SS 202B Signals & Systems

Final Exam (Page 1/ 11) Lecturer: Jung-Woo Choi

Solutions 2015 Fall (closed book)

[Score table] Total Score :

Prob 1 2 3 4
5
. a b c a b c a b c d e a b c d e
Total 5 10 10 5 5 10 5 5 5 5 5 5 5 5 5 5 5

Score

Your problems

[25 points] Chapter 4 Continuous-time Fourier transform

1. Figure 1 shows the magnitude and phase of the Fourier transform X ( j) of a signal x(t ) .

Figure 1

(a) [5 pts] Determine the value of this time signal x(t ) at t   / 200 .

The signal is the time shift of the signal with zero phase. From the properties of Fourier transform,
x(t  t0 )  X ( j )e jt0 , the time delay is t0   / 200 . Therefore, x(t0 ) is equivalent to x(0) with
zero time shift(figure shown below). From the properties of Fourier transform, x(0) is the area of X ( j)
divided by 2 . The half-circle area is  (100)2 / 2 , so x(t0 ) = (100)2 / 4  2500 .

(b) [10 pts] Express the signal x2 (t ) of Figure 2 in terms of x(t ) , when the Fourier transform X 2 ( j) is
given as follows:
SS 202B Signals & Systems
Final Exam (Page 2/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

Figure 2

The given signal is 1/10 scaling of the original signal, modulated by 100rad/s to both sides.

X (10 j) 1  1  t 
X1 ( j )   x1 (t ) 
10  10  10  
The scaling: x --- (1)
10

Modulation: the phase has opposite signs for positive and negative frequencies -> sinusoidal amplitude modulation

X 2 ( j)   X1 ( j)  S ( j) 

1 
2 
 
X1 ( j)  2  e 2  (  0 )  e 2  (  0 )  , 0  100
 j j


(continues on the next page)

From the multiplication-convolution property,

x2 (t )  x1 (t )  ( je j0t  je j0t )  x1 (t )  2sin(0t ) ----(2)

1  t 
From (1) and (2), x2 (t )  sin(100t ) x   `
50  10 
 SS 202B Signals & Systems
Final Exam (Page 3/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

(c) [10 pts] The signal x(t ) is fed into a system h(t ) and produces the output y(t ) . When the Fourier
transform of the output signal y(t ) is given as follows, derive the system’s impulse response h(t ) .

Figure 3

The phase for 0 is subtracted by  , which means that Y ( j)  H ( j) X ( j) with H ( j)
being the Signum function in the frequency domain defined by

H ( j)  2U () 1
where U () is the unit step function in the frequency domain.

(note: 1  e j  e j )

The inverse Fourier transform of the Signum function is given by

dH ( j ) dU ( )
2  2 ( )
d d
2 1 1
( H ( j) has no DC component, { ()} 
1
 jth(t )   ))
2  2
j
 h(t ) 
t

[20 points] Chapter 5 & 7 Discrete Fourier transform and Sampling

2. (a) [5 pts] Now the signal x(t ) of Figure 1 is sampled at a rate of s . Determine the Nyquist rate required
to avoid aliasing.

The signal is bandlimited within |  |  100  M , and hence, the Nyquist rate is s  2M  200
SS 202B Signals & Systems
Final Exam (Page 4/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

(b) [5 pts] Suppose that the sampling rate is s  1000 radssec, and the signal is converted to a discrete
sequence x[n] after the sampling. Plot the magnitude of the discrete Fourier transform X (e j ) . Also
indicate the zero-crossing point and peak amplitude of | X (e j ) | .

From the sampling T   ,

1   
X d (e j  )  
T k 
X ( j (  2 k ) / T )  s
2
 X ( j(  2 k )  )
k 
2
s

2 
Frequency scaling s  1000    2 :   100  
10 5
s 1000 50000
Amplitude scaling by : 100 
2 2 

(c) [10 pts] For a signal y[n] satisfying the following relation, plot magnitude of discrete Fourier transform
Y (e j ) .


 sin( 5 (n  5k )) 
y[n]   x[5k ]    (1)
k   5 (n  5k ) 
SS 202B Signals & Systems
Final Exam (Page 5/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

 x[n] for n  5k
Consider a signal x5 [n]  
 0 otherwise

sin( 5 n)
Equation (1) is then the convolution of x5[n] and

5 n


 sin( 5 (n  m))    sin( 5 (n  5k )) 
y[n]   5   (n  m)  k
x [ m]  x[5k ]    --(1)
m  5    5 (n  5k ) 
The Fourier transform of xN [n] is given by,

1 N 1
 X (e  N  ) (textbook Eq. 7.42)
j  k 2 
X N (e j  ) 
N k 0

Magnitude scaling: 1s N , repetition with period 2 / N (solid black curve shown below).

sin(Wn)
From the properties of Fourier transform,  Rect(W ) , where Rect(W ) is a rectangular
n
function of 2W width( W ). Therefore,

 sin( 5 n) 
  5Rect( 5 ) ,

  (red curve)
 5 n 
From (1) and convolution-multiplication property, it can be shown that the result is the same.

 sin( 5 n) 
 y[n]  x5[n]  
j
  X 5 (e )  5Rect( 5 )


 5 n 
1 N 1
 X (e  N  )  5Rect( 5 )  X (e j )
j  k 2 

5 k 0
Therefore, the answer is the same as that of (b).
 SS 202B Signals & Systems
Final Exam (Page 6/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

[25 points] Chapter 6 Time and Frequency Characterization

3. Consider a LTI system with the condition of initial rest. The system satisfies the following differential equation.

d 2 y(t ) dy(t ) dx(t )

2
2  y(t )  10 x(t )  (2)
dt dt dt
(a) [5 pts] Determine the Laplace transform H (s) and the frequency response H ( j)  Y ( j) / X ( j) of
the given system.

( s 2  2s  1)Y ( s)  (10  s) X ( s)
Y ( s) 10  s 10  s
H ( s)   2 
X ( s) s  2s  1 ( s  1) 2
10  j
 H ( j ) 
( j  1)2

(b) [5 pts] Draw the bode plot of H ( j) with respect to  . Show both magnitude and phase responses.

Magnitude response

- Starting point: H (0)  10  +20dB

Phase response

0.1 1 10 100 0.1 1 10 100 0.1 1 10 100

0 0 0

0.1 1 10 100
0

(c) [5 pts] Determine the impulse response h(t ) of this system.

SS 202B Signals & Systems
Final Exam (Page 7/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

LTI, Condition of initial rest  causal system  right-sided

10  j (1  j )  11 1 11
H ( j )    
( j  1) 2
( j  1) 2
j  1 ( j  1)2
 h(t )  et u(t )  11 tet u(t )

(d) [5 pts] Draw the pole-zero plot in the s-plane. Is this system stable?

ROC

Rational H (s) , Causal system  ROC is on the right-side of the rightmost pole.
ROC includes imaginary axis  Stable

(e) [5 pts] An inverse system Hinv (s) of a system H (s) is defined as the system of which transfer function is
1/ H ( s) . Determine whether the inverse of the given system( H (s)1 ) can be causal and stable at the same
time. Justify your answer.

Inverse of H ( s) changes the location of poles and zeros. The causal inverse system should have ROC on the
right-side from s=10, which does not include the imaginary axis  cannot be stable & causal
 SS 202B Signals & Systems
Final Exam (Page 8/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

[25 points] Chapter 10 Z-transform

4. The pole-zero plot of a LTI system with impulse response h[n] is shown in Figure 4.

Figure 4
j
(a) [10 pts] When the zero-frequency response is H (e ) | 0  2 , identify the system response H ( z ) From
H ( z ) , describe the system’s characteristic(ex. Low-pass, high-pass, band-pass, all-pass, low-shelf, high-
shelf … ).

z 1 1 3
H ( z)  G  G 1 1 , H (e0  1)  G 2  G
(z  4)
1
1 4 z 3/ 4 2

3 1
 H ( z )   1 1 (Low-pass: see textbook p. 468)
2 1 4 z

(b) Find the impulse response h[n] of the system. Verify your result at n  0 using initial value theorem.

ROC is outside from the pole  Causal system

n
31
h[n]    u[n]
2 4

3
lim H ( z )  at n  0
z  2

(c) [5 pts] Draw the pole-zero plot of (1)n h[n] in z -domain. Is this system stable?
SS 202B Signals & Systems
Final Exam (Page 9/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

 
H new ( z )   (1) h[n]z
n 
n n
  h[n]( z)
n 
n
 H ( z )

Now the pole is at z=-1s4 : ROC still includes unit circle  stable

(d) [5 pts] Draw the pole-zero plot of the 8n h[n] in z -domain. Is this system stable?

 
H new ( z )   (8) h[n]z
n 
n n
  h[n]( z / 8)
n 
n
 H ( z / 8)

Now the pole is at z=1s4*8=2: ROC excludes the unit circle  unstable

(e) [5 pts] Consider a discrete-time sequence h2 [n] defined as

h[n], n  2m
h2 [n]  
 0 otherwise

(1) Draw the pole locations of h2 [n] and discuss the ROC of this sequence. (2) Derive the difference
equation that can describe this system.
 SS 202B Signals & Systems
Final Exam (Page 10/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)

1 1
h2 [n]  h[n](1  (1)n )  H 2 ( z)   H ( z)  H ( z) 
2 2

3  1 1  3 1 3 z2
 H 2 ( z )    1 1  1 1     
4  1 4 z 1  4 z  2 (1  14 z 1 )(1  14 z 1 ) 2 ( z  14 )( z  14 )

 ROC is the same

3 1 Y ( z)
H 2 ( z)   1 2

2 (1  16 z ) X ( z )

2  y[n]  161 y[n  2]  3x[n]

 2 y[n]  18 y[n  2]  3x[n]

[5 points] Chapter 4 Continuous-time Fourier transform

5. [5 pts] Correlation of signals x(t ) and y(t ) is defined as follows:
 
cxx (t )   x( ) x(t   )d , cxy (t )   x(t   ) y( )d (3)
 

Suppose that an unknown LTI system with an impulse response h(t ) was driven by a input signal x(t ) , and
then its output y(t ) was measured. For a special input signal x(t ) , the impulse response h(t ) can be
directly obtained from the correlation of the output and input signal( cxy (t ) ). That is,

cxy (t )  h(t ) (4)

What is Cxx ( )  cxx (t ) ? (1) First, show that cxy (t )  x(t )  y (t ) . (2) From this result, find out the
characteristic of cxx (t ) to satisfy Eq. (4) and derive the Fourier transform of such cxx (t ) .
SS 202B Signals & Systems
Final Exam (Page 11/ 11) Lecturer: Jung-Woo Choi
Solutions 2015 Fall (closed book)


cxy (t )   x(t   ) y( )d


  x(t  ( )) y( )d


 x(t )  y(t )

Since y(t )  h(t )  x(t ) ,

cxy (t )  x(t )  y(t )

 x(t )  h(t )  x(t )
 x(t )  x(t )  h(t )  h(t )
cxx ( t )

Therefore, cxx (t )   (t ) and cxx (t )  1