Chapter 1: Problem Solutions

Review of Signals and Systems

Signals

à Problem 1.1

a) xn  0.5n  1  n  0.5n  1  n  2  0.8n  3

b) xn  0.5n  5  n  4  0.5n  3  n  2  0.8n  1

à Problem 1.2

a) I  e1
b) I  e1
c) I  0 since the interval of integration does not include the point t  1, where the impulse is
centered.
d) I  1
e) I  cos2 0.1
f) I  e
g) let    
1 t, then t  2 and dt  2d. Substitute in the integral to obtain
2

I   22   

1 2d 

2


2 d  2
  82   

1

h) let   3t, then t    3 and dt   

3 d. Then the integral becomes

1

3 d  
I   e3   1 

13
3 
1 e

2 Solutions_Chapter1[1].nb

à Problem 1.3

Amplitude A  2,
Period T0  0.01 sec , then frequency F0  
T0  100Hz  0.1kHz
1

Phase   0
Then the signal can be written as xt  2cos200t.

à Problem 1.4

a) Frequency F0  1  T0  1  3  103   

3  10 Hz. Then
1 3

3 1000t  15 
xt  2.5cos 
2 0

b) Digital Frequency 0  2F0  Fs  2  6    3 rad. Therefore the sampled sinusoid

becomes

3 n  15 

xn  2.5cos  0

à Problem 1.5

All sinusoids are distinct. In continuous time there is no ambiguity between frequency and signal.

à Problem 1.6

First bring all frequencies within the interval  to . This yields

x2 n  2cos1.5n  0.1  2n  2cos0.5n  0.1
 2cos0.5n  0.1
and also
x3 n  2cos1.5n  0.1  2n  2cos0.5n  0.1
 2cos0.5n  0.1
Therefore we can see that x1 n  x2 n and x3 n  x4 n.
A different way of solving this problem is graphically. The frequency plots for all four signals are
shown next. The key point is to understand that in discrete time, all frequencies within the interval
,  yield complex exponetials with distinct values:
Solutions_Chapter1[1].nb 3

e  j 0.1 e j 0.1

    
2 2
x1 n Frequency Representation

e j 0.1 e  j 0.1

3
  3 
2 2

shift by  2 shift by  2

e  j 0.1 e j 0.1

    
2 2
x2 n Frequency Representation
4 Solutions_Chapter1[1].nb

e  j 0.1 e j 0.1

3
  3 
2 2

shift by  2 shift by  2

e j 0.1 e  j 0.1

    
2 2
x3 n Frequency Representation

e j 0.1 e  j 0.1

    
2 2
x4 n Frequency Representation

Again you see that x1 n and x2 n have the same representation, and the same for x3 n and
x4 n.

à Problem 1.7

The sinusoid xn  3cos1.9n  0.2 has the same samples, since
3cos0.1n  0.2  3cos0.1n  2n  0.2
Solutions_Chapter1[1].nb 5

à Problem 1.8

a) since 150  15    180    12, we can write

xt   j12 ej100t  ej12 ej100t
2 e
3

b) xt  ej0.1 ej10t  ej0.1 ej10t

c) xt  2.5jej0.2 ej20t  2.5jej0.2 ej20t
d) xt  5jej1000t  5jej1000t
e) xt  1.5jej0.2t ej200t  1.5jej0.2tej200t

à Problem 1.9

a) since 2  j  2.2361ej0.4636 , we can write xt  4.4722cos100t  0.4636

 
b) since 1  2ej0.1  2.9672ej0.2098 , then xt  5.9344cos10t  0.2098

c) the signal can be written as xt = ÅÅÅÅ12 sin5 pt + 0.1 p  2 e j ÅÅ4ÅÅ e j10pt + 2 e- j ÅÅÅÅ4Å e-j10pt  and therefore
p p

 
 sin5t  0.1cos10t  
xt  
1
2 4

à Problem 1.10

a) 
xt  j20  xt
b)  xtdt  
j20 xt
1

c) xt  0.1  ej2xt  xt

à Problem 1.11

In order to solve this problem we have to decompose the signals into complex exponentials.
a) xt  2cos100t  0.2  ej0.2 ej100t  ej0.2 ej100t
6 Solutions_Chapter1[1].nb

magnitude

1 1

 50 50 F (Hz)

phase
0.2

 50
 0.2
50 F (Hz)

b) xt   j0.1 ej100t   j0.1 ej100t

2 e 2 e
3 3

  j0.2 ej110t

2j e
j0.2 ej110t  
2j e
3 3

This becomes
xt   3 ej0.1 ej100t  
3 ej0.1 ej100t
2 2

3
ej0.7 ej110t  
3
ej0.7 ej110t
2 2

magnitude
3 3 3 3
2 2 2 2

 55  50 50 55 F (Hz)

phase
0.7
0.1

 0.1
F (Hz)
 0.7

c) xt  ej0.2 ej100t  ej0.2 ej100t

  j0.1 ej100t
2j e
j0.1 ej100t  
2j e
3 3
Solutions_Chapter1[1].nb 7

Combining terms we obtain

xt  0.9071ej1.1800ej100t  0.9071ej1.1800ej100t
magnitude
0.9071 0.9071

 50 50 F (Hz)

phase
1.18

 50
 1.18
50 F (Hz)
  
d) xt  2  5ej 4 ej1000t  5 ej2.67795ej1100t
magnitude
2 5
5

 550 500 F (Hz)

phase

2.68 4

F (Hz)
e) xt  ej0t   j10t   j20t
2 e
j10t  
2 e 2j e
j20t  
2j e
1 1 1 1

This becomes
 
xt  ej0t   j10t  j 
2 ej20t   j 
2 ej20t
2 e
j10t  
2 e 2 e 2 e
1 1 1 1
8 Solutions_Chapter1[1].nb

magnitude
1
1 1 1 1
2 2 2 2

 10 5 5 10 F (Hz)

phase

2

 F (Hz)
2

à Problem 1.12

Magnitude A
A
2
2

 0  0  (rad )
Phase

 0 

0  (rad )

Referring to the figure above:
a) 0    2
b) 0  2  3
c) 0  
Solutions_Chapter1[1].nb 9

d) 0  0.75, since 1000  800  1.25  . We want the frequency to be between  and
 , we use the fact that this sinusoid has the same samples as 0  2  1.25  0.75.
e) 0  0, since 1000  500  2 and therefore 0  2  2  0.

à Problem 1.13

a) Since 0  0.2  2F0  2000 we solve for the frequency as F0  200Hz.

b) Sinusoids with frequencies F0  Fs  2.2kHz, or Fs  F0  1.8kHz have the same sample
values.

à Problem 1.14

a) Since  
xt  ej100t  ej100t  
2 e
3 j 
2 ej120t  
2 e
3 j 
2 ej120t  2ej150t  2ej150t

we btain the frequency plot shown below.

magnitude
2
1.5

 75  50 50 60 75 F
 60
b) The maximum frequency is FMAX  75Hz. Therefore the sampling frequency has to be such that
Fs  150Hz.

à Problem 1.15

The bandwith is 4kHz and therefore the sampling frequency has to be at least Fs  8kHz. This

16  8000  128kbytes  sec. For one minute of data we need at least

means that we need 8000 samples for every second of data, and therefore

60  120  7200kbytes  7.2Mbytes to store the file.

10 Solutions_Chapter1[1].nb

Systems

à Problem 1.16

If a property is not specified it is assumed Linear, Time Invariant, Causal, BIBO Stable.
b) Non Linear. In fact if you apply superposition:
x1 n  S  y1 n  2x1 n  1
x2 n  S  y2 n  2x2 n  1
and therefore
xn  x1 n  x2 n  S  yn  2x1 n  x2 n  1
which shows that the output is different from y1 n  y2 n  2x1 n  x2 n  2.
c) Time Varying, due to the time varying coefficient "t". Non BIBO Stable, since xt  ut , a
Bounded Input , yields yt  tut, a Non Bounded Output;
d) Time Varying,due to the coefficient et ;
e) Non Linear. If xt  x1 t  x2 t then the output is yt   x1 t  x2 t 
different from y1 t  y2 t   x1 t    x2 t ;
f) Non Causal, since the output at any time t depends on a future input at time t  1;
h) Non Causal since the output at time n depends on a future intput at time n  1;
i) Non Linear (due to the term x2 ). Time Varying (due to the time varying coefficient n). Not BIBO
Stable, since yn  nun  ... , ie Non Bounded, when the input xn  un, a Bounded
input.

à Problem 1.17

a) yn   uk0.5nk un  k.
k

If n  0:
n n n1
yn   0.5nk   0.5k  
10.5
  2  0.5n
10.5
k0 k0

If n  0 then ukun  k  0 for all k, and therefore yn  0.

Solutions_Chapter1[1].nb 11

Therefore yn  2  0.5n un.


b) yn   uk  20.5nk un  k.
k

If n  2:
n n
yn   0.5nk   0.5k  0.5n  0.5n1 

  0.5  0.5n1  21  0.5n2 

k2 k0
n1
 
10.5
10.5
n

If n  2 then uk  2un  k  2 for all k, and therefore yn  0.

Therefore yn  2  0.5n2 un  2.

c) yn   0.8k uk0.5nk un  k.
k

If n  0:
n n n1
yn   0.8k 0.5nk  0.5n   1.6k  0.5n  
11.6   1.25  0.5n  2  0.8n
11.6
k0 k0

If n  0 then ukun  k  0 for all k, and therefore yn  0.

Therefore yn  1.25  0.5n  2  0.8n un.

d) yn   1.2k uk0.5nk un  k
k

If n  0 then ukun  k  un  k. This implies

n n
yn   1.2k 0.5nk  0.5n   2.4k 
k k

0.5n   2.4k 

0.5n   2.4k   2.4k  

kn
 n1

 
k0 k0
12.4n
0.5  
n 1

12.41
 
12.41

 1.7143 1.2 n

If n  0 then ukun  k  uk. This implies

0 0
yn   1.2k 0.5nk  0.5n   2.4k 
k k

 0.5n   2.4k  1.7143  0.5n
k0

Finally we can write the answer yn  1.7143  1.2n un  1  1.7143  0.5n un.

e) yn   ej0.2nk 0.5k uk    0.5k  ej0.2k ej0.2n

 
k k0

Applying the geometric series we obtain

12 Solutions_Chapter1[1].nb

yn  
1
 ej0.2n  1.5059ej0.2n0.4585
10.5ej0.2

f) yn   ej0.2nk un  k0.5k uk    0.5k  ej0.2k ej0.2n if n  0.

 n
k k0
Then, applying the geometric series we obtain

yn    
 ej0.2n 
j0.2 n1

 1.5059ej0.2n0.4585  ej0.2 0.5n1 

10.5e 
10.5 e
j0.2

and yn  0 when n  0. Therefore

yn  1.5059ej0.2n0.4585  ej0.2 0.5n1 un

g) Since xn  ej0.2n  ej0.2n , using the solution to problem e) above we obtain
yn  1.5059  ej0.2n0.4585  ej0.2n0.4585 
 3.0118 cos0.2n  0.4585
h) Since xn  ej0.2n un  ej0.2nun, using the solution to problem f) above we obtain
yn  3.0118cos0.2n  0.4585  cos0.20.5n1 un
 3.0118cos0.2n  0.4585  1.2183  0.5n un

à Problem 1.18

a) hn  n  n  1  n  2

b) yn  un  un  1  un  2
c) yn  ej0.5n  ej0.5nn1  ej0.5nn2 . Therefore, since ej0.5  j, we can write
yn  1  j  1ej0.5n  ej0.5n0.5
d) yn  cos0.5n  0.5

à Problem 1.19

To determine the impulse response n substitute xn  n:

a) hn  3n  2n  1  n  2
b) hn has to satisfy the recursion hn  hn  1  n. Therefore:
if n  0 then hn  0;
Solutions_Chapter1[1].nb 13

if n  0 then h0  h1  1  1, since h1  0;

if n  0 then hn  hn  1 that is to say hn is a constant, which must be equal to
h0  1.
Putting everything together you can verify that hn  un;
c) hn  2n  2
For the next three problems recall that any signal xn can be written as

xn   xkn  k.
k

d) yn   0.5k n  k  0.5n un
k0

e) yn   0.5k n  k  0.5n


k
 
f) hn   0.5k2 uk  5n  k  2   0.5k uk  3n  k. Therefore the
k k
impulse response is hn  0.5n un  3.

à Problem 1.20

We have to check whether or not   hn   . Recall that the geometric series  an   if
 

and only if  a   1.
 n0


a)  0.5n   , then BIBO Stable;
n0
 1  
b)  0.5n   0.5n   0.5n   0.5n   , then BIBO Stable;
n0 n n0 n1
 1  
c)  0.5n   0.5n   0.5n   0.5n   then it is NOT BIBO Stable
n0 n n0 n1
0 
d)  0.5n   2n   then NOT BIBO Stable
n n0

e)   
n  , then NOT BIBO Stable
1
n1

f)   
n2  , then BIBO Stable
1
n1

g)  1  , then NOT BIBO Stable

h)   ej0.2n    1  , then NOT BIBO Stable

n0
 

i)   ej0.2n    1  then NOT BIBO Stable

n n
 

j)   0.8n ej0.2n    0.8n   then BIBO Stable

n0 n0
 
n0 n0
14 Solutions_Chapter1[1].nb

k)   0.8n cos0.2n    0.8n   0.8n   0.8n   0.8n   then

  1  
n n0 n n0 n1
BIBO Stable
0 
l)  2n   0.5n   then BIBO Stable
n n0

m)  2n   then NOT BIBO Stable
n0

z-Transforms

à Problem 1.21

z  , ROC  0.5z1   1, ie  z   0.5;


a) Xz   0.5n zn  
1  
10.5z 1 z0.5

z  , ROC  0.51 z   1, ie

n0
1 
b) Xz   0.5n zn   0.5n zn  
12z  1   
 z   0.5;
1
n n1 z0.5

12z , ROC  z   0.5;

0 
c)  0.5n zn   0.5n zn  
1

 , ROC  ej0.4 z1   1, ie  z   1;

n n0
 n
d) Xz   ej0.4 zn  
1ej0.4 z
1
1  
z
zej0.4
n0
1 n  n  n  n
e) Xz   ej0.4 zn   ej0.4 zn   ej0.4 zn   ej0.4 zn . The first
series converges when  z   1 and the second when  z   1. Therefore there is no common
n n0 n1 n0

ROC and Xz for this signal DOES NOT exist!

1 n  n
f) Xz   ej0.4 zn   ej0.4 zn  
1ej0.4 
z  1   
z , ROC

 ej0.4 z   1, ie  z   1;
1
n n1 zej0.4

1   
g) Xz   0.5n zn   0.5n zn   0.5n zn   0.5n zn . See the two series:

z2 , when  z   2;
n n0 n1 n0

 0.5n zn  
1
  1   
10.5z
z
n1

 0.5n zn  

1
10.5z 1  
z
z0.5  , when |z|>0.5.
n0

Therefore the ROC has to be the intersection of the ROC's, which yields
Xz   z2  , ROC: 0.5   z   2
z   
z
z0.5
Solutions_Chapter1[1].nb 15

h) Xz  1  z1  z2 , ROC: 0   z ;

i) Xz  z  z2  z3 , ROC  z   

à Problem 1.22

Recall

za ROC:  z    a 
Zan un  
z

za ROC:  z    a 
Zan un  1   
z

a) xn  2  0.5n1 un  1, then Xz  2z1  

z
z0.5   
2
z0.5 
z 0.5
b) xn  0.5n un  1  n, then Xz  
  1  

d) xn  0.5  ej0.1  un, then Xz  
z0.5 z0.5
n z

e) xn  0.5n un, then Xz  
z0.5ej0.1
z

1 0.5ej0.1  un   1 0.5ej0.1  un, then
z0.5
n n
f) xn  
2 2
Xz  
2  
1 z
z0.5ej0.1  
2  
1 z
 . Combining terms w eobtain
z0.5ej0.1
z cos0.1z
2
Xz  

z2 cos0.10.25

0.5   z   2
g) xn  0.5n un  0.5n un  1, then Xz  
z
z0.5   
z 1.5z
z2  
z2 2.5z1  , ROC:

1.5z 1.5
h) Using the result in g) above, Xz  z1  
z2 2.5z1   
z2 2.5z1 

à Problem 1.23
j2.6180 j2.6180
a) 
z   
Xz
z  
2 0.5774e
zej2.0944  
0.5774e
 therefore
zej2.0944
z z
  0.5774ej2.6100 
Xz  1  0.5774ej2.6100  
z  ej2.0944 z  ej2.0944
Since  z   1 all terms are causal. Therefore
xn  n  0.5774ej2.0944n2.6100 un  0.5774ej2.0944n2.6100 un
Combining terms we obtain
xn  n  1.0548cos2.0944n  2.6100un
b) Same Xz as in the previous problem, but different ROC. All terms are noncausal, and therefore
xn 
n  0.5774ej2.0944n2.6100 un  1  0.5774ej2.0944n2.6100 un  1
and, after combining terms
16 Solutions_Chapter1[1].nb

xn  n11.0548cos2.0944n  2.6100un  1

c) 
z   
Xz
z  
0.5
z1  
2
z2 , therefore Xz  0.5  2 
1.5
z1  1.5 
z z
z2 . Form the ROC all
terms are causal. Thereofore
xn  0.5n  2un  1.5  2n un
d) Same Xz with different ROC. In this case all terms are anticausal,
xn  0.5n  2un  1  1.5  2n un  1
e) Same Xz with different ROC. In this case 
z1 yields a causal term, and 
z z
z2 yields an anti-
causal term. Therefore
xn  0.5n  2un  1.5  2n un  1

 z   1 all terms are causal, therefore

f) 
z   
Xz
z2 1  
1
zj   
0.5j
zj  therefore Xz  0.5j 
0.5j
zj  0.5j 
z z
zj . From the ROC

 
xn  0.5jej 2 n un  0.5jej 2 n un
Combining terms:

2 un
 
xn  cos 
2 n  

g) Same Xz but all anticausal:

 
xn  0.5jej 2 n un  1  0.5jej 2 n un  1
Combine terms:

2 un  1
 
xn  cos 
2 n  

à Problem 1.24

a) Hz  

1
10.5z 1  
z
z0.5  Since it is causal, the impulse response is hn  0.5n un
2 2
b) Hz  
1
  
10.5z z2 . Since 
z   
Hz
  
zz2 z  
1
z2 then Hz  1  
1 z
z2 . The
impulse response is anticausal, form the difference equation, therefore the impulse response is
hn  n  2n un  1.
1z z 1 2 z z12 z z1 2 1
c) Hz  
1z
2   
z2 1 . Therefore 
z   
Hz
zz21
  
z  
z1  
1.5 0.5
z1 which leads to
the decomposition Hz  1  1.5  z1  0.5 
z z
z1 . Since the difference equation is causal, then
the impulse response is causal, which leads to hn  n  1.5un  0.51n un.
2
d) Hz  
1z1
1
z2  
z
 , then 
z2 z1 z   
Hz z
  
z2 z1   
0.2764
z0.6180  and therefore
0.7236
z1.6180
Hz  0.2764    0.7236 
z
z0.6180  . The difference equation is causal and therefore
z
z1.6180
the impulse response is hn  0.27640.6180n un  0.7236  1.6180n un
Solutions_Chapter1[1].nb 17

à Problem 1.25
z
z0.5  , with ROC  z   0.5. The unit
a) The system is causal with transfer function Hz  
circle is in the ROC and therefore the system is BIBO stable. The impulse response is
hn  0.5n un.

depends from future values only) the ROC is  z   0.5 and the impulse response
b) Now Hz  2 
12z  
z z  , same as before. But since it is clearly anticausal (the output
z0.5

hn  0.5n un  1. The system is NOT BIBO Stable, since the unit circle is not within the
ROC.

z1 with ROC  z   1.

c) If the input is xn  un, its z-Tranform is Xz  
z

When the system is causal then

Yz  
  
z1    z1 , ROC  z   1
  2 
z z z z
z0.5 z0.5

therefore yn  0.5n un  2un;

intersection of  z   1 and  z   0.5, which is empty. Therefore in this case the output yn
When the system is anticausal, the z-Transform is the same, but with different ROC given by the

does not have a z-Transform.

à Problem 1.26

a) The difference equation can be implemented in a number of ways, so we have not enough informa-
tion to assess causality or the ROC of the transfer function.

z2z0.5
1
b) The transfer function is Hz  
z
12.5z1 z 
2  
z
z2 2.5z1   
z
 . The possible
ROC's are the following:
 z   2, the system is causal , and NOT BIBO stable, since the unit circle is not in the ROC;
0.5   z   2, the system is non causal, but BIBO Stable, since the unit circle is in the ROC;
 z   0.5, the sytem is anticausal, NOT BIBO Stable since the unit circle is not in the ROC.
c) The step response has transfer function
z2z0.5z1
z2
Yz  

  0.6667 
z
  2 
z1  1.333 
z z

When the sytem is causal the ROC is  z   2, then

z0.5 z2

yn  0.66670.5n un  2un  1.333  2n un;

When the system is non casual, the ROC is 1   z   2, then
yn  0.66670.5n un  2un  1.333  2n un  1;
When the system is anticausal, the ROC of the transfer function (0.5   z ) and the ROC of the
input (  z   1) have empty intersection, and therefore the ouput hs no z-Transform.
18 Solutions_Chapter1[1].nb

Frequency Response

à Problem 1.27

The transfer function of the system is

Hz  

1
10.5z 1  
z
z0.5 
Therefore the system is stable (pole at z  0.5 inside the unit circle) and its frequency response is
j
H  
e

ej 0.5

a) the input is a complex exponential with frequency   0.2 and therefore the corresponding
output is yn  H0.2ej0.2n . Substituing for the frequency response H we obtain
yn  1.5059ej4585ej0.2n  1.5059ej0.2n0.4585 .
b) Since xn  ej0.1 ej0.5n  ej0.1 ej0.5n the response is
yn  H0.5ej0.1 ej0.5n  H0.5ej0.1 ej0.5n . Substituing for H this
yields
yn  0.8944ej0.4636ej0.1 ej0.5n  0.8944ej0.4636ej0.1 ej0.5n
 0.4472cos0.5n  0.1495
1
10.5  2 and H0.3  1.2289ej5202 then
c) Since H0  
yn  1  2  5  1.2289cos0.3n  0.5  0.5202 which yields
yn  2  6.1443cos0.3n  2.0910;
d) H0  2 and H  0.6667, then yn  2  0.66671n ;
e) H0.2  1.5059ej4585 , therefore
yn  1.5059sin0.2n  0.4585  1.5059cos0.2n  0.4585;

à Problem 1.28

The system has transfer function Hz  Zhn  Z0.8n un  1.25n un  1. This
yields
Hz  
z
z0.8
z
z1.25 , ROC: 0.8   z   1.25
  
The unit circle is within the ROC therefore the system is stable and it has a Frequency Response
ej ej
H  
  
ej 0.8  .
ej 1.25
Solutions_Chapter1[1].nb 19

a) H0.2  1.0417, then yn  1.0417ej0.2n ;

b) H0.5  0.2195, then yn  0.4390cos0.5n  0.1;
c) H0  9, H0.3  0.5146, then yn  9  2.5731cos0.3n  0.5;
d) H0  9, H  0.1111, therefore yn  9  0.11111n ;
e) H0.2  1.0417, then yn  1.0417sin0.2n  1.0417cos0.2n

à Problem 1.29

given by  z   0.5 and the unit circle is NOT in the ROC. As a consequence the system is
2z
The system ha stransfer function Hz  
12z  
z
z0.5  and it is anticausal.Therefore the ROC is

UNSTABLE and the steady state frequency response does not exist.

à Problem 1.30

Therefore ROC:  z   1, the unit circle is NOT in the ROC and the system is NOT stable. Again
The transfer function is Hz  
1z1
z
z2 , the system is causal and the poles are on the unit circle.

the frequency response does not exist!

à Problem 1.31

a) From the graph H0.1  6.5ej . Therefore the output is

yn  13cos0.1n  1
b) From the graph we obtain H0  8, H0.5  1.8ej and H  0. Therefore the
output is
yn  16  1.8cos0.5n  

Time, z-, and Frequency Domains

à Problem 1.32

a) First we have to determine the digital frequency of the disturbance, as 0  2F0  Fs  0.4.
Therefore the filter must have at least two zeros at z1,2  ej0.4 , and its transfer function becomes
20 Solutions_Chapter1[1].nb

b0 z b1 zb2
Hz  b0  b1 z1  b2 z2  
2

  b0 1  0.6180z1  z2 

z2 
ze j0.4ze j0.4 
 b0  
z2 
The poles are both z  0.
b) With the constraint that H0  1, ie that Hz  1 when z  1 we obtain an ewuation to solve
for the coefficient b0 . Therefore
H1  b0 1  0.6180  1  1

which yields b0  0.7236. Therefore the difference equation of the filter becomes
yn  0.7236xn  0.4472xn  1  0.7236xn

à Problem 1.33

If we keep the same zeros at z1,2  ej0.4 , we add two poles close to the zeros as, say,
p1,2  0.95ej0.4 , then the frequency response becomes more selective. In this way the Transfer
Function becomes
zz1 zz2 
zp1 zp 2
z 0.6180z1
2
Hz  b0  

  b0  

z2 0.5871z0.9025

In order to have the Frequency Response with H0  1, we need to solve for the coefficient b0
from the equation
b0  
10.61801
  1
10.58710.9025

which yields b0  0.9518. Finally the difference equation for the filter becomes
yn  0.5871yn  1  0.9025yn  2 
0.9518xn  0.5883xn  1  0.9518xn  2

à Problem 1.34

Since xn is periodic with period N, by defintion xn  xn  N for all n, and therefore
xn  xn  N  0 for all n. Therefore in this case the difference equation becomes
yn  a1 yn  1  a2 yn  2  0
and if the system is stable yn  0 from any initial condition.
a) When N  4 the filter has Transfer Function
z 1 
Hz  b0  
4
z2 a1
za2

The zeros must be the solutions of the equation z4  1, ie z4  ejk2 with k integer. This yields
z  ejk2 , for k  0, 1, 2, 3, Therefore the four zeros are z  1, j, 1, j
b) Choose the poles close to the zeros, inside the unit circle, at z  , j, , j, with
0    1, close to one.
Solutions_Chapter1[1].nb 21

à Problem 1.35

The digital frequencies of the disturbance are 1  2100  1000  0.2,

2  2150  1000  0.3, 3  2200  1000  0.4.
a) zeros at z1,2  ej0.2 , z3,4  ej0.3 , z5,6  ej0.4 , and poles at p1,2  ej0.2 ,
p3,4  ej0.3 , p5,6  ej0.4 with  again positive, close to one. Say   0.9.
b) From the zeros and the poles the transfer function becomes
z 3.4116z 6.6287z 7.9988z 6.6287z 3.4116z1
6 5 4 3 2
Hz  b0  
z6 3.0705z5 5.3692z 
4 5.8312z3 4.3491z 
2 2.0145z0.5314 
If we want the DC gain to be one, ie H  1 when   0, then b0  0.7664;
c) The difference equation then becomes
yn  3.0705yn  1  5.3692yn  2  5.8312yn  3 
4.3491yn  4  2.0145yn  5  0.5314yn  6 
 0.7664xn  3.4116xn  1  6.6287xn  2 
7.9988xn  3  6.6287xn  4  3.4116xn  5  xn  6

à Problem 1.36

The digital frequency of the signal we want to enhance is 0  2  1.5  5  0.6 radians.
a) The poles are p1,2  ej0.6 with 0    1, close to one. We can choose any zeros we like,
say z1,2   1 to attenuate the low frequencies (  0) and the high frequencies (  ).
b) Choose (say)   0.9, and the transfer function becomes
z 1 2
Hz  b0  

z2 0.5562z0.81

response at frequency 0 to have unit magnitude, we impose  H0.6   1 which yields

We can choose b0 to satisfy any normalization we like. For example, if we want the frequency

b0  0.0950.
c) The difference equation becomes
yn  0.5562yn  1  0.81yn  2  0.0950xn  0.0950xn  2

à Problem 1.37

a) The filter has zeros at z1,2   j, poles at p1,2   j0.9. Its frequency response is as shown;
b) Notch Filter;
c) yn  0.81yn  2  xn  xn  2.
22 Solutions_Chapter1[1].nb

Fourier Analysis of Discrete Time Signals

à Problem 1.38
2
a) Period N  3, therefore Xk   xnej 3 kn , for k  0, 1, 2. Just substitute for one
2

n0
period as x0  1, x1  2, x2  3 to obtain
X0  6.000, X1  1.5  j0.8660, X2  1.5  j0.8660.

b) Period N  2, then Xk   xnej 2 kn  x0  1k x1, for k  0, 1. Therefore
1 2

X0  1  1  0, X1  1  1  2.

n0

3 3
c) Period N  4, then Xk   xnej 4 kn   xnjkn , for k  0, 1, 2, 3.
2

n0 n0
Substitute the numerical values of the sequence to obtain X0  X2  0, X1  X3  2 or,
in vector form,
X  0, 2, 0, 2
d) The signal xn   cos 
4 n  is shown below.


| cos 4 n  |
1 1

 2

n
Plot of xn   cos 
4 n 


The period can be seen by inspection as N  4 and therefore the expansion is of the form
3 
1   Xkej 
xn   2 kn , with the DFS being
4 k0
3
Xk   xnjkn , for k  0, 1, 2, 3
n0

since ej 
2  j. Therefore
Solutions_Chapter1[1].nb 23

X  2.4142, 1, 0.4142, 1

à Problem 1.39

The transfer function of the system is Hz  

0.2z
z0.8  , by inspection. The system is causal and stable
therefore we can define the frequency response as
j
H  
0.2e

a) The DFS of the signal is X  6, 1.5  j0.866, 1.5  j0.866. If we call yn the
ej 0.8

N Xk, for k  0, 1, 2. By substitution we find Y0  

output signal, it is going to be periodic with the same period N  3 and DFS given by
Yk  Hk  2 0.2
10.8  6  6,
2
j 
 
Y1  
0.2e
2  1.5  j0.866  0.123  j0.1846 and
3

ej  0.8

3
2
j2 
3
Y2  
0.2e  1.5  j0.866  0.123  j0.1846. In vector form:

Y  6, 0.123  j0.1846, 0.123  j0.1846

2
ej2  0.8
3

b) Period N  2, therefore Yk  Hk 2 Xk for k  0, 1. This yields

2

1k 0.8
k
Yk  
 
0.21
  Xk for k  0, 1. Substitute te numerical values of Xk to obtain

Y  0, 0.5
c) Period N  4, therefore Yk  Hk  4 Xk for k  0, 1, 2, 3. This yields
2

jk 0.8
k
Yk  
 Xk. Substitute for Xk to obtain
0.2j

Y  0, 0.2439  j0.1951, 0, 0.2439  j0.1951

à Problem 1.40
3
a) The length of the sequence is N  4. Therefore the DFT becomes Xk   xnej 4 kn for
2

n0
k  0, 1, 2, 3. Substitute for the sequence to obtain
X  0, 0, 4, 0
3
b) Length N  4, then again Xk   xnej 4 kn for k  0, ..., 3 to obtain
2

X  4, 0, 0, 0
n0

2
c) Length N  3, then Xk   xk  ej 
3  kn for n  0, 1, 2. This yields
2

X  6.0000, 1.5000  j0.8660, 1.5000  j0.8660

n0

3
d) Length N  4, then again Xk   xnej 4 kn for k  0, ..., 3 to obtain
2

n0
24 Solutions_Chapter1[1].nb

X  0, 2, 0, 2
3
e) Length N  4, then again Xk   xnej 
4  kn for k  0, ..., 3 to obtain
2

X  1  j2.4142, 1  j2.4142, 1  j0.4142i, 1  j0.4142

n0

3
f) Length N  4, then again Xk   xnej 
4  kn for k  0, ..., 3 to obtain
2

X  4.0961, 4.1904  j2.4938, 2.7687, 4.1904  j2.4938

n0

7
g) Length N  8, then Xk   xnej 
8  kn for k  0, ..., 7 to obtain
2

X  0, 7.0534  j9.7082, 0, 0, 0, 0, 0, 7.0534  j9.7082

n0

à Problem 1.41

Recall the definition X  DTFTxn   xnejn . Then:
n

a) X   0.8n ejn   0.8n ejn   0.8n ejn . After some simple
 1 
n n n0
manipulations:
 
X   0.8n ejn  1   0.8n ejn  

1
10.8e j  1  
1
10.8e   
j 
0.36
11.6cos0.64
n0 n0

An alternative way would be to use the z-Transforms. Since, in this case

xn  1.25n un  1  0.8n un
the z-Transform is Xz   
z
 , ROC : 0.8   z   1.25
  
z

the unit circle  z   1 is within the ROC. Therefore just substitute z  ej to obtain:
z1.25 z0.8

j j
X   
e
  
ej 1.25
e

ej 0.8

and the result follows with some algebra.


b) X   0.5n ejn  
j
1
10.5e   
j
e

ej 0.5
n0

c) xn  1 when 0  n  4 and xn  0 otherwise. Therefore

4 j5
X   ejn  
1e
1ej  ej1.5  
sin
sin2.5
n0
Solutions_Chapter1[1].nb 25

magnitude  X 

d) xn  1.5ej0.2 ej0.1n  1.5ej0.2 ej0.1n .

Recall that DTFTej0 n   2  0  and therefore we obtain
X  3ej0.2   0.1  3ej0.2   0.1
26 Solutions_Chapter1[1].nb

| X ( ) |

3 3

  0.1 0.1  

 X ( )
0.2
 0.1
 0.1  
 0.2
X magnitude and phase

e) xn  0.5jej0.5n  0.5jej0.5n  0.5ej0.25n  0.5ej0.25n . Therefore the

DTFT becomes
X  ej0.5   0.5  ej0.5   0.5 
  0.25    0.25
Solutions_Chapter1[1].nb 27

| X ( ) |
3
3
   
 
  2  4 4 2  

 X ( )

0.5

  
 0.5
X magnitude and phase

f) xn  20.5jej0.5n  0.5jej0.5n 

2

 20.25ejn  0.25ejn  2  0.25 

 1  ejn

Therefore X  2  2   

 2  ej     ej   
Notice that we split the term 2    ej     ej  ) to preserve the
simmetry of the DTFT.
28 Solutions_Chapter1[1].nb

| X ( ) |

2
 
  

 X ( )



 

X magnitude and phase
g) xn  1n  ejn  0.5ejn  0.5ejn therefore
X        
again we split it to preserve the simmetry of the DTFT.
Solutions_Chapter1[1].nb 29

| X ( ) |

 
  

 X ( )


 
X magnitude and phase

Fourier Transform

à Problem 1.42

You know that

FTrectt  sincF
30 Solutions_Chapter1[1].nb

rectt sincF
2 2
1.5 1.5
1 1
0.5 0.5
t
3 2 1 1 2 3 6 4 2 2 4
0.5 0.5

Apply the properties of the Fourier Transform:

a) FTrect2t   2 
2 sinc 
1 F

rect2t 0.5sincF2
2 2
1.5 1.5
1 1
0.5 0.5
t
3 2 1 1 2 3 6 4 2 2
0.5 0.5

b) FTrect2t  5  FTrect2t  2.5  ej2F 2.5 FTrect2t which

yields
FTrect2t  5  ej5F  2 
2 sinc 
1 F

j2 10t  ej 2 10 t . Since

5  then
c) FTrect5tcos20t  FTrect5t 
2 e
1

FTrect5t  
5 sinc 
1 F

FTrect5tcos20t  
10 sincF  10  
1
10 sincF  10
1
Solutions_Chapter1[1].nb 31

xt X
1.5 0.2
1 0.15
0.5 0.1
t 0.05
0.4 0.2 0.2 0.4
0.5
15 10 5
1 0.05
1.5 0.1

d) FTsinc0.1t. By the duality property, FTsinct  rectF. Since the "rect"

function is symmetric, we can write FTsinct  rectF. Therefore
FTsinc0.1 t  10 rect10F

sinc0.1t 10 rect
1.5
1.25
1
0.75
0.5
0.25
t
20 10
0.25 10 20
0.5 0.4 0.2

e) FTsinc0.1tcos20t  0.1  . Therefore

ej0.1 FTej210t sinc0.1t  e j0.1 FTe j210t sinc0.1t
XF  ej0.1 0.1rect10F  10  ej0.1 0.1rect10F  10

a  , then IFTsinc 

a sinc  a    a  rectat.
Therefore let a  1  10 and then
f) Since FTrectat  
1 F F

IFTsinc10F  0.1rect0.1t
g) IFTrect10F  0.1sinc0.1t for the same reasons as in f);
32 Solutions_Chapter1[1].nb

h) Recall that FT  t  nT0   F0   F  kF0 , where F0  1  T0 . Also recall that
 

 xnT0 t  nT0   xt  t  nT0 .

n k
 
n n

With this in mind we can write

 
 sinc0.2nt  0.1n  sinc2t  t  0.1n, and therefore
n n
 
FT  sinc0.2nt  0.1n  FTsinc2t  FT  t  0.1n
n n
 
 0.5rect0.5F  10  F  10k  5  rect0.5F  10k
k k

An alternative way to arrive at the same inswer is by using the "rep" and "comb" operators. Recall that
FTcombT0 xt  F0 repF0 XF
In our case T0  0.1, F0  1  T0  10, xt  sinc2t,
XF  FTsinc2t  0.5rect0.5F. Therefore
FTcomb0.1sinc2t  10rep10 0.5rect0.5F
i) Again apply duality to obtain
FTrepT0 xt  F0 combF0 XF

Therefore FT  rectt  2n  FTcomb2 rectt  0.5 rep0.5sincF, and
n
therefore
 
FT  rectt  2n  0.5  sinc0.5kF  0.5k
n k

j) By the same arguments as in the previous problem, we can write


FT  rectt  n  FTcomb1 rectt  rep1 sincF
n

and therefore
 
FT  rectt  n   sinckF  k
n k

But now recall that sinck  0 for all k  0. Also it is easy to see that  rectt  n  1
n
for all t. Therefore
FT1  F
since all other terms sinckF  k , for k  0, in the summation, are zero.
Solutions_Chapter1[1].nb 33

à Problem 1.43

Let a signal xt have Fourier Transform as shown.

X (F )
1

1 1 F
Sketch the following:
 
a)  x0.01nt  0.01n  xt  t  0.01n  comb0.01xt. There-
n n
fore its Fourier Transform is given by

FT  x0.01nt  0.01n  100rep100XF
n

shown below

100rep100 X ( F )

100

 200 100 100 200 F (H


b)  xt  0.01n  rep0.01xt. Therefore its Fourier Transform is 100comb100XF
n
shown below
34 Solutions_Chapter1[1].nb

100comb100 X ( F )
100

1 0 1 F (Hz )
0.01

c) FTxtcos100t  
2 XF  50  
1
2 XF  50
1

1
2 X ( F  50)  12 X ( F  50)
1 1
2 2

F (Hz )
 50 50
d) Since cos2 100t    j100t   j200t  2  ej200t then
2
2 e
j100t  
2 e 4 e
1 1 1

FTxtcos2 100t  

2 XF  
1
4 XF  100  
1
4 XF  100
1

shown below.
Solutions_Chapter1[1].nb 35

1
2 X ( F )  14 X ( F  100)  14 X ( F  100)
1
2
1
1 4
4

 100 100 F (H
e) From the propertises, FTx5t   5  shown below:
5 X 
1 F

1
5 X  F5 
1
5

5 5 F (Hz )
f) From the properties again, FTx0.1t  10X10F, as shown:
36 Solutions_Chapter1[1].nb

10 10 X 10 F 

 0 . 1 0. 1 F (Hz )