u[k ]} − ℑ{0.2k u[k ]}
1 1 0.2e − jΩ
= − =
(1 − 0.2e ) (1 − 0.2e− jΩ )
− jΩ
− jΩ 2 1 − 0.2e 2

1 −5e − jΩ
H ( Ω ) = ℑ{5k u[ −k ]} = ℑ{δ [k ] + 5k u[−k − 1]} = 1 − − jΩ
=
1 − 5e 1 − 5e − jΩ

Therefore, the DTFT Y ( Ω ) can be expressed as

−e − j 2 Ω k1 k2 k3 .
Y ( Ω ) = X (Ω ) H ( Ω) = = + +
(1 − 5e )(1 − 0.2e )
− jΩ − jΩ 2 1 − 5e − jΩ (1 − 0.2e − jΩ )2 1 − 0.2e − jΩ

The first two partial fraction coefficients can be calculated as k1 = −25 , k2 = 25 .

576 24
Substituting k1 = −25
, k2 =
25 in Y (Ω) expression and comparing the numerators, we get k3 = −575 .
576 24 576
 Solutions 37

Substituting the values of A and B and C, we get

−25 / 576 25 / 24 −575 / 576
Y (Ω) = + + .
(1 − 0.2e )
− jΩ − jΩ
1 − 5e − j Ω 2
1 − 0.2 e
Using the DTFT pairs mentioned above, y[k ] can be calculated as

y[ k ] = 25
576 × 5k u[ − k − 1] + 24
25
( k + 1) × 0.2 k u[k ] − 576
575
× 0.2 k u[k ]
= 25
576 × 5k u[ − k − 1] + 24
25
( k + 241 ) × 0.2 k u[k ].

(v) In the textbook, there are typos in the x[k ] and h[k ] expressions. The correct expressions are as
follows:
⎧ 1 −2 ≤ k ≤ 2
x[k ] = u[k + 2] − u[k − 3] = ⎨ and h[k ] = u[k − 5] − u[k − 6] = δ [k − 5]
⎩0 otherwise
∞ 2
1 − ( e − jΩ ) 5 j 2Ω 1 − e
− j 5Ω
X (Ω) = ∑
k =−∞
x[k ]e − jΩk = ∑ 1 ⋅ e − jΩ k = e j 2 Ω
k =−2 1 − e − jΩ
= e
1 − e − jΩ
H ( Ω ) = ℑ{δ [k − 5]} = e − j 5Ω

Therefore, Y ( Ω ) can be expressed as

1 − e − j 5Ω 4 7
Y ( Ω ) = X ( Ω ) H ( Ω ) = e − j 3Ω − jΩ
= e − j 3Ω ∑ e − jkΩ = ∑1.e − jk Ω ,
1− e k =0 k =3

resulting in the following output:

⎧1 3 ≤ k ≤ 7
y[ k ] = ⎨ = u[k − 3] − u[k − 8] . ▌
⎩0 otherwise

Problem 11.14:
1 1 + 3e jΩ 1 + 3cos(Ω) + j 3sin(Ω)
H (Ω) = − jΩ
= − jΩ jΩ
=
1 + 3e (1 + 3e )(1 + 3e ) 10 + 6 cos(Ω)
1 + 3cos(Ω)
(i) Re { H (Ω)} =
10 + 6 cos(Ω)
3sin(Ω)
(ii) Im { H (Ω)} =
10 + 6 cos(Ω)
1 1 1 1
(iii) H (Ω) = − jΩ
= = = .
1 + 3e 1 + 3cos(Ω) − j 3sin(Ω) (1 + 3cos(Ω) )
2
+ 9sin 2 (Ω) 10 + 6 cos(Ω)

⎛ 3sin(Ω) ⎞
(iv) H (Ω) = tan −1 ⎜ ⎟
⎝ 1 + 3cos(Ω) ⎠
38 Chapter 11

The frequency responses Re { H (Ω)} , Im { H (Ω)} , H (Ω) , and H (Ω) , for −π ≤ Ω ≤ π , are shown in
Fig. S11.14. ▌

Fig. S11.14: Frequency response plots of H (Ω) in Problem 11.14.

% Problem 11.14
%The MATLAB code to plot the frequency responses in Problem 11.14.
omega = [-pi:0.01:pi] ;
ReH = (1+3*cos(omega))./(10+6*cos(omega));
ImH = 3*sin(omega)./(10+6*cos(omega));
MagH = 1./sqrt(10+6*cos(omega));
PhaseH = atan(3*sin(omega)./(1+3*cos(omega)));

subplot(2,2,1), plot(omega, ReH), grid

xlabel('Omega (in rad/s)') % Label of X-axis
%axis([-4 4 0 5])
ylabel('Re(H)') % Label of Y-axis
%
subplot(2,2,2), plot(omega, ImH), grid
xlabel('Omega (in rad/s)') % Label of X-axis
ylabel('Im(H)') % Label of Y-axis %
%
subplot(2,2,3), plot(omega, MagH), grid
xlabel('Omega (in rad/s)') % Label of X-axis
ylabel('Mag(H)') % Label of Y-axis %

subplot(2,2,4), plot(omega, PhaseH), grid

xlabel('Omega (in rad/s)') % Label of X-axis
ylabel('Phase(H)') % Label of Y-axis %

print -dtiff plot.tiff % Save the figure as a TIFF file

 Solutions 39

Problem 11.15:

(i) H1 ( Ω ) = −3.75e − jΩ −3.75e− jΩ −3.75e− jΩ

= =
( )(
1 − 0.25e − jΩ 1 − 4e− jΩ )
1 − 4.25e − jΩ + e − j 2Ω e − jΩ ⎡⎣e jΩ − 4.25 + e − jΩ ⎤⎦

−3.75 3.75 .
= =
2 cos(Ω) − 4.25 4.25 − 2 cos(Ω)
Note that [ 4.25 − 2 cos(Ω)] is a positive real-valued function. Therefore,

3.75 3.75
H1 ( Ω ) = = , and H1 ( Ω ) = 0 .
4.25 − 2 cos(Ω) 4.25 − 2 cos(Ω)

The magnitude and phase response plots are shown in Fig. S11.15(i).

(ii) H 2 ( Ω ) = − 1
1 − 2e− jΩ
1 1
H2 (Ω) = =
1 − 2 cos(Ω) + j 2sin(Ω) 5 − 4 cos(Ω)

⎛ 2sin(Ω) ⎞ .
H2 (Ω) = ( −1) − (1 − 2 cos(Ω) + j 2sin(Ω) ) = π − tan −1 ⎜ ⎟
⎝ 1 − 2 cos(Ω) ⎠
The magnitude and phase response plots are shown in Fig. S11.15(ii).

e− j 5Ω
(iii) H 3 (Ω) = −35
1 − 3e − jΩ
35 243
H3 ( Ω) = =
1 − 3cos(Ω) + j 3sin(Ω) 10 − 6 cos(Ω)

⎛ 3sin(Ω) ⎞ .
H3 ( Ω) = ( −3 ) + ( e
5 − j 5Ω
) − (1 − 3cos(Ω) + j3sin(Ω) ) = π − 5Ω − tan −1
⎜ ⎟
⎝ 1 − 3cos(Ω) ⎠
The magnitude and phase response plots are shown in Fig. S11.15(iii).
− jΩ
(iv) H 4 ( Ω ) = −5e − jΩ
1 − 5e
5 1
H4 (Ω) = =
1 − 5cos(Ω) + j5sin(Ω) 26 − 10 cos(Ω)

⎛ 5sin(Ω) ⎞ .
H4 (Ω) = ( −5) + ( e ) − (1 − 5cos(Ω) + j5sin(Ω) ) = π − Ω − tan
− jΩ −1
⎜ ⎟
⎝ 1 − 5cos(Ω) ⎠
The magnitude and phase response plots are shown in Fig. S11.15(iv).

(v) H 5 ( Ω ) = e − j 5Ω [Assuming h[k ] = u[k − 5] − u[k − 6] = δ [k − 5] ]

H5 ( Ω) = 1
40 Chapter 11

H5 (Ω) = (e − j 5Ω
) = −5Ω
The magnitude and phase response plots are shown in Fig. S11.15(v). ▌

(i) (ii)

(iii) (iv)

(v)
Fig. S11.15: Magnitude and Phase response plots for Problem 11.15.
 Solutions 41

% Problem 11.15
%The MATLAB code to plot the amplitude and phase responses in Problem 11.15.
omega = [-pi:0.01:pi] ;

%part(i)
NumH=3.75;
DenH=(4.25-2*cos(omega));
H1 = NumH./(DenH+eps);
MagH1 = H1;
PhaseH1 = 0*omega;
subplot(2,1,1), plot(omega, MagH1), grid
xlabel(' Frequency (in rad/s)') % Label of X-axis
ylabel('Mag(H1)') % Label of Y-axis %
subplot(2,1,2), plot(omega, PhaseH1), grid
xlabel(' Frequency (in rad/s)') % Label of X-axis
ylabel('Phase(H1)') % Label of Y-axis %
print -dtiff plot.tiff % Save the figure as a TIFF file
%
%part(ii)
NumH=-1;
DenH=1-2*exp(-j*omega);
H2 = NumH./(DenH+eps);
MagH2 = abs(H2);
PhaseH2 = phase(H2);
subplot(2,1,1), plot(omega, MagH2), grid
xlabel(' Frequency (in rad/s)') % Label of X-axis
ylabel('Mag(H2)') % Label of Y-axis %
axis([-4 4 0.2 1.2])
subplot(2,1,2), plot(omega, PhaseH2), grid
xlabel('Frequency (in rad/s)') % Label of X-axis
ylabel('Phase(H2)') % Label of Y-axis %
print -dtiff plot.tiff % Save the figure as a TIFF file

%part(iii)
omega = [-pi:0.01:pi] ;
NumH=(-3^5)*exp(-j*5*omega);
DenH=1-3*exp(-j*omega);
H3 = NumH./(DenH+eps);
MagH3 = abs(H3);
PhaseH3 = phase(H3);
%Unwrapping phase over [-pi,pi] %P=unwrap(PhaseH3,pi);
bin = round(PhaseH3/(2*pi));
PhaseH3 = PhaseH3-bin*2*pi; %mod(PhaseH3,pi)
subplot(2,1,1), plot(omega, MagH3), grid
xlabel(' Frequency (in rad/s)') % Label of X-axis
ylabel('Mag(H3)') % Label of Y-axis %
%axis([-4 4 0.2 1.2])
subplot(2,1,2), plot(omega, PhaseH3), grid
xlabel(' Frequency (in rad/s)') % Label of X-axis
ylabel('Phase(H3)') % Label of Y-axis %
print -dtiff plot.tiff % Save the figure as a TIFF file
%
%part(iv)
omega = [-pi:0.01:pi] ;
NumH=-5*exp(-j*omega);
DenH=1+NumH;
H4 = NumH./(DenH+eps);
MagH4 = abs(H4);
PhaseH4 = phase(H4);
%Unwrapping phase over [-pi,pi]
%bin = round(PhaseH4/(2*pi));
42 Chapter 11

%PhaseH4 = PhaseH4-bin*2*pi;
subplot(2,1,1), plot(omega, MagH4), grid
xlabel(' Frequency (in rad/s)') % Label of X-axis
ylabel('Mag(H4)') % Label of Y-axis %
axis([-4 4 0.8 1.4])
subplot(2,1,2), plot(omega, PhaseH4), grid
xlabel(' Frequency (in rad/s)') % Label of X-axis
ylabel('Phase(H4)') % Label of Y-axis %
axis([-4 4 -0.3 0.3])
print -dtiff plot.tiff % Save the figure as a TIFF file
%
%part(v)
omega = [-pi:0.01:pi] ;
H5=exp(-j*5*omega);
MagH5 = abs(H5);
PhaseH5 = phase(H5);
%Unwrapping phase over [-pi,pi]
bin = round(PhaseH5/(2*pi));
PhaseH5 = PhaseH5-bin*2*pi;
subplot(2,1,1), plot(omega, MagH5), grid
xlabel(' Frequency (in rad/s)') % Label of X-axis
ylabel('Mag(H5)') % Label of Y-axis %
axis([-4 4 0.8 1.2])
subplot(2,1,2), plot(omega, PhaseH5), grid
xlabel('Frequency (in rad/s)') % Label of X-axis
ylabel('Phase(H5)') % Label of Y-axis %
print -dtiff plot.tiff % Save the figure as a TIFF file

Problem 11.16:

h[k ] = 3δ[k + 3] − 2δ[k + 2] + δ[k + 1] + 5δ[k ] − δ[k − 1] − 2δ[k − 2] − 3δ[k − 3] + 4δ[k − 4]

∞
H (Ω) = ∑ h[k ] ⋅ e
k =−∞
− jΩk

∞
(i) H (Ω) Ω=0 = ∑ h[k ] = 3 − 2 + 1 + 5 − 1 − 2 − 3 + 4 = 5
k =−∞

∞
(ii) H (Ω) Ω=π = ∑ h[k ] ⋅ e
k =−∞
− jπ k
= ... + h[ −2] − h[−1] + h[0] − h[1] + h[2] − h[3] + ...
∞ ∞
= ∑
k =−∞
h[k ] − ∑ h[k ] = (−2 + 5 − 2 + 4) − (3 + 1 − 1 − 3) = 5
k =−∞
k = even k =odd

(iii) It is difficult to calculate the H (Ω) without calculating the DTFT. The DTFT can be calculated as
follows:
 Solutions 43

H (Ω) = 3e j 3Ω − 2e j 2Ω + e jΩ + 5 − e − jΩ − 2e − j 2Ω − 3e− j 3Ω + 4e− j 4Ω

= (3e j 3Ω − 3e − j 3Ω ) − (2e j 2Ω + 2e− j 2Ω ) + (e jΩ − e− jΩ ) + 5 + 4e− j 4Ω
= j 6sin(3Ω) − 4 cos(2Ω) + j 2sin(Ω) + 5 + 4 cos(4Ω) − j 4sin(4Ω)
= [5 − 4 cos(2Ω) + 4 cos(4Ω)] + j [ 2sin(Ω) + 6sin(3Ω) − 4sin(4Ω)]
=> 0

⎛ 2sin(Ω) + 6sin(3Ω) − 4sin(4Ω) ⎞

Therefore, H (Ω) = tan −1 ⎜ ⎟.
⎝ 5 − 4 cos(2Ω) + 4 cos(4Ω) ⎠
The phase H (Ω) is sketched in Fig. S11.16(a).

π
∫ π H ( Ω )e d Ω
jk Ω
(iv) From Eq. 11.28b), we know h[ k ] = 1
2π −
π π
∫ π H ( Ω )e dΩ = π ∫ H (Ω) dΩ
j ×0×Ω
In other words, h[0] = 1
2π 2
1
− π −
π
Therefore, ∫ π H ( Ω ) d Ω = 2π h[0] = 10π .
−
=π
∞
(v) H (Ω) = ∑ h[k ] ⋅ e
k =−∞
− jΩk
.

Substituting Ω= − Ω on both sides of the DTFT equation, we obtain:

∞ ∞
H (−Ω) = ∑
k =−∞
h[k ] ⋅ e jΩk = ∑ h[−(−k )] ⋅ e
k =−∞
− jΩ( − k )

∞
= ∑ h[−m] ⋅ e − jΩm
[substituting m = −k ]
m =−∞
∞
= ∑ h[−k ] ⋅ e
k =−∞
− jΩk

= ℑ{h[−k ]}
Therefore, the DT sequence is given by h[−k ] , which is sketched in Fig. S11.16(b).

⎧ ∞
− jΩk ⎫
∞
(vi) Re { H (Ω)} = Re ⎨ ∑
⎩k =−∞
h[ k ] ⋅ e ⎬ ∑ h[k ]cos(Ωk )
=
⎭ k =−∞
∞
1 1 ∞ 1 ∞
= ∑ h[k ] ⋅ ⎡⎣e jΩk + e− jΩk ⎤⎦ = ∑ h[k ]e jΩk + ∑ h[k ]e − jΩk
2 k =−∞ 2 k =−∞ 2 k =−∞
1 1 1
= ℑ{h[−k ]} + ℑ{h[k ]} = ℑ{h[k ] + h[−k ]}
2 2 2
Therefore, the DT sequence is given by 1
2 ( h[k ] + h[−k ]) , which is sketched in Fig. S11.16(c). ▌
44 Chapter 11

(a) (b)

(c)
Fig. S11.16: The DT sequences obtained in Problem 11.16.

% P11.16, part (iii) Phase plotting

omega = [-pi:0.01:pi] ;
R=5-4*cos(2*omega)+4*cos(4*omega);
I=2*sin(omega) + 6*sin(3*omega)-4*sin(4*omega);
H = R + j*I;
PhaseH = phase(H);
%Unwrapping phase over [-pi,pi] %P=unwrap(PhaseH,pi);
bin = round(PhaseH/(2*pi));
PhaseH = PhaseH-bin*2*pi; %mod(PhaseH3,pi)
plot(omega, PhaseH), grid
xlabel(' Frequency (\Omega)') % Label of X-axis
ylabel(' <H(\Omega)') % Label of Y-axis %
%axis([-4 4 0.2 1.2])
print -dtiff plot.tiff

Problem 11.17:
There is a typo in the question. This question can be easily solved using the Convolution (and not
Parseval’s theorem) property of the DTFT as follows.
∞ ∞
sin( π5 k )sin( π7 k )
Note that the sum is of the form:
k =−∞
∑ x[k ] = ∑
k =−∞ k2
. The sum of a discrete sequence can

easily calculated if the DTFT is known. Substituing Ω=0 on both sides of the DTFT expression
∞
X (Ω) = ∑ x[k ] ⋅ e
k =−∞
− jΩk
, we obtain
 Solutions 45

∑ x[k ] = X (Ω)
k =−∞
Ω=0
.

Noting that x[k ] is a product of two sinc functions, we expect that the X (Ω) will be the convolution of
two rectangular fiunctions, and in this case, it will be a trapezoidal shape. It can be shown that (see the
solution of P11.4(a)(vi))

⎧ 21π ( Ω+ 1235π ) − 1235π ≤ Ω< − 235π

⎪
⎡ sin( π5 k )sin( π7 k ) ⎤ 2⎪
1
7 − 235π ≤ Ω< 235π
ℑ⎢ ⎥⎦ = X (Ω) Ω ≤π =π ⎨
⎣ k2 ⎪ 2π ( Ω − 35 ) 35 ≤ Ω< 35
1 12π 2π 12π

⎪
⎩ 0 otherwise.
∞
sin ( π5k ) sin ( π7k ) π2
Therefore, ∑
k =−∞ k2
= X (Ω) Ω=0
=
7
. ▌

Problem 11.18:

h[k ] = sinc ( 34k ) = sinc ( (3ππ/ 4) k ) .

Using the DTFT pair (16) in Table 11.2, the transfer function H (Ω) is obtained as follows.

π ⎧⎪1 Ω ≤ 3π / 4 ⎧⎪4 / 3 Ω ≤ 3π / 4
H (Ω) Ω ≤π = ×⎨ =⎨
3π / 4 ⎪⎩0 3π / 4 < Ω ≤ π ⎪⎩ 0 3π / 4 < Ω ≤ π
Note that the DT system is a lowpass filter, which passes the DT frequency components below 3π / 4
rad/s, and stops the frequency above it.

⎛ 11π k ⎞ ⎛ 3π k ⎞ 1 ⎡ ⎛ 7π k ⎞ ⎛ π k ⎞⎤
(i) x[k ] = cos ⎜ ⎟ cos ⎜ ⎟ = ⎢cos ⎜ ⎟ + cos ⎜ ⎟⎥
⎝ 16 ⎠ ⎝ 16 ⎠ 2 ⎣ ⎝ 8 ⎠ ⎝ 2 ⎠⎦
π
X (Ω) Ω ≤π = ⎡⎣δ ( Ω + π2 ) + δ ( Ω − π2 ) + δ ( Ω + 78π ) + δ ( Ω − 78π ) ⎤⎦
2

Y (Ω) Ω ≤π = X (Ω) Ω ≤π ×H (Ω) Ω ≤π = ( 2π / 3) ⎡⎣δ ( Ω + π2 ) + δ ( Ω − π2 ) ⎤⎦

Using the DTFT pair (10) in Table 11.2, the output sequence is obtained as
2 ⎛πk ⎞
y[k ] = cos ⎜ ⎟
3 ⎝ 2 ⎠

(ii) The DTFT X (Ω) is given by (see solution of P11.4(b)(i))

46 Chapter 11

∞
X ( Ω ) = 2π ∑ D δ (Ω − π )
n =−∞
n
2 n
6

⎡.... + D−3δ ( Ω + π ) + D−2δ ( Ω + 23π ) + D−1δ ( Ω + π3 ) + D0δ ( Ω ) + ⎤

= 2π ⎢ ⎥
⎣⎢ D1δ ( Ω − 3 ) + D2δ ( Ω − 3 ) + D3δ ( Ω − π ) + ...
π 2π
⎦⎥
1 e − jnπ / 3 − 6e − j 6 nπ / 3 + 5e − j 7 nπ / 3
where Dn = . Substituting different values of n, we obtain,
6 (1 − e − jnπ / 3 ) 2

D0 = 25 ; D±1 = − 12 ± j 2
3
; D±2 ≈ − 12 ± j 0.2887 .
The DTFT of the output is given by
⎡ D−2δ ( Ω + 23π ) + D−1δ ( Ω + π3 ) + D0δ ( Ω ) + ⎤
Y (Ω) Ω ≤π = X (Ω) Ω ≤π ×H (Ω) Ω ≤π = ( 8π / 3) ⎢ ⎥
⎢⎣ D1δ ( Ω − 3 ) + D2δ ( Ω − 3 )
π 2π
⎥⎦

Using the DTFT pair (8) in Table 11.2, the output sequence is obtained as
y[k ] = ( 4π / 3) ⎡ D−2e ⎤
j 2π k jπ k jπ k j 2π k
− −
3
+ D−1e 3
+ D0 + D1e 3
+ D2e 3

⎣⎢ ⎦⎥
⎡( − 1 − j 0.2287)e − 3 + ( − 1 − j 3 )e − 3 + 5 + ⎤
j 2π k jπ k

= ( 4π / 3) ⎢ ⎥
2 2 2 2

⎢( − 1 + j 3 )e 3 + ( − 1 + j 0.2287)e 3 ⎥
jπ k j 2π k

⎣ 2 2 2 ⎦
= ( 4π / 3) ⎡⎣ − cos( 2π3 k ) − 0.4574sin( 2π3 k ) − cos( π3k ) − 3 sin( π3k ) + 25 ⎤⎦

= ( 4π / 3) ⎡⎣ 25 − cos( π3k ) − 3 sin( π3k ) − cos( 2π3 k ) − 0.4574sin( 2π3 k ) ⎤⎦ .

(iii) The DTFT X (Ω) is given by (see solution of P11.4(b)(ii))

∞
X ( Ω ) = 2π ∑ D δ (Ω − π )
n =−∞
n
2 n
9

⎡.... + D−4δ ( Ω + 89π ) + D−3δ ( Ω + 69π ) + D−2δ ( Ω + 49π ) + D−1δ ( Ω + 29π ) + ⎤

= 2π ⎢ ⎥
⎣⎢ D0δ ( Ω ) + D1δ ( Ω − 9 ) + D2δ ( Ω − 9 ) + D3δ ( Ω − 9 ) + D4δ ( Ω − 9 ) + ...⎦⎥
2π 4π 6π 8π

1
where Dn =
9
(1 + 0.5e − j 6 nπ / 9 )(1 + e − j 2 nπ / 9 + e − j 4 nπ / 9 ) . Substituting different values of n, we obtain,

D0 = 0.5; D±1 ≈ 0.0833 ∓ j 0.2290; D±2 ≈ 0.0833 ∓ j 0.0993, D±3 = 0 .

The DTFT of the output is given by
Y (Ω) Ω ≤π = X (Ω) Ω ≤π ×H (Ω) Ω ≤π

⎡ D−3δ ( Ω + 69π ) + D−2δ ( Ω + 49π ) + D−1δ ( Ω + 29π ) + D0δ ( Ω ) + ⎤

= ( 8π / 3) ⎢ ⎥
⎢⎣ D1δ ( Ω − 9 ) + D2δ ( Ω − 9 ) + D3δ ( Ω − 9 )
2π 4π 6π
⎥⎦
= ( 8π / 3) ⎡⎣ D−2δ ( Ω + 49π ) + D−1δ ( Ω + 29π ) + D0δ ( Ω ) + D1δ ( Ω − 29π ) + D2δ ( Ω − 49π ) ⎤⎦
 Solutions 47

Using the DTFT pair (8) in Table 11.2, the output sequence is obtained as
y[k ] = ( 4π / 3) ⎡⎢ D−2e ⎤
j 4π k j 2π k j 2π k j 4π k
− −
9
+ D−1e 9
+ D0 + D1e 9
+ D2e 9

⎣ ⎦⎥
⎡ (0.0833 + j 0.0993)e − 9 + (0.0833 + j 0.2290)e − 9 + 0.5 + ⎤
j 4π k j 2π k

≈ ( 4π / 3) ⎢ j 2π k j 4π k
⎥
⎢ (0.0833 − j 0.2290)e 9 + (0.0833 − j 0.0993)e 9 ⎥
⎣ ⎦
= ( 4π / 3) [0.1666cos( 9 ) + 0.1986sin( 9 ) + 0.1666cos( 9 ) + 0.1986sin( 2π9 k ) + 0.5]
4π k 4π k 2π k

= ( 2π / 3) [1 + 0.3332 cos( 4π9 k ) + 0.3972sin( 4π9 k ) + 0.3332 cos( 2π9 k ) + 0.3972sin( 2π9 k )].

(iv) The DTFT X (Ω) is given by (see solution of P11.4(b)(v))

2π ∞
X (Ω) = ∑ δ ( Ω − 2π5n )
5 n =−∞
2π
= ⎡.... + δ ( Ω + 45π ) + δ ( Ω + 25π ) + δ ( Ω ) + δ ( Ω − 25π ) + δ ( Ω − 45π ) + ...⎤⎦
5 ⎣
The DTFT of the output is given by
Y (Ω) Ω ≤π = X (Ω) Ω ≤π ×H (Ω) Ω ≤π = (8π /15) ⎡⎣δ ( Ω + 25π ) + δ ( Ω ) + δ ( Ω − 25π ) ⎤⎦

= ( 154 ) × 2πδ ( Ω ) + ( 158 ) × π ⎡⎣δ ( Ω + 25π ) + δ ( Ω − 25π ) ⎤⎦

Using the DTFT pairs (1) and (10) in Table 11.2, the output sequence is obtained as
y[k ] = 154 + 158 cos( 25π k ) .

(v) x[k ] = sinc ( k3 ) = sinc ( ( π / 3) k

π ) . The DTFT X (Ω) is given by.
π ⎧⎪1 Ω ≤π /3 ⎧⎪ 3 Ω ≤π /3
X (Ω) Ω ≤π = ×⎨ =⎨
π / 3 ⎩⎪0 π / 3 < Ω ≤ π ⎩⎪0 π / 3 < Ω ≤ π
The DTFT of the output is given by
⎪⎧4 Ω ≤π /3
Y (Ω) Ω ≤π = X (Ω) Ω ≤π ×H (Ω) Ω ≤π = ⎨ = ( 43 ) X (Ω) Ω ≤π
⎪⎩0 π / 3 < Ω ≤ π
Calculating the inverse DTFT, the output sequence is obtained as
y[k ] = 43 x[k ] = 43 sinc ( k3 ) . ▌

Problem 11.19:
1 1 2 (1 − 247 e − jΩ )
X ( Ω ) = ℑ{4 u[k ] + 3 u[k ]} =
−k −k
+ =
1 − 14 e − jΩ 1 − 13 e − jΩ (1 − 14 e − jΩ )(1 − 13 e − jΩ )
−2 (1 + 14 e − jΩ )
{
Y ( Ω ) = ℑ 2 ( 14 ) u[k ] − 4 ( 43 ) u[k ] =
k k
} 2
−
4
=
1 − 14 e − jΩ 1 − 43 e − jΩ (1 − 14 e − jΩ )(1 − 43 e − jΩ )
.

(i) The transfer function of the system is calculated as follows.

48 Chapter 11

Y ( Ω ) − (1 + 41 e )(1 − 13 e ) −1 + 121 e − jΩ + 121 e − j 2 Ω

− jΩ − jΩ

H (Ω) = = =
X ( Ω ) (1 − 43 e − jΩ )(1 − 247 e − jΩ ) 25 − jΩ
1 − 24 e + 327 e − j 2 Ω

(ii) Using the partial fraction approach, the impulse response can be obtained.

H (Ω) =
−1 + 121 e − jΩ + 121 e − j 2 Ω
=
8
21 (1 − 25
24 e − jΩ + 327 e − j 2 Ω ) − (1 + 218 ) + ( 121 + 63
25
)e − jΩ
25 − jΩ
1 − 24 e + 327 e − j 2 Ω 25 − jΩ
1 − 24 e + 327 e − j 2 Ω
121 − jΩ
− 29
21 + 252 e k1 k2
= 8
+ = 8
+ +
21
(1 − 43 e− jΩ )(1 − 247 e− jΩ ) 21
1− e3
4
− jΩ
1 − 247 e − jΩ

− 29
21 + 252 × 3
121 4
40 − 29
21 + 252 × 7
121 24
13
where k1 = = − and k 2 = = −
1 − 247 × 43 33 1 − 43 × 247 77
Substituting the values of k1 and k 2 , we get

40 / 33 13/ 77
H (Ω) = 8
− − Ω
− .
1 − 247 e − jΩ
21
1 − 43 e j

Calculating the inverse DTFT of H ( Ω ) , the impulse response h[k ] is obtained as follows.

h[k ] = 8
21 δ [k ] − ⎡⎣ 40 k 13 k
⎤
33 × (3/ 4) + 77 × (7 / 24) ⎦ u[ k ] .

Y ( Ω ) −1 + 121 e − jΩ + 121 e − j 2 Ω
(iii) H ( Ω ) = =
X ( Ω ) 1 − 24
25 − jΩ
e + 327 e − j 2 Ω
By cross-multiplication, we obtain:

(1 − 25
24 e − jΩ + 327 e − j 2 Ω ) Y ( Ω ) = ( −1 + 121 e − jΩ + 121 e − j 2 Ω ) X ( Ω )

By calculating the inverse DTFT of both sides, we obtain the following difference equation:
y[k ] − 24
25
y[k − 1] + 327 y[k − 2] = − x[k ] + 121 x[k − 1] + 121 x[k − 2] .
(iv) In part (ii), the impulse response was obtained as

h[k ] = 8
21 δ [k ] − ⎡⎣ 40 k 13 k
⎤
33 × (3/ 4) + 77 × (7 / 24) ⎦ u[ k ]

Because, h[k ] = 0 for all k < 0 , the system is causal. ▌

Problem 11.20:

The three LTID systems have the transfer functions as follows.

2
H 1 (Ω ) = − jΩ
.
1− e 3
4 + 18 e − j 2Ω
 Solutions 49

⎧1 Ω ≤ π6
H 2 (Ω ) = ⎨
⎩0 6 < Ω ≤ π.
π

⎧0 Ω ≤ π6
H 3 (Ω ) = ⎨
⎩1 6 < Ω ≤ π.
π

The problem includes several CT sinusoidal functions which are discretized using a sampling frequency
of 8000 Hz. Let us assume that a sinusoidal signal with frequency f 0 Hz is passed through the three
systems. After sampling the CT sinusoid, the corresponding DT frequency of the DT signal is
2π f 0
given by Ω0 = . The gains of the three systems will be as follows.
8000
2
H1 ( Ω ) Ω=Ω = − jΩ 0
.
0 1− e 3
4 + 18 e − j 2 Ω0
⎧1 Ω0 ≤ π6
H 2 ( Ω ) Ω=Ω = ⎨
0
⎩0 otherwise
⎧0 Ω0 ≤ π6
H 3 ( Ω ) Ω=Ω = ⎨
0
⎩1 otherwise

(i) x1 (t ) = 2 + 3 cos(400πt ) + 7 cos(800πt )

The DT sequence is given by, x1[k ] = x1 (kTs ) = x1 ( 8000

k
) = 2 + 3cos ( π20k ) + 5cos ( π10k ) .
Ω0 = 0
π
Ω1 = 20 π
Ω2 = 10

(a) The gains of the system at Ω = 0, 20

π π
, 10 are given by:
2 16
H1 ( Ω ) Ω=0 = = ;
1− +3
4
1
8 3
2 2
H1 ( Ω ) Ω= π = π π ≈ ≈ 5.0698 − j1.0552 = 5.1784 − 0.2052
20
1− e3
4
− j 20
+ e1
8
− j 10
0.3781 + j 0.0787
2 2
H1 ( Ω ) Ω= π = π − j π5
≈ ≈ 4.4205 − j1.8042 = 4.7745 − 0.3875
10
1 − 43 e
− j 10
+ 18 e 0.3878 + j 0.1583

The output signal can then be expressed as:

y1, H 1[k ] = 2 × 163 + 3 × 5.1784 cos ( π20k − 0.2052 ) + 5 × 4.7745cos ( π10k − 0.3875 )
= 32
3 + 15.5352 cos ( π20k − 0.2052 ) + 23.8725cos ( π10k − 0.3875 ) .

(b) The gains of the system at Ω = 0, 20

π π
, 10 are given by:
H 2 ( Ω ) Ω=0 = 1 ; H 2 ( Ω ) Ω= π = 1 ; H 2 ( Ω ) Ω= π = 1
20 10

As all three components will pass with a gain of 1, the output signal can then be expressed as:
y1, H 2 [k ] = 2 + 3cos ( π20k ) + 5cos ( π10k ) .
50 Chapter 11

(c) The gains of the system at Ω = 0, 20

π π
, 10 are given by:
H 3 ( Ω ) Ω=0 = 0 ; H 3 ( Ω ) Ω= π = 0 ; H 3 ( Ω ) Ω= π = 0 .
20 10

As all three components will pass with a gain of 0, the output signal can then be expressed as:
y1, H 3 [k ] = 0 .

(ii) x 2 (t ) = 2 cos(4000πt ) + 5 cos(6000πt )

The DT sequence is given by, x2 [k ] = x2 (kTs ) = x2 ( 8000

k
) = 2 cos ( π2k ) + 5cos ( 3π4 k ) .
Ω0 = π2 Ω1 = 34π

(a) The gains of the system at Ω = π2 , 34π are given by:

2 2
H1 ( Ω ) Ω= π = π ≈ ≈ 1.3176 − j1.1294 = 1.7354 − 0.7086
2 3 −j2
1− 4 e + 8 e 1 − jπ 0.875 + j 0.75
2 2
H1 ( Ω ) Ω= 3π = − j 34π − j 32π
≈ ≈ 1.1044 − j 0.4729 = 1.2014 − 0.4046
4 1− 4 e
3
+8e
1 1.5303 + j 0.6553

The output signal can then be expressed as:

y2, H 1[k ] = 2 ×1.7354 cos ( π2k − 0.7086 ) + 5 ×1.2014 cos ( 3π4 k − 0.4046 )
= 3.4708cos ( π2k − 0.7086 ) + 6.007 cos ( 3π4 k − 0.4046 ) .

(b) The gains of the system at Ω = π2 , 34π are given by:

H 2 ( Ω ) Ω= π = 0 ; H 2 ( Ω ) Ω= 3π = 0 .
2 4

As all three components will pass with a gain of 0, the output signal can then be expressed as:
y2, H 2 [k ] = 0 .

(c) The gains of the system at Ω = π2 , 34π are given by:

H 3 ( Ω ) Ω= π = 1 ; H 3 ( Ω ) Ω= 3π = 1 .
2 4

As all three components will pass with a gain of 1, the output signal can then be expressed as:
y2, H 3[k ] = 2 cos ( π2k ) + 5cos ( 3π4 k ) .

(iii) x3 (t ) = 5 cos(600πt ) + 9 cos(900πt ) + 2 cos(3000πt )

The DT sequence is given by, x3 [k ] = x3 (kTs ) = x3 ( 8000

k
) = 5cos ( 340
πk
) + 9 cos ( 980π k ) + 2 cos ( 3π8 k ) .
Ω0 = 340π Ω1 = 980π Ω1 = 38π

(a) The gains of the system at Ω = 340π , 980π , 38π are given by:
2 2
H1 ( Ω ) Ω= 3π = 3π 3π ≈ ≈ 4.7762 − j1.4792 = 5 − 0.3003
1 − 34 e 40 + 18 e 20 0.3821 + j 0.1183
− j − j
40
 Solutions 51

2 2
H1 ( Ω ) Ω= 9π = − j 980π − j 980π
≈ ≈ 4.2308 − j1.9284 = 4.6495 − 0.4277
80
1− e 3
4 + e
1
8
0.3914 + j 0.1784
2 2
H1 ( Ω ) Ω= 3π = − j 38π − j 34π
≈ ≈ 1.6533 − j1.6002 = 2.3009 − 0.7691
8
1 − 34 e + 18 e 0.6246 + j 0.6045

The output signal can then be expressed as:

y3, H 1[k ] = 5 × 5cos ( 340
πk
− 0.3003) + 9 × 4.6495cos ( 980π k − 0.4277 ) + 2 × 2.3009 cos ( 3π8 k − 0.7691)
= 25cos ( 340
πk
− 0.3003) + 41.8455cos ( 980π k − 0.4277 ) + 4.6018cos ( 3π8 k − 0.7691)

(b) The gains of the system at Ω = 3π

40 , 980π , 38π are given by:
H 2 ( Ω ) Ω= 3π = 1 ; H 2 ( Ω ) Ω= 9π = 1 ; H 2 ( Ω ) Ω= 3π = 0 .
40 80 8

As the first two components will pass with a gain of 1, and the third component will pass with a gain of 0,
the output signal can then be expressed as:

y3, H 2 [k ] = 5cos ( 340

πk
) + 9 cos ( 980π k ) .
(c) The gains of the system at Ω = 3π
40 , 980π , 38π are given by:
H 3 ( Ω ) Ω= 3π = 0 ; H 3 ( Ω ) Ω= 9π = 0 ; H 3 ( Ω ) Ω= 3π = 1 .
40 80 8

As the first two components will pass with a gain of 0 and the third component will pass with a gain of 1,
the output signal can then be expressed as:

y3, H 3 [ k ] = 2 cos ( 3π8 k ) .

(iv) x 4 (t ) = 4 cos(600πt ) + 6 cos(12000πt )

The DT sequence is given by,

x4 [k ] = x4 (kTs ) = x4 ( 8000
k
) = 4 cos ( 340
πk
) + 6 cos ( 3π2k ) = 4 cos ( 340π k ) + 6 cos ( π2k ) .
Ω0 = 340π Ω1 = π2

(a) The gains of the system at Ω = 340π , π2 are given by:

2 2
H1 ( Ω ) Ω= 3π = − j 340π − j 320π
≈ ≈ 4.7762 − j1.4792 = 5 − 0.3003
40
1− 4 e
3
+8e
1 0.3821 + j 0.1183
2 2 2
H1 ( Ω ) Ω= π = π = ≈7 ≈ 1.3176 − j1.1294 = 1.7354 − 0.7086
2
−
1 − 43 e 2 + 81 e
j − jπ
1 + 4 j − 8 8 + j 0.75
3 1

The output signal can then be expressed as:

y4, H 1[k ] = 5 × 4 cos ( 340
πk
− 0.3003) + 6 × 1.7354 cos ( π2k − 0.7086 )
= 20 cos ( 340
πk
− 0.3003) + 10.4124 cos ( π2k − 0.7086 ) .

(b) The gains of the system at Ω = 3π

40 , π2 are given by:
52 Chapter 11

H 2 ( Ω ) Ω= 3π = 1 ; H 2 ( Ω ) Ω= π = 0 .
40 2

As the first component will pass with a gain of 1, and the second component will pass with a gain of 0,
the output signal can be expressed as:
y4, H 2 [k ] = 4 cos ( 340
πk
).
(c) The gains of the system at Ω = 3π
40 , π2 are given by:
H 3 ( Ω ) Ω= 3π = 0 ; H 3 ( Ω ) Ω= π = 1 .
40 2

As the first component will pass with a gain of 0 and the second component will pass with a gain of 1, the
output signal can be expressed as:
y4, H 3 [k ] = 6 cos ( π2k ) . ▌

Problem 11.21:

The three LTID systems have the transfer functions as follows.

2
H1 (Ω ) = .
1 − 34 e − jΩ + 18 e − j 2Ω
⎧1 Ω ≤ π6
H 2 (Ω ) = ⎨
⎩0 6 < Ω ≤ π.
π

⎧0 Ω ≤ π6
H 3 (Ω ) = ⎨
⎩1 6 < Ω ≤ π.
π

The problem includes several CT sinusoidal functions which are discretized using a sampling frequency
of 22000 Hz. Let us assume that a sinusoidal signal with frequency f 0 Hz is passed through the three
systems. After sampling the CT sinusoid, the corresponding DT frequency of the DT signal is
2π f 0
given by Ω0 = . The gains of the three systems will be as follows.
22000
2
H1 ( Ω ) Ω=Ω = − jΩ 0
.
0 1− e 3
4 + 18 e − j 2 Ω0
⎧1 Ω0 ≤ π6
H 2 ( Ω ) Ω=Ω = ⎨
0
⎩0 otherwise
⎧0 Ω0 ≤ π6
H 3 ( Ω ) Ω=Ω = ⎨
0
⎩1 otherwise

(i) x1 (t ) = 2 + 3 cos(8000πt ) + 7 cos(18000πt )

The DT sequence is given by, x1[k ] = x1 (kTs ) = x1 ( 22000

k
) = 2 + 3cos( 411π k ) + 7 cos( 911π k ) .
Ω1 = 0
Ω2 = 411π Ω3 = 911π

(a) The gains of the system at Ω = 0, 411π , 911π are given by:
 Solutions 53

2 16
H1 ( Ω ) Ω=0 = = ;
1− +
3
4
1
8 3
2 2
H1 ( Ω ) Ω= 4π = − j 411π − j 811π
≈ ≈ 1.7005 − j1.6478 = 2.3679 − 0.7696
11 1− e3
4 + e
1
8
0.6066 + j 0.5878
2 2
H1 ( Ω ) Ω= 9π = − j 911π − j 1811π
≈ ≈ 1.0852 − j 0.3348 = 1.1356 − 0.2992
11 1 − 34 e + 18 e 1.6829 + j 0.5192

The output signal can then be expressed as:

y1, H 1[k ] = 163 × 2 + 3 × 2.3679 cos( 411π k − 0.7696) + 7 ×1.1356 cos( 911π k − 0.2992)
= 323 + 7.1037 cos( 411π k − 0.7696) + 7.9492 cos( 911π k − 0.2992).

(b) The gains of the system at Ω = 0, 411π , 911π are given by:
H 2 ( Ω ) Ω=0 = 1 ; H 2 ( Ω ) Ω= 4π = 0 ; H 2 ( Ω ) Ω= 9π = 0 .
11 11

As the first two components will pass with a gain of 1, and the third component will pass with a gain of 0,
the output signal can be expressed as:
y1, H 2 [k ] = 2.

(c) The gains of the system at Ω = 0, 411π , 911π are given by:
H 3 ( Ω ) Ω=0 = 0 ; H 3 ( Ω ) Ω= 4π = 1 ; H 3 ( Ω ) Ω= 9π = 1 .
11 11

As the first component will pass with a gain of 0 and the next two components will pass with a gain of 1,
the output signal can be expressed as:
y1, H 3 [k ] = 3cos( 411π k ) + 7 cos( 911π k ) .

(ii) x 2 (t ) = 2 cos(10000πt ) + 5 cos(30000πt )

The DT sequence is given by,

x2 [k ] = x2 (kTs ) = x2 ( 22000
k
) = 2 cos( 511π k ) + 5cos( 1511π k ) = 2 cos( 511π k ) + 5cos( 711π k ) .
Ω2 = 511π Ω3 = 711π

(a) The gains of the system at Ω = 511π , 711π are given by:
2 2
H1 ( Ω ) Ω= 5π = − j 511π − j 1011π
≈ ≈ 1.4085 − j1.2880 = 1.9086 − 0.7407
11 1− 4 e
3
+8e
1 0.7733 + j 0.7071
2 2
H1 ( Ω ) Ω= 7 π = − j 711π − j 1411π
≈ ≈ 1.1626 − j 0.7343 = 1.3751 − 0.5633
11 1− 4 e
3
+8e
1 1.2297 + j 0.7767

The output signal can then be expressed as:

y2, H 1[k ] = 2 ×1.9086 cos( 511π k − 0.7407) + 5 ×1.3751cos( 711π k − 0.5633)
=3.8172 cos( 511π k − 0.7407) + 6.8755cos( 711π k − 0.5633).

(b) The gains of the system at Ω = 5π

11 , 711π are given by:
54 Chapter 11

H 2 ( Ω ) Ω= 5π = 0 ; H 2 ( Ω ) Ω= 7 π = 0 .
11 11

As both the components will pass with a gain of 0, the output y2, H 2 [ k ] = 0.

(c) The gains of the system at Ω = 5π

11 , 711π are given by:
H 3 ( Ω ) Ω= 5π = 1 ; H 3 ( Ω ) Ω= 7 π = 1 .
11 11

As both the components will pass with a gain of 1, the output

y2, H 1[k ] = 2 cos( 511π k − 0.7407) + 5cos( 711π k − 0.5633) .

(iii) x3 (t ) = 5 cos(600πt ) + 9 cos(900πt ) + 2 cos(3000πt )

The DT sequence is given by, x3 [k ] = x3 (kTs ) = x3 ( 22000

k
) = 5cos( 110
3π
k ) + 9 cos( 220
9π
k ) + 2 cos( 322π k ) .
3π
Ω0 = 110 9π
Ω1 = 220 Ω 2 = 322π

(a) The gains of the system at Ω = 110 3π 9π 3π

, 220 , 22 are given by:
2 2
H1 ( Ω ) Ω= 3π = 3π 3π ≈ ≈ 5.2520 − j 0.5989 = 5.2860 − 0.1135 ;
1 − 34 e 110 + 18 e 55 0.3759 + j 0.0429
− j − j
110

2 2
H1 ( Ω ) Ω= 9π = 9π 9π ≈ ≈ 5.1538 − j 0.8795 = 5.2283 − 0.1690
220
1− e 3
4
− j 220
+ e
1
8
− j 110
0.3771 + j 0.0643
2 2
H1 ( Ω ) Ω= 3π = − j 322π − j 311π
≈ ≈ 3.8642 − j 2.0992 = 4.3976 − 0.4976
22 1− e 3
4 + e 1
8
0.3996 + j 0.2171

The output signal can then be expressed as:

y3,H 1[k ] = 5 × 5.2860cos( 110
3π
k − 0.1135) + 9 × 5.2283cos( 220
9π
k − 0.1690) + 2 × 4.3976cos( 322π k − 0.4976)
= 26.43cos( 110
3π
k − 0.1135) + 47.0547 cos( 220
9π
k − 0.1690) + 8.7952 cos( 322π k − 0.4976).

(b) The gains of the system at Ω = 110

3π 9π 3π
, 220 , 22 are given by:
H 2 ( Ω ) Ω= 3π = 1 ; H 2 ( Ω ) Ω= 9π = 1 ; H 2 ( Ω ) Ω= 3π = 1 .
110 220 22

As all there components will pass with a gain of 1, the output is given by:
y3,H 2 [k ] = x3[k ] = 5cos( 110
3π
k ) + 9 cos( 220
9π
k ) + 2 cos( 322π k ) .

(c) The gains of the system at Ω = 110

3π 9π 3π
, 220 , 22 are given by:
H 3 ( Ω ) Ω= 3π = 0 ; H 3 ( Ω ) Ω= 9π = 0 ; H 3 ( Ω ) Ω= 3π = 0 .
110 220 22

As all there components will pass with a gain of 0, the output is given by:
y3,H 3[k ] = 0 .

(iv) x 4 (t ) = 4 cos(28000πt ) + 6 cos(18000πt )

 Solutions 55

The DT sequence is given by,

x4 [k ] = x4 ( kTs ) = x4 ( 22000
k
) = 4 cos( 1411π k ) + 9 cos( 911π k ) = 4 cos( 811π k ) + 9 cos( 911π k ) .
Ω0 = 811π Ω1 = 911π

(a) The gains of the system at Ω = 8π

11 , 911π are given by:
2 2
H1 ( Ω ) Ω= 8π = − j 811π − j 1611π
≈ ≈ 1.1130 − j 0.5216 = 1.2291 − 0.4383 ;
11 1− e 3
4 + 81 e 1.4734 + j 0.6905
2 2
H1 ( Ω ) Ω= 9π = − j 911π − j 1811π
≈ ≈ 1.0852 − j 0.3348 = 1.1356 − 0.2992
11 1 − 43 e + 18 e 1.6829 + j 0.5192

The output signal can then be expressed as:

y4,H 1[k ] = 5 × 5.2860cos( 110
3π
k − 0.1135) + 9 × 5.2283cos( 220
9π
k − 0.1690) + 2 × 4.3976cos( 322π k − 0.4976)
= 26.43cos( 110
3π
k − 0.1135) + 47.0547 cos( 220
9π
k − 0.1690) + 8.7952 cos( 322π k − 0.4976).

(b) The gains of the system at Ω = 8π

11 , 911π are given by:
H 2 ( Ω ) Ω= 8π = 0 ; H 2 ( Ω ) Ω= 9π = 0 .
11 11

As both the components will pass with a gain of 0, the output is given by:
y4,H 2 [k ] = 0 .

(c) The gains of the system at Ω = 8π

11 , 911π are given by:
H 3 ( Ω ) Ω= 8π = 1 ; H 3 ( Ω ) Ω= 9π = 1 .
11 11

As both the components will pass with a gain of 1, the output is given by:
y4,H 3[k ] = x4 [k ] = 4 cos( 811π k ) + 9 cos( 911π k ) . ▌