Chapter 2: Introduction to Systems

Problem 2.1
(i)

The currents flowing out of node 1 along resistors R1, R2, and capacitor C, are given by

iR1 =

y (t ) v (t )
,
R1

iR 2 =

y (t )
,
R2

iC = C

dy
dt

Applying the Kirchoffs current law to node 1 and summing up all the currents, gives
y (t ) v(t ) y (t )
dy
+
+C
=0
R1
R2
dt

or,

dy 1
1
v(t )
+ + y (t ) =
dt R1 R2
R1

dy R1 + R2
1
+
y (t ) =
v(t ) .
dt
CR1 R2
CR1

or,

(ii)
(a) Linear: For v1(t) applied as the input, the output y1(t) is given by
dy1 R1 + R2
dy
R + R2
1
y1 (t ) =
v1 (t ) v1 (t ) = CR1 1 + 1
+
y1 (t )
dt
CR1 R2
CR1
dt
R2
For v2(t) applied as the input, the output y2(t) is given by
dy 2 R1 + R2
dy
R + R2
1
+
y 2 (t ) =
v 2 (t ) v 2 (t ) = CR1 2 + 1
y 2 (t ) .
dt
CR1 R2
CR1
dt
R2

For v3(t) = v1(t) + v2(t) applied as the input, the output y3(t) is given by
dy 3 R1 + R2
1
(v1 (t ) + v 2 (t ) ) .
+
y 3 (t ) =
dt
CR1 R2
CR1
Substituting the values of v1(t) and v2(t) from the earlier equations, we get

dy

dy

dy 3 R1 + R2
R + R2
R + R2
+
y 3 (t ) = 1 + 1
y1 (t ) + 2 + 1
y 2 (t ) .
dt
CR1 R2
CR1 R2
CR1 R2
dt

dt

Rearranging the terms on the right hand side of the equation, we get
dy 3 R1 + R2
d (y1 + y 2 ) R1 + R2
(y1 (t ) + y 2 (t ) ) ,
+
y 3 (t ) =
+
dt
CR1 R2
dt
CR1 R2

which implies that

y 3 (t ) = y1 (t ) + y 2 (t ) .

The system is, therefore, linear.

(b) Time-invariance: For v(t t0) applied as the input, the output y1(t) is given by

(S2.1)

42

Chapter 2
dy1 R1 + R2
1
+
y1 (t ) =
v(t t 0 ) .
dt
CR1 R2
CR1

Substituting = t t0 (which implies that dt = d), we get

dy1 ( + t 0 ) R1 + R2
1
+
y1 ( + t 0 ) =
v() .
d
CR1 R2
CR1
Comparing with Eq. (S2.1), we get

y () = y1 ( + t0 ) or,

y1 () = y ( t0 ) ,

proving that the system is time-invariant.

(c) Memoryless: Express Eq. (S2.1) as
y (t ) =

1
CR1

v()d

R1 + R2
CR1 R2

y()d .

The output y(t) at t = t0 is given by

y (t ) t =t =
0

1
CR1

t0

v()d

R1 + R2
CR1 R2

t0

y()d

From the first integral on the right hand side of the equation, it is clear that all previous values of the input
v(t), for t t0, are needed to calculate the output y(t) at t = t0. The system has, therefore, a memory
and is not memoryless.
(d) Causal: From the previous result, we deduce that the system is causal since only the past values of the
input v(t), for t t0, are needed to calculate the output y(t) at t = t0.
(e) Invertible: The system is invertible as v(t) can be determined from the following relationship
v(t ) = CR1

dy R1 + R2
+
y (t ) .
dt
R2

(f) Stable: The system is BIBO stable since a bounded input will always produce a bounded output as
shown below.
Using Theorem 3.1, the output of the system defined by Eq. (S2.1) is given by
y (t ) = e p

p (t ) = exp

where

1 v ( ) d
CR
1

R1 + R2
CR1 R2

t .

From the above solution, it is clear that the output y(t) is bounded in the input v(t) is bounded.
Problem 2.2

(a)

The currents flowing in resistor R, inductor L, and capacitor C, are given by

t

iR =

y (t )
1
dy
, iL = y ( ) d , iC = C .
R
L
dt

Solutions

43

Applying the Kirchoffs current law, we obtain

t

y (t ) 1
dy
+ y ( )d + C
= i (t ) ,
R
L
dt

or, by differentiating,

d 2 y 1 dy 1
di
+
+ y (t ) =
2
dt
R dt L
dt

d2y
1 dy 1
1 di
+
+
y (t ) =
.
2
dt
RC dt LC
C dt

or,

(S2.2)

(ii)
(a) Linear: For i1(t) applied as the input, the output y1(t) is given by

d 2 y1
dt 2

1 dy1
1
1 di1
+
y1 (t ) =
RC dt
LC
C dt

di1
d 2 y1 1 dy1 1
=C
+
+ y1 (t )
dt
R dt
L
dt 2

For i2(t) applied as the input, the output y2(t) is given by

d 2 y2
dt 2

1 dy 2
1
1 di 2
+
y 2 (t ) =
RC dt
LC
C dt

di 2
d 2 y 2 1 dy 2 1
=C
+
+ y 2 (t ) .
dt
R dt
L
dt 2

For i3(t) = i1(t) + i2(t) applied as the input, the output y3(t) is given by
d 2 y3
dt
or,

d 2 y3
dt

1 dy 3
1
1 d (i1 + i 2 )
+
,
y 3 (t ) =
RC dt
LC
C
dt

1 dy 3
1
di1 di 2
+
+
y 3 (t ) =
RC dt
LC
C dt C dt

Substituting the values of di1/dt and di2/dt from the earlier equations, we get
d 2 y3
dt 2

d 2 y

d 2 y

1 dy 3
1
1 dy1
1
1 dy 2
1
+
+
+
y 3 (t ) = 21 +
y1 (t ) + 2 2 +
y 2 (t ) .
RC dt
LC
RC dt
LC
RC dt
LC
dt

dt

Rearranging the terms on the right hand side of the equation, we get
d 2 y3
dt 2

d 2 (y1 + y 2 )
1 dy 3
1
1 d (y1 + y 2 )
1
(y1 + y 2 ) .
+
y 3 (t ) =
+
+
2
RC
dt
LC
RC dt
LC
dt
y 3 (t ) = y1 (t ) + y 2 (t ) .

Comparing with (S2.2) implies that

The system is therefore linear.

(b) Time-invariance: For i1(t t0) applied as the input, the output y1(t) is given by
d 2 y1
dt

1 dy1
1
1 di1 (t t 0 )
y1 (t ) =
+
.
RC dt
LC
C
dt

Substituting = t t0 (which implies that dt = d), we obtain

d 2 y1 ( + t 0 )
d

1 dy1 ( + t 0 )
1
1 di1 ( )
+
y1 ( + t 0 ) =
.
RC
d
LC
C d

44

Chapter 2

Comparing with Eq. (S2.2.2), we obtain

y () = y1 ( + t0 ) or,

y1 () = y ( t0 ) ,

proving that the system is time-invariant.

(c) Memoryless: Express Eq. (S2.2) as
y (t ) =

1
C

i ( )d

1
LC

y ( )dd

1
RC

y()d .

The output y(t) at t = t0 is given by

y (t ) t =t =
0

1
C

t0

i ( ) d

1
LC

t0

y ( )dd

1
RC

t0

y ( ) d

From the first integral on the right hand side of the equation, it is clear that all previous values of the input
i(t), for t t0, are needed to calculate the output y(t) at t = t0. The system has, therefore, memory and
is not memoryless.
(d) Causal: From the previous result, we deduce that the system is causal since only the past values of the
input i(t), for t t0, are needed to calculate the output y(t) at t = t0.
(e) Invertible: The system is invertible as i(t) can be determined from the following relationship
d 2 y 1 dy 1
di
=C 2 +
+ y (t ) .
dt
R dt L
dt
(f) Stable: The system is BIBO stable since a bounded input will always produce a bounded output.

Problem 2.3

(i)

For input x1(t), x2(t), and x1(t) + x2(t), the respective outputs are given by

x1 ( t ) c1 x1 (t ) + c2 x12 (t ) = y1 ( t )
x2 ( t ) c1 x2 (t ) + c2 x22 (t ) = y2 ( t )
x1 ( t ) + x2 ( t ) c1 x1 ( t ) + x2 ( t ) + c2 x1 ( t ) + x2 ( t ) = y(t )
2

Because y(t ) y1 (t ) + y2 (t ) , the demodulator is a non-linear device.

(ii)
(a) Time Invariance: For input x1(t) and x2(t) = x1(t T), the respective outputs are given by

x1 ( t ) c1 x1 (t ) + c2 x12 (t ) = y1 ( t )
x2 ( t ) = x1 ( t T ) c1 x2 (t ) + c2 x22 (t ) = c1 x1 ( t T ) + c2 x12 ( t T ) = y2 ( t )
Since y1 ( t T ) = y2 ( t ) , the system is time invariant.
(b) Memory: The output v2(t) depends only on the current value of the input v1(t). Therefore, the system
is memoryless. All memoryless systems are causal. Therefore, the system is also causal.
(c) Invertible: The input to the system, v1(t) can be calculated using the following equation

Solutions

45

c1 c12 + 4c2 v2 (t )
.
v1 (t ) =
2c2

For a given value of v2(t), two possible values of v1(t) exist As the input v1(t) cannot be uniquely
determined from the output v2(t), the system is NOT invertible.
(d) Stable: Assuming that |v1(t)| M < , the output v2(t) is bounded by
v2 ( t ) = c1v1 (t ) + c2 v12 (t ) < c1 v1 (t ) + c2 v12 (t ) < c1M + c2 M 2 < .

Therefore, the system is stable.

Problem 2.4

(i)

The modulated signal is given by

s ( t ) = A [1 + km(t )] cos ( 2 f c t ) = 5 [1 + 2ksin(200 t ) ] cos ( 2000000 t ) .

1 + 2ksin(200 t ) 0 implies
(ii)

2ksin(200 t ) 1 , or, k 0.5 .

Any value of k in the range (0 k 0.5) can be used. We use k = 0.5 in the rest of the problem. The
AM signal is given by

s (t ) = A [1 + km(t ) ] cos ( 2 f c t ) = 5 [1 + 0.8sin(200 t ) ] cos ( 2000000 t ) .

(iii) Expanding the above equation, we get

s (t ) = 5cos ( 2000000 t ) + 4sin(200 t ) cos ( 2000000 t )

= 5cos ( 2000000 t ) + 2sin(2000200 t ) + 2sin(1999800 t )
It is observed that the AM signal has three frequency components at 1,000,000 Hz, 1,000,200 Hz, and
999,900 Hz. The frequency component at 1,000,000 Hz represents the carrier signal, while the remaining
two frequency components at 1,000,200 Hz and 999,900 Hz represent the sinusoidal tone. Therefore, the
frequency of the sinusoidal tone is shifted to a higher frequency range.

Problem 2.5

(i)

Dividing both sides by M, Eq. (2.16) can be expressed as

d2y
1
r dy k
y (t ) =
x(t ) .
+
+
M dt M
M
dt

Comparing with the given expression, the coefficients are given by

k
M
n
r
=
Q M
n2 =

and
(ii)

k
M
kM
Q=
.
r

n =

k
, the natural frequency n can be increased either by: (a) increasing the value of
M
the spring constant k, or (b) by decreasing the value of the mass M.

Since n =

46

Chapter 2

kM
, the value of Q can be reduced either by: (a) reducing the value of the spring
r
constant k, (b) reducing the value of mass M, or (c) by increasing the value of r.

Since Q =

(iii)

d2y
dy
+
+ y (t ) = x(t ) ,
2
dt
dt

with =

r
k
1
, = , =
M
M
M

(S2.5.1)

(a) Linear: For x1(t) applied as the input, the output y1(t) is given by

d 2 y1
dy
+ 1 + y1 (t ) = x1 (t )
2
dt
dt

x1 (t ) = 1

d 2 y1 dy1
+
+ y1 (t )
dt 2 dt

x2 (t ) = 1

d 2 y2 dy2
+
+ y2 (t ) .
dt 2 dt

For x2(t) applied as the input, the output y2(t) is given by

d 2 y2
dy
+ 2 + y2 (t ) = x2 (t )
2
dt
dt

For x3(t) = x1(t) + x2(t) applied as the input, the output y3(t) is given by

d 2 y3
dy
+ 3 + y3 (t ) = x3 (t ) = ( x1 (t ) + x2 (t ) ) .
2
dt
dt
Substituting the values of x1 (t ), x2 (t ) from the earlier equations, we obtain

d 2 y1

d 2 y2

d 2 y3
dy3
dy1
dy

y
(
t
)

y
(
t
)

+
+
=
+
+
+
+ 2 + y2 (t )
3
1

2
2
2
dt
dt
dt
dt
dt

dt

d 2 ( y1 + y2 )
d ( y1 + y2 )
+
+ ( y1 (t ) + y2 (t ) )
2
dt
dt

Comparing the left-hand and right-hand sides, we obtain

y 3 (t ) = y1 (t ) + y 2 (t ) .
The system is therefore linear.
(b) Time-invariance: For x1(t) applied as the input, the output y1(t) is given by

d 2 y1
dy
+ 1 + y1 (t ) = x1 (t )
2
dt
dt

(S2.5.2)

For x2(t) =x1(t t0) applied as the input, the output y2(t) is given by

d 2 y2
dy
+ 2 + y2 (t ) = x2 (t ) = x1 (t t0 ) .
2
dt
dt
Substituting = t + t0 (which implies that dt = d) in Eq. (S2.5.2), we obtain

d 2 y1 ( t0 )
dy ( t0 )
+ 1
+ y1 ( t0 ) = x1 ( t0 )
2
d
d
Substituting t = , we obtain

d 2 y1 (t t0 )
dy (t t0 )
+ 1
+ y1 (t t0 ) = x1 (t t0 )
2
dt
dt

(S2.5.3)

Solutions

47

Comparing with Eq. (S2.5.2), we obtain

y2 (t ) = y1 (t t0 ) ,
proving that the system is time-invariant.
(c) Memoryless: Express Eq. (S2.5.1) as

y (t ) =

x( )d d

x( )d d

y( )d .

The output y(t) at t = t0 is given by

y (t ) t =t =
0

t0

x( )d d

t0

x( ) d d

t0

y( )d

From the first integral on the right hand side of the equation, it is clear that all previous values of the input
x(t), for t t0, are needed to calculate the output y(t) at t = t0. The system has, therefore, memory and
is not memoryless.
(d) Causal: From the previous result, we deduce that the system is causal since only the past values of the
input x(t), for t t0, are needed to calculate the output y(t) at t = t0.
(e) Invertible: The system is invertible as i(t) can be determined from the following relationship

x(t ) = 1

d 2 y dy
+
+ y (t ) .
dt 2 dt

(f) Stable: The system is BIBO stable since a bounded input will always produce a bounded output.

Problem 2.6

(i)

Substituting, t = kt in

d2y
dy
+5
+ 6 y (t ) = 0 , yields
dt
dt

d2y
dt

+5
t = kt

dy
dt

t = kt

+ 6 y (t ) t = kt = 0 .

Substituting the values of the first and second derivative from the backward finite difference
scheme, we get

y (k t ) 2 y ((k 1)t ) + y ((k 2)t )

( t )

+5

y (k t ) y ((k 1)t )
+ 6 y (k t ) = 0 ,
t

or, y ( k t ) 2 y (( k 1) t ) + y (( k 2) t ) + 5 [ y ( k t ) y (( k 1)t ) ] t + 6 y ( k t )(t ) 2 = 0 .

Substituting y(kt) = y[k], y((k 1)t) = y[k 1], and y((k 2)t) = y[k 2], the above equation
reduces to

y[k ] 2 y[k 1] + y[k 2] + 5 { y[k ] y[ k 1]} t + 6 y[k ](t ) 2 = 0 ,

which simplifies to

48

Chapter 2

(1 + 5t + 6(t ) )y[k ] + ( 2 5t )y[k 1] + y[k 2] = 0 .

2

(ii)

Substituting the value of the first CT initial condition y(0) = 3 in

y[k] = y(kt)
for k = 0, we obtain

y[0] = y(0) = 3.

Similarly, substituting the value of the second CT initial condition y (0) = 7 in

dy
dt

t = kt

y (0) =

for k = 0, we get

y (kt ) y ((k 1)t )

t

y[0] y[1]
= 7 .
t

Simplifying the above, we obtain

y[1] = y[0] ty (0) = 3 + 7t .

(iii) By substituting t = 0.02s in the difference equation,

(1 + 5t + 6(t ) )y[k ] + ( 2 5t )y[k 1] + y[k 2] = 0 ,

2

(1.1024) y[k ] + ( 2.1)y[k 1] + y[k 2] = 0 ,

y[k ] = 1.90493 y[k 1] 0.90711y[k 2]

we obtain
or,

with y[0] = 3 and y[1] = 3 + 7(0.02) = 3.14.

The Matlab code used to compute and plot the DT solution is given in Program 2.6. For
comparison, we also plot the CT solution. The two plots are included in Fig. S2.6.

CT Solution of Problem 2.6

3

yct

1
0

0.1

0.2

0.3

10

15

0.4

0.5
0.6
0.7
t
DT Solution of Problem 2.6

0.8

0.9

40

45

ydt

2
1
0

20
25
30
k: Recall t = k t

35

Fig. S2.6: Plots obtained from the CT (top) and DT (bottom) solution
of the 2nd order differential equation in Problem 2.6.

Solutions

49

Program 2.6: MATLAB code for Problem 2.6.

% MATLAB code for Problem 2.6
% plot CT result
N = 50;
% No of points to be plotted
k = 0:N-1;
% time index
dt = 0.02;
% discretization step
t = k*dt;
% time instants
yct = exp(-3*t)+2*exp(-2*t);
% CT answer
subplot(2,1,1)
plot(t,yct), grid on
xlabel('t')
% Label of X-axis
ylabel('yct')
% Label of Y-axis
title('CT Solution of Problem 2.6');
axis ([0 max(t) 0 max(yct)])
% compute finite difference approximation
ydt(1) = 3+7*dt;
% this is actually y[-1]
ydt(2)= 3;
% this is actually y[0]
for i = 3:N+1
ydt(i) = 1/(1+5*dt+6*dt^2)*((2+5*dt)*ydt(i-1)- ydt(i-2));
end
subplot(2,1,2)
stem(k,ydt(2:N+1),'fill'), grid on
xlabel('k: Recall t = k\times\Deltat');% Label of X-axis
ylabel('ydt')
% Label of Y-axis
title('DT Solution of Problem 2.6');
axis ([0 max(k) 0 max(ydt)])
Problem 2.7

Starting from the initial value x(0) = 0V, The output of the delta modulator receiver is computed from the
recursive expression
x (kT ) = x ((k 1)T ) + bk
where bk is 1 if bit 1 is received, while bk is 1 if bit 0 is received. The computed values are shown in
Table S2.7 with the waveform plotted in Fig. S2.7.

Table S2.7: Decoded output of the delta modulator in Problem 2.7

k

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.0

bk
x (t )

1
1

2
1

3
1

4
1

5
1

6
0

7
1

8
1

9
1

10 11 12 13 14 15 16 17 18 19 20
1

0 0.1 0.2 0.3 0.4 0.5 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2

50

Chapter 2

Reconstructed waveform for x(t)

1
0.9
0.8
0.7

x(t)

0.6
0.5
0.4
0.3
0.2
0.1
0

0.2

0.4

0.6

0.8

1
time (t)

1.2

1.4

1.6

1.8

Fig. S2.7:Waveform reconstructed from the delta-modulated bit stream in Problem 2.7.
Problem 2.8

Eq. (2.27) represents a lowpass filter that averages the four neighboring pixels to compute the value of the
reference pixel located at (m, n). Specifically,
y[m, n] =

1
(x[m, n] + x[m, n 1] + x[m 1, n] + x[m 1, n 1]) .
4

The system is invertible as the input x[m, n] can be constructed in a recursive way from the current value
of the input and the previously reconstructed values of the input pixels based on the following
relationship

x[m, n] = 4 y[m, n] x[m, n 1] x[m 1, n] + x[m 1, n 1] .

Since both the original filter and the inverse system use the values of the past pixels to compute the
reference pixel, both systems are causal in nature.

Problem 2.9

(i)

y (t ) = x (t 2)

(a) Linearity: Since

x1 (t ) x1 (t 2) = y1 (t )
x2 (t ) x2 (t 2) = y2 (t )
x1 (t ) + x2 (t ) x1 (t 2) + x2 (t 2) = y1 (t ) + y2 (t )
therefore, the system is a linear system.
(b) Time Invariance: For inputs x1(t) and x2(t) = x1(t T), the outputs are given by

x1 (t ) x1 (t 2) = y1 (t )

x2 (t ) = x1 (t T ) x2 (t 2) = x1 (t T 2) = y2 ( t )

and y1 (t T ) = x1 (t T 2) = y2 (t ) , the system is time invariant.

Solutions

51

(c) Stability: Assume that the input is bounded |x(t)| M. Then, the output

y (t ) = x(t 2) M
is also bounded proving that the system is BIBO stable.
(d) Causality: Since the output depends only on the past input and does not depend on the future
values of the input, therefore, the system is causal.
(ii)

y (t ) = x(2t 5)
(a) Linearity: Since

x1 (t ) x1 (2t 5) = y1 ( t )
x2 (t ) x2 (2t 5) = y2 ( t )
x1 (t ) + x2 (t ) x1 (2t 5) + x2 (2t 5) = y1 (t ) + y2 (t )
therefore, the system is a linear system.
(b) Time Invariance: For inputs x1(t) and x2(t) = x1(t T), the outputs are given by

x1 (t ) x1 (2t 5) = y1 (t )

x2 (t ) = x1 (t T ) x2 (2t 5) = y2 ( t )

which implies that

x2 (t ) = x1 (t T ) y2 (t ) = x2 (2t 5) x

( t ) = x1 (t T )

= x1 ( 2t 5 T ) .

On the other hand,

y1 (t T ) = x1 (2t 5) t =t T = x1 (2(t T ) 5) = x1 (2t 2T 5) ,

and y1 (t T ) y2 ( t ) . Therefore, the system is NOT time invariant.
(c) Stability: Assume that the input is bounded |x(t)| M. Then, the output

y (t ) = x( 2t 5) M
is also bounded proving that the system is BIBO stable.
(d) Causality: For (t > 5), the output depends on the future values of the input, therefore, the system
is NOT causal.
(iii)

y (t ) = x(2t ) 5
(a) Linearity: Since

x1 (t ) x1 (2t ) 5 = y1 ( t )
x2 (t ) x2 (2t ) 5 = y2 ( t )
x1 (t ) + x2 (t ) x1 (2t ) + x2 (2t ) 5 y1 (t ) + y2 (t )
because y1 (t ) + y2 (t ) = x1 (2t ) + x2 (2t ) 5( + ) . Therefore, the system is NOT linear.
(b) Time Invariance: For inputs x1(t) and x2(t) = x1(t T), the outputs are given by

52

Chapter 2
x1 (t ) x1 (2t ) 5 = y1 (t )

x2 (t ) = x1 (t T ) x2 (2t ) 5 = y2 ( t )

which implies that

x2 (t ) = x1 (t T ) y2 (t ) = x2 (2t ) 5 x

(t ) = x1 (t T )

= x1 (2t T ) 5 .

On the other hand,

y1 (t T ) = x1 (2t ) t =t T 5 = x1 (2(t T )) 5 = x1 (2t 2T ) 5 .

Clearly, y1 (t T ) y2 (t ) , therefore, the system is NOT time invariant.
(c) Stability: Assume that the input is bounded |x(t)| M. Then, the output

y (t ) = x(2t ) 5 x(2t ) + 5 M + 5
is also bounded proving that the system is BIBO stable.
(d) Causality: For (t > 0), the system requires future values of the input to calculate the current
value of the input. Therefore, the system is NOT causal.
(iv)

y (t ) = tx (t + 10)

(a) Linearity: Since

x1 ( t ) tx1 ( t + 10 ) = y1 ( t )
x2 ( t ) tx2 ( t + 10 ) = y2 ( t )
x1 ( t ) + x2 ( t ) tx1 ( t + 10 ) + tx2 ( t + 10 ) = y1 (t ) + y2 (t )
therefore, the system is a linear system.
(b) Time Invariance: For inputs x1(t) and x2(t) = x1(t T), the outputs are given by

x1 (t ) tx1 (t + 10) = y1 (t )
x2 (t ) = x1 (t T ) tx2 (t + 10) = tx1 (t T + 10) = y2 (t )
We also note that

y1 ( t T ) = (t T ) x1 ( t T + 10 ) y2 ( t ) ,

therefore, the system is NOT time invariant.

(c) Stability: Assume that the input is bounded |x(t)| M. Then, the output

y (t ) = tx(t + 10) = t x(t + 10) M t

is unbounded as t . Therefore, the system is NOT BIBO stable.
(d) Causality: Since the output depends on the future values of the input, and therefore the system is
NOT causal.
(v)

2
y (t ) = 2u ( x ( t ) ) =
0

(a) Linearity: Since

x (t ) 0
x (t ) < 0

Solutions

53

x1 (t ) 2u ( x1 ( t ) ) = y1 (t )
x2 (t ) 2u ( x2 ( t ) ) = y2 (t )
x1 (t ) + x2 (t ) 2u ( x1 ( t ) + x2 ( t ) ) = y (t )

and y1 (t ) + y2 (t ) = 2 u x1 ( t ) + 2 u x2 ( t ) y(t ) . Therefore, the system is NOT linear.

Also, we note that

y2 (t ) y1 (t ) = 2u ( x2 ( t ) ) 2u ( x1 ( t ) ) [ x2 (t ) x1 (t )] .
Therefore, the system is NOT an incrementally linear system either.
(b) Time Invariance: For inputs x1(t) and x2(t) = x1(t T), the outputs are given by

x1 (t ) 2u ( x1 ( t ) ) = y1 (t )
x2 (t ) = x1 (t T ) 2u ( x2 ( t ) ) = 2u ( x1 ( t T ) ) = y2 (t )
We note that y1 (t T ) = 2u ( x1 ( t T ) ) = y2 (t ) , therefore, the system is time invariant.
(c) Stability: Since |y(t)| 2, therefore, the system is BIBO stable.
(d) Causality: The output at any time instant does not depend on future value of the input. The
system is, therefore, causal.
(vi)

0
y (t ) =
x(t ) x(t 5)

t<0
t0

= [ x(t ) x(t 5) ] u (t )

(a) Linearity: Since

x1 (t ) [ x1 (t ) x1 (t 5)] u (t ) = y1 (t )
x2 (t ) [ x2 (t ) x2 (t 5)] u (t ) = y2 (t )
x1 (t ) + x2 (t ) { x1 (t ) + x2 (t )} { x1 (t 5) + x2 (t 5)} u (t ) = y(t )

= [ x1 (t ) x1 (t 5)] u (t ) + [ x1 (t ) x1 (t 5)] u (t )
= y1 (t ) + y2 (t ),

the system is linear.

(b) Time Invariance: For inputs x1(t) and x2(t) = x1(t T), the outputs are given by

x1 (t ) [ x1 (t ) x1 (t 5)] u (t ) = y1 (t )
x2 (t ) = x1 (t T ) [ x2 (t ) x2 (t 5)] u (t ) = y2 (t )
We note that y1 (t T ) y2 ( t ) since

y1 (t T ) = y1 (t ) t =t T = [ x1 (t ) x1 (t 5)] u (t ) t =t T = [ x1 (t T ) x1 (t T 5)] u (t T )
and

y2 (t ) = [ x2 (t ) x2 (t 5)] u (t ) == [ x1 (t T ) x1 (t T 5)] u (t ) .

Therefore, the system is NOT time invariant.

54

Chapter 2
(c) Stablity: Assume that the input is bounded |x(t)| M. Then, the output

y (t ) = x(t ) x(t 5) x(t ) + x(t 5) 2 M

is also bounded. Therefore, the system is BIBO stable.
(d) Causality: The output does not depend on the future values of the input, therefore, the system is
causal.
(vii) y (t ) = 7 x 2 (t ) + 5 x(t ) + 3
(a) Linearity: For x1(t) applied as the input, the output y1(t) is given by

y1 (t ) = 7 x12 (t ) + 5 x1 (t ) + 3
For x2(t) applied as the input, the output y2(t) is given by

y2 (t ) = 7 x22 (t ) + 5 x2 (t ) + 3 .
For x3(t) = x1(t) + x2(t) applied as the input, the output y3(t) is given by

or,

y3 (t ) = 7(x1 (t ) + x2 (t ) )2 + 5(x1 (t ) + x2 (t ) ) + 3 ,

) (

y3 (t ) = 7 x12 (t ) + 5 x1 (t ) + 3 + 7 x12 (t ) + 5 x2 (t ) + 3 + 14 x1 (t ) x2 (t ) + 3(1 )

y1 ( t )

y 2 (t )

The above result implies that

y3 (t ) y1 (t ) + y2 (t ) ,
And hence the system is not linear.
(b) Time Invariance: For x1(t) and x2(t) applied as the inputs, the outputs are given by

x1 (t ) 7 x12 (t ) + 5 x1 (t ) + 3 = y1 (t )
x2 (t ) = x1 (t T ) 7 x22 (t ) + 5 x2 (t ) + 3 = y2 (t ).
Substituting x2 (t ) = x1 (t T ) we obtain,

y2 (t ) = 7 x12 (t T ) + 5 x2 (t T ) + 3 .
We also note that

y1 (t T ) = 7 x12 (t T ) + 5 x2 (t T ) + 3 ,

implying that y1 (t T ) = y2 (t ) . The system is, therefore, time invariant.

(c) Stablity: Assume that the input is bounded |x(t)| M. Then, the output

y (t ) = 7 x 2 (t ) + 5 x(t ) + 3 7 x(t ) x(t ) + 5 x(t ) + 3 7 M 2 + 5M + 3

is also bounded. Therefore, the system is BIBO stable.
(d) Causality: The output y(t) at t = t0 requires only one value of the input y(t) at (t = t0). Therefore,
the system is causal.
(viii) y(t) = sgn(x(t))
(a) Linearity: For x1(t) applied as the input, the output y1(t) = sgn(x1(t)).
For x2(t) applied as the input, the output y2(t) = sgn(x2(t)).

Solutions

55

For x3(t) = x1(t) + x2(t) applied as the input, the output y3(t) is given by

y3 (t ) = sgn (x1 (t ) + x2 (t ) ) sgn (x1 (t ) ) + sgn (x2 (t ) ) .

The above result implies that

y3 (t ) y1 (t ) + y2 (t ) ,
And hence the system is not linear.
(b) Time Invariance: For x1(t) and x2 (t ) = x1 (t T ) applied as the inputs, the outputs are given by

x1 (t ) sgn (x1 (t ) ) = y1 (t )
x2 (t ) = x1 (t T ) sgn (x2 (t ) ) = y2 (t )
Substituting x2 (t ) = x1 (t T ) we obtain,

y2 (t ) = sgn (x1 (t T ) ) .
y1 (t T ) = sgn (x1 (t T ) ) ,

We also note that

implying that y1 (t T ) = y2 (t ) . The system is, therefore, time invariant.

(c) Stablity: The system is stable as the output is always bounded between the values of 1 and 1.
(d) Causality: The output y(t) at (t = t0) requires only one value of the input x(t) at (t = t0), therefore,
the system is causal.
t0

(ix)

y (t ) =

x()d + 2 x(t )

t0

(a) Linearity: For x1(t) and x2(t) applied as the inputs, the outputs are given by
t0

x1 ()d + 2 x1 (t ) ,

y1 (t ) =

t0

t0

y2 (t ) =

x2 ()d + 2 x2 (t ) .

t0

For x3(t) = x1(t) + x2(t) applied as the input, the output y3(t) is given by

y3 (t ) =

t0

( x ( ) + x ( ) ) d + 2 ( x (t ) + x (t ) )
1

t0

t0

t0

= x1 ( )d + 2 x1 (t ) + x2 ( )d + 2 x2 (t ) ,
t0

t0

y1 ( t )

y2 ( t )

= y1 (t ) + y2 (t )
Therefore, the system is linear.
(b) Time Invariance: For x1(t) and x2 (t ) = x1 (t T ) applied as the inputs, the outputs are given by

56

Chapter 2

t0

x1 (t ) y1 (t ) =

x1 ()d + 2 x1 (t )

t0

t0

x2 (t ) = x1 (t T ) y2 (t ) =

x2 ()d + 2 x2 (t ) ,

t0

Substituting x2 (t ) = x1 (t T ) we obtain,
t0

y2 (t ) =

x ( T )d + 2 x (t T ) .
1

t0

t0 T

By substituting = T, we get y2 (t ) =

x1 ( )d + 2 x1 (t T ) .

t0 T

t0

y1 (t T ) =

We also note that

x1 ()d + 2 x1 (t T ) ,

t0

implying that y1 (t T ) y2 (t ) . The system is, therefore, NOT time invariant.

(c) Stablity: Assume that the input is bounded |x(t)| M. Then, the output

y (t ) =

t0

t0

t0

t0

x()d + 2 x(t ) x() d + 2 x(t ) 2Mt0 + 2M

is also bounded. Therefore, the system is BIBO stable.

(d) Causality: To solve the integral, the output y(t) always requires the values of the input x(t)
within the range (t0 t t0) no matter when y(t) (even for t < t0) is being determined. Therefore,
the system is NOT causal.
t0

(x)

y (t ) =

dx

x()d + dt

(a) Linearity: For x1(t) and x2(t) applied as the inputs, the outputs are given by

y1 (t ) =

t0

x ( ) d +
1

y2 (t ) =

t0

x ( ) d +
2

dx1
,
dt
dx2
.
dt

For x3(t) = x1(t) + x2(t) applied as the input, the output y3(t) is given by

Solutions

y3 (t ) =

d ( x1 (t ) + x2 (t ) )
dt

t0

( x ( ) + x ( ) ) d +
1

57

t0
t0
d ( x1 (t ) )
d ( x2 (t ) )
= x1 ( )d +
+ x2 ( )d +
,
dt
dt


y1 ( t )

y2 ( t )

= y1 (t ) + y2 (t )
Therefore, the system is linear.
(b) Time Invariance: For x1(t) and x2 (t ) = x1 (t T ) applied as the inputs, the outputs are given by

x1 (t ) y1 (t ) =

t0

x ( ) d +
1

x2 (t ) = x1 (t T ) y2 (t ) =

dx1
dt

t0

x ( ) d +
2

dx2
.
dt

Substituting x2 (t ) = x1 (t T ) we obtain,

y2 (t ) =

t0

x ( T ) d +
1

By substituting = T, we get y2 (t ) =

dx1 (t T )
.
dt

t0 T

x1 ( )d +

We also note that

y1 (t T ) =

t0

x ( ) d +
1

dx1 (t T )
.
dt

dx1 (t T )
,
dt

implying that y1 (t T ) y2 (t ) . The system is, therefore, NOT time invariant.

(c) Stablity: Assume that the input is bounded |x(t)| M. Then, the output
t0

y (t ) =

dx(t )
x ( ) d + 2

dt

t0

x ( ) d + 2

dx(t )
dt

is unbounded because of the integral which integrates x(t) from ( t t0). Therefore, the system
is NOT stable.
(d) Causality: To solve the integral, the output y(t) always requires only the values of the input x(t)
within the range ( t t0) no matter when y(t) (even for t < t0) is being determined. Therefore,
the system is NOT causal.
(xi)

d4y
d3y
d2y
dy
d 2x
+
3
+
5
+
3
+
y
(
t
)
=
+ 2 x(t ) + 1
dt
dt 4
dt 3
dt 2
dt 2
(a) Linearity: For x1(t) applied as the input, the output y1(t) is given by

58

Chapter 2

d 4 y1
d 3 y1
d 2 y1
dy1
d 2 x1
y
t
+
3
+
5
+
3
+
(
)
=
+ 2 x1 (t ) + 1 .
1
dt
dt 4
dt 3
dt 2
dt 2

(S2.9.1)

For x2(t) applied as the input, the output y2(t) is given by

d 4 y2
d 3 y2
d 2 y2
dy2
d 2 x2
+
3
+
5
+
3
+
y
(
t
)
=
+ 2 x2 (t ) + 1 .
2
dt
dt 4
dt 3
dt 2
dt 2

(S2.9.2)

For y3(t) = x1(t) + x2(t) applied as the input, the output y3(t) is given by
d 4 y3
dt 4

+3

d 3 y3

+5

dt 3

d 2 y3
dt 2

+3

dy3
d 2 (x1 (t ) + x2 (t ) )
+ y2 (t ) =
+ 2(y1 (t ) + y2 (t ) ) + 1 ,
dt
dt 2

2
3
4

2
or, d y3 + 3 d y3 + 5 d y3 + 3 dy3 + y2 (t ) = d x1 + 2 x1 (t ) + 1 + d x2 + 2 x2 (t ) + 1 + [1 ] .
2
2
2
3
4

dt

dt

dt

dt

dt

dt

Term I

Term II

Substituting the values of the derivative terms (Terms I and II) from Eqs. (S2.9.1) ans (2.9.2), we
obtain

d 4 y3
dt 4

+3

d 3 y3
dt 3

+5

d 2 y3
dt 2

+3

d4y

dy3
d3y
d2y
dy
+ y2 (t ) = 41 + 3 31 + 5 21 + 3 1 + y1 (t )
dt
dt
dt
dt
dt

Term I

d4y

d3y
d2y
dy
+ 42 + 3 32 + 5 22 + 3 2 + y2 (t ) + [1 ]
dt
dt
dt
dt

Term II

y3 (t ) y1 (t ) + y2 (t ) .

which implies that

The system is, therefore, NOT linear. Note that the dc term of (+ 1) on the right hand side of the
differential equation contributes to the nonlinearity of the system
(b) Time-invariance: The system is time-invariant. The proof is similar to Problem 2.1.
(b) Time-invariance: For x1(t) and x2 (t ) = x1 (t T ) applied as the inputs, the outputs are given by
d 4 y1
d 3 y1
d 2 y1
dy1
d 2 x1
y
t
+
3
+
5
+
3
+
(
)
=
+ 2 x1 (t ) + 1
1
dt
dt 4
dt 3
dt 2
dt 2

(S2.9.3)

d 4 y2
d 3 y2
d 2 y2
dy2
d 2 x2
+
3
+
5
+
3
+
y
(
t
)
=
+ 2 x2 (t ) + 1 .
2
dt
dt 4
dt 3
dt 2
dt 2
Substituting x2 (t ) = x1 (t T ) we obtain,

d 4 y2
d 3 y2
d 2 y2
dy2
d 2 x1 (t T )
+
3
+
5
+
3
+
y
(
t
)
=
+ 2 x1 (t T ) + 1 .
2
dt 4
dt 3
dt 2
dt
dt 2

(S2.9.4)

Substituting = t + T (which implies that dt = d) in Eq. (S2.9.3), we obtain

d 4 y1 ( T )
d 3 y1 ( T )
d 2 y1 ( T )
dy1 ( T )
d 2 x1 ( T )
+
3
+
5
+
3
+
y
(

T
)
=
+ 2 x1 ( T ) + 1 .
1
d 4
d 3
d 2
d
d 2

Solutions

Or,

59

d 4 y1 (t T )
d 3 y1 (t T )
d 2 y1 (t T )
dy1 (t T )
d 2 x1 (t T )
+
3
+
5
+
3
+
y
(
t

T
)
=
+ 2 x1 (t T ) + 1 .
1
dt 4
dt 3
dt 2
dt
dt 2

Comparing with Eq. (S2.9.4), we obtain

y2 (t ) = y1 (t T ) ,
proving that the system is time-invariant.
(c) Stablity: The system is BIBO stable since a bounded input will always produce a bounded output.
(d) Causality: Express Eq. (S2.2) as follows:
t

y (t ) = 3 y ( ) d 5

+5

y ( )d d 3

y ( )d d d

x ( )d d + 2

y( )d d d d

y ( )d d d +

d d d d

The output y(t) at t = t0 is given by

t0

t0

y (t ) t =t = 3 y ( )d 5
0

t0

t0

y( )d d 3 y( )d d d y( )d d d d

t0

t0

+5

t0

x( )d d + 2 y( )d d d + d d d d

The system is causal since only the past values of the input x(t), for t t0, are needed to calculate the
output y(t) at t = t0.

Problem 2.10

(i)

y[k ] = ax[k ] + b

(a)

Linearity: Since
x1[ k ] ax1[ k ] + b = y1[ k ]
x2 [ k ] ax2 [ k ] + b = y1[k ]

( x1[k ] + x2 [k ]) a ( x1[k ] + x2 [k ]) + b y1[k ] + y2 [k ]

the system is NOT a linear system.
(b)

Time Invariance: For inputs x1[k] and x2[k] = x1[k K], the outputs are given by
x1[ k ] ax1[ k ] + b = y1[ k ]
x2 [ k ] = x1[ k K ] ax2 [k + 2] + b = y2 [ k ]

We note that

y2 [ k ] = ax2 [ k ] + b = ax1[k K + 2] + b = y1[k K ]

which implies that the system is time invariant.

(c)

Stability: Assume that the input is bounded |x[k]| M. Then, the output
y[k ] = ax[k ] + b ax[k ] + b aM + b

is also bounded. Therefore, the system is BIBO stable.

60

Chapter 2

(d)

Causality: Since the output requires only the current value of input, the system is causal.

(ii)

y[k ] = 5 x[3k 2]

(a)

Linearity: Since
x1[ k ] y1[k ] = 5 x1[3k 2]
x2 [k ] y 2 [k ] = 5 x2 [3k 2]
(x1[k ] + x2 [k ]) 5(x1[3k 2] + x2 [3k 2]) = y1[k ] + y2 [k ]

the system is a linear system.

(b)

Time Invariance: For inputs x1[k] and x2[k] = x1[k K], the outputs are given by
x1[k ] 5 x1[3k 2] = y1[k ]
x2 [k ] = x1[k K ] 5 x2 [3k 2] x

[ k ] = x1 [ k K ]

= 5 x1[3k 2 K ] = y 2 [k ]

y1[k K ] = 5 x1[3(k K ) 2] y2 [k ]

We also note that

Therefore, the system is NOT time invariant.

(c)

Stability: Assume that the input is bounded |x[k]| M. Then, the output
y[k ] = 5 x[3k 2] = 5 x[3k 2] 5M

is also bounded. Therefore, the system is BIBO stable.

(d) Causality: When k 2, the output requires the future value of the input. Therefore, the system is
NOT causal.
(iii)

y[k ] = 2 x[ k ]

(a)

Linearity: Since
x1[ k ] 2 x1 [ k ] = y1[ k ]
x2 [ k ] 2 x2 [ k ] = y2 [ k ]

( x1[k ] + x2 [k ]) 2 x [ k ]+ x [ k ] = 2 x [ k ]2 x [ k ] = ( 2 x [ k ] )
1

( 2 ) = ( y [ k ]) ( y [k ])
x2 [ k ]

y1[ k ] + y2 [ k ]

the system is NOT a linear system.

(b) Time Invariance: For inputs x1[k] and x2[k] = x1[k K], the outputs are given by
x1[ k ]

2 x1 [ k ] = y1[k ]

x2 [ k ] = x1[k K ] 2 x2 [ k ] = y2 [ k ]

Note that

y2 [ k ] = 2 x2 [ k ] = 2 x1 [ k K ]

and

y1[ k K ] = 2 x1 [ k ]

k =k K

= 2 x1 [ k K ] .

Since y2 [ k ] = y1[ k K ] , the system is time invariant.

(c)

Stability: Assume that the input is bounded |x[k]| M. Then, the output

Solutions

y[k ] = 2 x[ k ] 2

x[ k ]

61

2M

is also bounded. Therefore, the system is BIBO stable.

(d) Causality: Since the output requires only the current value of the input, therefore, the system is causal.
(iv)

y[k ] =

x[m]

m =

(a)

Linearity: Since
k

x1[ k ]

m =

x2 [ k ]

m =

x1[m] = y1[k ]
x2 [ m] = y2 [ k ]
k

m =

m =

( x1[k ] + x2 [k ]) ( x1[m] + x2 [m]) =

x1[ m] +

m =

x2 [ m]

= y1[ k ] + y2 [ k ]

the system is a linear system.

(b)

Time Invariance: For inputs x1[k] and x2[k] = x1[k K], the outputs are given by
x1[ k ]

m =

x1[m] = y1[ k ]

x2 [ k ] = x1[k K ]

Note that

y2 [ k ] =

m =

x2 [ m] =

m =

m =

x2 [ m] = y2 [ k ]

x1[ m K ] =

k K

m =

x1[ m]= y1[ k K ]

And hence the system is time invariant.

(c)

Stability: Assume that the input is bounded |x[k]| M. Then, the output
y[k ] =

x[m]

m =

x[m]

m =

may become unbounded as an infinite number of absolute values of x[k] are being added. Therefore, the
system is NOT BIBO stable.
(d)
Causality: Since the output does not depend on the future values of the input, therefore, the system
is causal.
(v)

y[k ] =

k +2

x[m] 2 | x[k ] |

m=k 2

(a)

Linearity: Since

62

Chapter 2
k +2

x1[m] 2 x1[k ] = y1[k ]

x1[k ]
x2 [ k ]

(x1[k ] + x2 [k ])

m=k 2
k +2

x2 [ m ] 2 x2 [ k ] = y 2 [ k ]

m=k 2
k +2

(x1[m] + x2[m]) 2 (x1[k ] + x2[k ])

m=k 2
k +2

k +2

x1[m] 2 x1[k ] +
x2 [ m ] 2 x2 [ k ]

m=k 2

m=k 2

the system is NOT a linear system. Note that the absolute term on the right hand side of the equation
makes the system nonlinear.
(b)

Time Invariance: For inputs x1[k] and x2[k] = x1[k K], the outputs are given by
x1[k ]

k +2

x1[m] 2 x1[k ]
m=k 2
k +2

x2 [k ] = x1[ k K ]

= y1[k ]

x 2 [ m ] 2 x2 [ k ] = y 2 [ k ]

m=k 2

Note that
k +2

y2 [k ] =
x 2 [ m ] 2 x2 [ k ]
m = k 2
x [ k ]= x [ k K ]
2
1

k +2

k K +2

m=k 2

p=k K 2

x1[m K ] 2 x1[k K ] = x1[ p] 2 x1[k K ]

which is the same as y1[k K]. Therefore, the system is time invariant.
(c)

Stability: Assume that the input is bounded |x[k]| M. Then, the output
y[ k ] =

k +2

x[m] 2 x[k ]

m=k 2

k +2

x[m] + 2 x[k ] 7M

m=k 2

is also bounded. Therefore, the system is BIBO stable.

(d)

Causality: Since the output requires future values of the input, therefore, the system is NOT causal.

(vi)

y[k ] + 5 y[ k 1] + 9 y[ k 2] + 5 y[k 3] + y[k 4] = 2 x[k ] + 4 x[k 1] + 2 x[k 2]

(a)

Linearity: For x1[k] applied as the input, the output y1[k] is given by
y1[k ] + 5 y1[ k 1] + 9 y1[k 2] + 5 y1[k 3] + y1[k 4] = 2 x1[k ] + 4 x1[k 1] + 2 x1[ k 2] .

For x2[k] applied as the input, the output y2[k] is given by

y2 [k ] + 5 y2 [k 1] + 9 y2 [k 2] + 5 y2 [k 3] + y2 [k 4] = 2 x2 [k ] + 4 x2 [k 1] + 2 x2 [k 2] .
For x3[k] = x1[k] + x2[k] applied as the input, the output y3[k] is given by

Solutions

63

y3[k ] + 5 y3[k 1] + 9 y3[k 2] + 5 y3[k 3] + y3[k 4]

= 2(x1[k ] + x2 [k ]) + 4(x1[k 1] + x2 [ k 1]) + 2(x1[k 2] + x2 [k 2]),
which can be expressed as
y3[k ] + 5 y3[k 1] + 9 y3[k 2] + 5 y3[k 3] + y3[k 4]
= (2 x1[k ] + 4 x1[k 1] + 2 x1[k 2]) + (2 x1[k ] + 4 x1[k 1] + 2 x1[k 2]).

Substituting the value of the input terms, we get

y3[k ] + 5 y3[k 1] + 9 y3[k 2] + 5 y3[k 3] + y3[k 4]
= ( y1[k ] + 5 y1[k 1] + 9 y1[ k 2] + 5 y1[k 3] + y1[k 4])
+ ( y2 [k ] + 5 y2 [k 1] + 9 y2 [k 2] + 5 y2 [k 3] + y2 [k 4])
or,
y3[k ] + 5 y3[k 1] + 9 y3[k 2] + 5 y3[k 3] + y3[k 4]
= (y1[k ] + y2 [k ]) + 5(y1[ k 1] + y2 [ k 1]) + 9(y1[k 2] + y2 [k 2])
+ 5(y1[k 3] + y2 [ k 3]) + (y1[k 4] + y2 [ k 4]) ,
which implies that y3[k] = y1[k] + y2[k]. Therefore, the system is linear.
(b)

Time Invariance: For inputs x1[k] and x2[k] = x1[k K], the outputs are given by
x1[k] y1[k ] + 5 y1[k 1] + 9 y1[k 2] + 5 y1[k 3] + y1[k 4] = 2 x1[k ] + 4 x1[k 1] + 2 x1[k 2]
x2[k] = x1[k K]
y2 [k ] + 5 y2 [k 1] + 9 y2 [ k 2] + 5 y2 [ k 3] + y2 [k 4] = 2 x2 [k ] + 4 x2 [k 1] + 2 x2 [k 2]
From the above equations, we can prove that y2[k] = y1[k K] implying that the system is time
invariant.

(c)

Stability: Assume that the input is bounded |x[k]| M. Taking the absolute value of both sides of the
difference equation, we get
y1[k ] + 5 y1[k 1] + 9 y1[k 2] + 5 y1[k 3] + y1[k 4] = 2 x1[k ] + 4 x1[ k 1] + 2 x1[k 2]

implying that

y1[k ] + 5 y1[k 1] + 9 y1[ k 2] + 5 y1[k 3] + y1[k 4] 8M .

Since sum of several output samples are bounded, y[k] itself must be bounded. Therefore, the system is
BIBO stable.
(d)

Causality: Since the output can be computed iteratively from

y[k ] = (2 x[ k ] + 4 x[k 1] + 2 x[k 2]) (5 y[k 1] + 9 y[k 2] + 5 y[k 3] + y[k 4]) ,

which requires only the previous values of the input, therefore, the system is causal.
(vii) y[k ] = 0.5 x[6k 2] + 0.5 x[6k + 2]
(a)

Linearity: For x1[k] applied as the input, the output y1[k] is given by

y1[k ] = 0.5 x1[6k 2] + 0.5 x1[6k + 2] .

For x2[k] applied as the input, the output y2[k] is given by
y2 [k ] = 0.5 x2 [6k 2] + 0.5 x2 [6k + 2] .

64

Chapter 2

For x3[k] = x1[k] + x2[k] applied as the input, the output y3[k] is given by
y3[k ] = 0.5(x1[6k 2] + x2 [6k 2]) + 0.5(x1[6k + 2] + x2 [6k + 2])
which can be expressed as
y3[k ] = (0.5 x1[6k 2] + 0.5 x1[6k + 2]) + (0.5 x2 [6k 2] + 0.5 x2 [6k + 2]) ,
y1 [ k ]

y2 [k ]

implying that y3[k] = y1[k] + y2[k]. Therefore, the system is linear.

(b)

Time Invariance: For inputs x1[k] and x2[k] = x1[k K], the outputs are given by
x1[k ] y1[k ] = 0.5 x1[6k 2] + 0.5 x1[6k + 2]
x2 [k ] = x1[k K ] 0.5 x2 [6k 2] + 0.5 x2 [6k + 2] = y2 [k ]

The second equation implies that

y2 [k ] = [0.5 x2 [6k 2] + 0.5 x2 [6k + 2]]x [ k ]= x [ k K ]
2
1
= [0.5 x1[6k 2 K ] + 0.5 x2 [6k + 2 K ]]

We also note that

y1[ k K ] = 0.5 x1[6k 6 K 2] + 0.5 x1[6k 6 K + 2]

Since y2[k] y1[k K], the system is NOT time invariant.

(c)

Stability: Assuming that the input is bounded |x[k]| M, the output

y1[k ] = 0.5 x1[6k 2] + 0.5 x1[6k + 2] 0.5 x1[6k 2] + 0.5 x1[6k 2] M

is also bounded. Therefore, the system is BIBO stable.

(d)

Causality: Since the output requires future values of the input, the system is NOT causal.

Problem 2.11

(i)

Using the linearity property

5 x(t ) 5 y (t ) .
The output waveform is drawn in Fig. S2.11(a).

(ii)

Using the time invariance property

x(t 1) y (t 1)

and

x(t + 1) y (t + 1) .

Using the above result and the linearity property

[0.5 x(t 1) + 0.5 x(t + 1)] [0.5 y(t 1) + 0.5 y(t + 1)]
The output waveform is drawn in Fig. S2.11(b).

Solutions

5x(t) 5y(t)
5

(a)
0.5x(t1) + 0.5x(t+1) 0.5y(t1) + 0.5y(t+1)
1
0.5
2

(b)
x(t+1) x(t1) y(t+1) y(t1)
1
1
2

2
1

(c)
dx (t )
dt

+ 3 x(t )

(t)

dy (t )
dt

4
2

+ 3 y (t )

(d)
Fig. S2.11: Output waveforms for CT transformations in Problem 2.11.
(iii) Using the time invariance property, we obtain
x(t 1) y (t 1)
and

x(t + 1) y (t + 1) .

Using the above result and the linearity property, we obtain

[ x(t + 1) x(t 1)] [ y (t + 1) + y (t 1)] .

The output waveform is drawn in Fig. S2.11(c).
(iv)

Since differentiation is a linear operation,

dx(t )
dy (t )

dt
dt
Using the above result and the linearity property

65

66

Chapter 2

dx(t )

dy (t )

dt + 3x(t ) dt + 3 y (t ) .

The output waveform is drawn in Fig. S2.11(d).

Problem 2.12

In terms of the DT impulse function [k], the output can be represented by

y[k ] = 4[k + 2] 2[k + 1] + 4[k ] + 2[k 1] + 4[k 2].
(i)

Using the linearity and time invariance properties, we obtain

4 x[k 1] 4 y[k 1] = 16[k + 1] 8[k ] + 16[k 1] + 8[k 2] + 16[k 3].
The output waveform is drawn in Fig. S2.12(a).

(ii)

Using the time invariance property, we obtain

x[k 2] y[k 2] = 4[k ] 2[k 1] + 4[ k 2] + 2[k 3] + 4[k 4]
and

x[k + 2] y[ k + 2] = 4[k + 4] 2[k + 3] + 4[k + 2] + 2[k + 1] + 4[k ] .

Using the above result and the linearity property, we obtain

[0.5 x[k 2] + 0.5 x[k + 2]] [0.5 y[k 2] + 0.5 y[k + 2]]
or, [0.5 x[k 2] + 0.5 x[k + 2]] 2 [k + 4] [k + 3] + 2 [k + 2] + [k + 1]
+4 [k ] [k 1] + 2 [k 2] + [k 3] + 2 [k 4].
The output waveform is drawn in Fig. S2.12(b).
(iii) Using the time invariance property, we obtain
x[k + 1] y[k + 1] = 4[k + 3] 2[k + 2] + 4[k + 1] + 2[k ] + 4[k 1]
and

x[k 1] y[ k 1] = 4[k + 1] 2[k ] + 4[k 1] + 2[k 2] + 4[ k 3] .

Using the above result and the linearity property, we obtain

[x[k 1] 2 x[k ] + x[k + 1]] [ y[k 1] 2 y[k ] + y[k + 1]]

or, [x[ k 1] 2 x[k ] + x[ k + 1]] 4[k + 3] 10[ k + 2] + 12[k + 1] 8[k ] + 4[ k 1] 6[ k 2] + 4[k 3].
The output waveform is drawn in Fig. S2.12(c).
(iv)

Using the time invariance property, we obtain

x[k ] y[ k ] = 4[ k + 2] 2[ k + 1] + 4[ k ] + 2[k 1] + 4[ k 2] .
Using the property [k + m] = [k m] , the above equation reduces to
x[k ] y[k ] = 4[k 2] 2[k 1] + 4[k ] + 2[k + 1] + 4[k + 2]
The output waveform is drawn in Fig. S2.12(d).

Solutions

0.5y[k 2] + 0.5y[k + 2]

4y[k 1]

16

16

67

16
8

2
1

2
1

(a)

(b)
y[k]

y[k 1] 2y[k ] + y[k + 1]

10 12
4

4
2

(c)

(d)

Fig. S2.12: Output waveforms for DT transformations in Problem 2.12.

Problem 2.13

(i)

The system is invertible with the inverse system given by

x(t ) = 13 y (t 2) .

(ii)

To calculate the inverse system, we differentiate the integral to get

dy (t )
= x(t 10) .
dt
The inverse system is obtained through two steps. Step 1 compute z(t) = dy/dt, while Step 2
computes x(t) from the relationship x(t) = z(t + 10).

(iii) The system y(t) = |x(t)| is not invertible as x(t) = a produces the same output y(t) = a.
(iv)

If y(t) is differentiable then x(t) can always be calculated uniquely from the expression
x(t ) =

dy (t )
+ y (t )
dt

and the system is invertible. However, if y(t) is not differentiable (for example, it contains a
discontinuity), then x(t) cannot always be calculated uniquely and the system is not invertible.
(v)

System represented by y(t) = cos(2x(t)) is not invertible as different values of x(t) = ( + 2m),
where m is an integer, produce the same output.

Problem 2.14

(i)

Not invertible as y[1] is always 0 irrespective of the value of x[k].

(ii)

The input-output relationship can be expressed as:

68

Chapter 2
y[0] = x[2]
y[ 1] = y[1] = x[2] + x[3]
y[ 2] = y[2] = x[2] + x[3] + x[4]
............

It is observed that the output y[k ], <k < depends on the x[k ], k 2 . From the output values, the
value of x[k ], k 2 can be uniquely calculated. However, the input value x[k ], k < 2 cannot be
calculated from y[k ] . Therefore the system is not invertible.
(iii) The system is not invertible as
x[k ] k = 0, 2, 4,...
y[k ] =
elsewhere
0
All odd values of x[k] are lost and can not be recovered from y[k].
(iv)

System is invertible with the inverse system given by

x[k + 2] = y[k ] (2 x[k + 1] 6 x[k ] + 2 x[k 1] + x[k 2]) ,

Or, x[k ] = y[k 2] 2 x[k 1] + 6 x[k 2] 2 x[k 3] x[k 4] .
(v)

System is invertible with the inverse system given by

x[k ] = y[k ] + 2 y[k 1] + y[k 2] .

Problem 2.15

The proof can be made as follows.

x (t ) y (t )

[ time invariance property ]

or, x (t + t ) y (t + t )

x (t + t ) x (t )
y ( t + t ) y ( t )

[linearity property ]
t
t
x ( t + t ) x ( t )
y ( t + t ) y ( t )
or, lim
lim
t 0

0
t
t
dx
dy

or,
dt
dt
or,

Problem 2.16

The periodic signal xp(t) can be expressed in terms of x(t) as follows

x p (t ) =

x(t 2m),

m = integers .

m =

Applying the linearity property, the output for the periodic signal xp(t) is given by
x p (t ) =

which is shown in Fig. S2.16.

m =

m =

x (t 2m )

y (t 2 m ) = y p ( t )

Solutions

69

yp(t)
1

Fig. S2.16: Output yp(t) for the periodic signal xp(t) in Problem 2.16.
Problem 2.17

The impulse response of the system is defined in the problem as follows:

(t ) h(t ) or, h(t ) = y (t ) x ( t )= ( t ) .

(i)

Substituting x(t) = (t), we obtain

h(t ) = (t + 2) 2(t ) + 2(t 2) .

(ii)

Substituting x(t) = (t), we obtain

h (t ) =

t + to

( 4)d =

t to

t + to 4

1 t to 4 < 0 < t + to 4
otherwise
0

( )d =

t to 4

1 4 to < t < 4 + to
=
otherwise.
0

(iii) Substituting x(t) = (t), we obtain

t

h (t ) =

e 2( t ) ( 4)d =

( 4)d = 0

Noting that

e 2( t 4) ( 4)d = e 2( t 4) ( 4)d

t>4
= u (t 4) , the impulse response is obtained as follows:
otherwise

h(t ) = e 2( t 4)u(t 4) .
(iv)

Substituting x(t) = (t), we obtain

h (t ) =

f (T ) (t )d =

f (T t ) (t )d = f (T t ) (t )d

=1

= f (T t )
Problem 2.18

The output h[k] of a DT LTI system to an unit impulse function [k] is shown in Fig. P2.18. Find the
output for the following set of inputs.
(i)

x[k ] = [k + 1] + [ k ] + [k 1]

(ii)

x[k ] =

[k 4m]

m =

70

Chapter 2

(iii) x[k] = u[k]

h[k]
1

Fig. P2.18: Output h[k] for input x[k] = [k] in Problem 2.18.
Solution:

In terms of the DT impulse function [k], the impulse response h[k] is expressed as
[k ] h[k ] = [k + 1] 2[k ] + [ k 1].
(i)

Using the linearity and time invariance property

[k ] h[k ] = [k + 1] 2[k ] + [k 1]
[k + 1] h[k + 1] = [k + 2] 2[k + 1] + [k ]
and [k 1] h[k 1] = [ k ] 2[k 1] + [ k 2]
Using the linearity property
[k + 1] + [k ] + [k 1] h[k + 1] + h[k ] + h[k 1]
or, [k + 1] + [k ] + [k 1] [k + 2] [ k + 1] [k 1] + [k 2].
(ii)

Using the linearity and time invariance property

m =

m =

[k 4m] h[k 4m] = y[k ]

which will be a periodic signal (with period K = 4). One period of the output signal y[k ] is given
below.

1 k = 1
2 k = 0

y[ k ] =
k =1
1
0 k = 2.
(iii) Recall that
u[k ] =

[k m] y[k ] =

m =0

h[k m] ,

m =0

which results in the output

y[k ] = h[k ] + h[k 1] + h[k 2] + h[k 3] +
Substituting different values of k, we get

Solutions
(k = 2):

y[2] = h[2] + h[3] + h[4] + h[5] + = 0.

(k = 1):

y[1] = h[1] + h[2] + h[3] + h[4] + = 1.

(k = 0):

y[0] = h[0] + h[1] + h[2] + h[3] + = 1.

(k = 1):

y[1] = h[1] + h[0] + h[1] + h[2] + = 0.

By expanding for other values of k, it can be shown that y[k] = 0 for |k| 2. In other words,

y[k ] = [k + 1] [k ] .
Problem 2.19

(i)

Substituting x[k] = [k], we obtain

y[k ] = [ k ] 2[k 1] + [k 2] .

(ii)

Substituting x[k] = [k 1] + [k + 1], we get

y[k ] = ([k 1] + [k + 1]) 2([k 2] + [k ]) + ([k 3] + [k 1])
= [k 3] 2[k 2] + 2[ k 1] 2[k ] + [k + 1].
x[k ] = 3[k + 3] + 2[k + 2] + [ k + 1] + [k 1] + 2[k 2] + 3[k 3].

(iii) Express

Substituting the above value of x[k] in the difference equation, we get

y[k ] = ( 3 [k + 3] + 2 [k + 2] + [k + 1] +
2 (

[k 1] + 2 [k 2] + 3 [k 3])
[k 2] + 2 [k 3] + 3 [k 4])

3 [k + 2] + 2 [k + 1] + [k ] +

+(

[k 3] + 2 [k 4] + 3 [k 5])

3 [k + 1] + 2 [k ] + [k 1] +

y[k ] = 3[k + 3] 4[k + 2] + 2[k 1] 4[k 4] + 3[k 5].

or,
Problem 2.20

(i)

Linearity: For x1[k] applied as the input, the output y1[k] is given by
y1[k ] =

1 4
x1[k m] .
5 m=0

For x2[k] applied as the input, the output y2[k] is given by

y2 [k ] =

1 4
x2 [ k m ]
5 m=0

For x3[k] = x1[k] + x2[k] applied as the input, the output y3[k] is given by
y3 [ k ] =

4
4
1 4
(x1[k m] + x2 [k m]) = 1 x1[k m] + 1 x2 [k m]
5 m=0
5 m=0
5 m =0

which implies that y3[k] = y1[k] + y2[k]. Therefore, the system is linear.
Time Invariance: For inputs x1[k] and x2[k] = x1[k K], the outputs are given by
x1[k ] y1[k ] =

1 4
x1[k m]
5 m=0

71

72

Chapter 2

x2 [ k ] = x1[k K ] y2 [k ] =

and

1 4
1 4
x2 [ k m ] =
x1[ k K m] .
5 m =0
5 m=0

From the above equations, it is clear that y2[k] = y1[k K] implying that the system is time
invariant.
(ii)

The impulse response of the system is given by

h[k ] =

1 4
[k m] = 0.2 ( [k ] + [k 1] + [k 2] + [k 3] + [k 4])
5 m =0

0.2
=
0

k = 0,1, 2,3, 4
otherwise

(iii) By expressing
u[k ] =

[k m] y[k ] =

m =0

h[k m] ,

m =0

the output y[k] to the unit step function is given by

y[k ] = h[k ] + h[k 1] + h[k 2] + h[k 3] + h[k 4] + h[k 5] +

Substituting different values of k, we get

(iv)

(k 1):

y[k] = 0.

(k = 0):

y[0] = h[0] + h[1] + h[2] + h[3] + h[4] + h[5] = 0.2.

(k = 1):

y[1] = h[1] + h[0] + h[1] + h[2] + h[3] + h[4] + = 0.4.

(k = 2):

y[2] = h[2] + h[1] + h[0] + h[1] + h[2] + h[3] + = 0.6.

(k = 3):

y[3] = h[3] + h[2] + h[1] + h[0] + h[1] + h[2] + = 0.8.

(k 4):

y[k] = 1

Using the linearity and time invariance property

[k ] = ( u[k ] u[k 1]) h[k ] = ( y[k ] y[k 1])

which leads to

0 k 1 0 k 0
0.2 k = 0 0.2 k = 1

k 1
0
0.4 k = 1 0.4 k = 2

= 0.2 0 k 4
h[k ] =
=
=
0.6
k
2
0.6
k
3

0
k 5.
0.8 k = 3 0.8 k = 4

1.0 k 4 1.0 k 5
y[k ]

y [ k 1]

Note that the above value of the impulse response is the same as the value obtained in part (ii).
Problem 2.21

Solutions

73

(i)

The linearity and time invariance properties can be proved directly from the definitions. Instead, we
use the property of linear, constant coefficient finite difference equations, which always represent
linear and time invariant systems.

(ii)

Series Configuration: Denoting the output of system S1 by w[k], we obtain

w[k ] = x[ k ] 2 x[k 1] + x[ k 2] .
For system S2, w[k] is the input and y[k] is the output. The output in given by
y[k ] = w[k ] + w[k 1] 2 w[k 2] .
Substituting the value of w[k] from the earlier equation, we obtain
y[k ] = (x[k ] 2 x[k 1] + x[k 2]) + (x[k 1] 2 x[k 2] + x[k 3]) 2(x[k 2] 2 x[k 3] + x[k 4])

which reduces to
y[k ] = x[k ] x[ k 1] 3x[ k 2] + 5 x[k 3] 2 x[ k 4] .
(iii) Parallel Configuration: Denoting the output of system S1 by w1[k], we obtain

w1[k ] = x[k ] 2 x[k 1] + x[k 2] .

For system S2, x[k] is the input and w2[k] is the output. The output is given by
w2 [k ] = x[ k ] + x[k 1] 2 x[k 2] .
The overall output of the parallel configuration is given by

y[k ] = w1 [k ] + w2 [k ] = ( x[k ] 2 x[k 1] + x[k 2]) + (x[k ] + x[k 1] 2 x[k 2]) ,

which reduces to
y[k ] = 2 x[k ] x[ k 1] x[k 2] .
(iv)

Since both series and parallel configurations are represented by linear, constant coefficient finite
difference equations, both systems are linear and time invariant.

74

Chapter 2

Additional work:
Supplement to Problem 2.4 How do we demodulate the AM signal?

The modulated signal can be demodulated as follows:

Multiplying the modulated signal s (t ) = A[1 + km(t )] cos(2f c t ) with cos(2fct), we get
s (t ) = A[1 + km(t )] cos 2 (2f c t ) =

1
2

A[1 + km(t )]

Low Frequency Component

1
2

A[1 + km(t )] cos(4f c t ) .

High Frequency Component

By using a low pass filter, the high frequency component can be filtered out. The low frequency
component can be used be extract the information bearing signal m(t).

Supplement to Problem 2.9 Determine if the systems are memoryless and invertible. If invertible, find
the inverse system.

(i)

y (t ) = x (t 2)

Also, note that the system requires past memory and is not memoryless. Further, the system is invertible
as the input can be reconstructed using the relationship x(t ) = y (t + 2) .
(ii)

y (t ) = x(2t 5)
Since all noncausal systems must have memory, the system is NOT memoryless. Furthermore, the
system is invertible with the inverse system given by

y (t ) = x(2t 5) x(t ) = y ( t +25 ) = y (0.5t + 2.5) .

(iii)

y (t ) = x(2t ) 5
Since all noncausal systems must have memory, the system is NOT memoryless. Furthermore the
system is invertible with the inverse system given by

y (t ) = x(2t ) 5 x(2t ) = y (t ) + 5 x(t ) = y (0.5t ) + 5 .

(iv)

y (t ) = tx (t + 10)

Since all noncausal systems must have memory, therefore, the system is NOT memoryless.
Futhermore, the system is NOT invertible as its inverse

x (t ) =

y ( t 10)
t 10

is not defined for t = 10. Therefore, the system is NOT invertible.

(v)

2
y (t ) = 2u ( x ( t ) ) =
0

x (t ) 0
x (t ) < 0

Also note that the system is memoryless since the output at any time instant does not depend on the
past or future values of the input. The system is not invertible since it is not possible to calculate the
input x(t) uniquely from y(t). The output is always 0 for x(t) < 0. Likewise, the output is always 2
for x(t) > 0.
(vi)

0
y (t ) =
x(t ) x(t 5)

t<0
t0

= [ x(t ) x(t 5) ] u (t )

Solutions

75

The system is NOT memoryless since it requires past values of the input to compute the current
value of the output. For (t < 0), the output is always 0. Hence, it is not possible to calculate the input
x(t) uniquely from y(t) for (t < 0). Therefore, the system is NOT invertible.
(vii) y (t ) = 7 x 2 (t ) + 5 x(t ) + 3
The system is also memoryless since only the current value of the input is required to calculate the
output. The system is NOT invertible as the inverse system given by
x(t ) =

5 25 28(3 y (t )
14

produces two possible values for inverting each value of y(t).

(viii) y(t) = sgn(x(t))
The system is also memoryless since only the current value of the input is required to calculate the
output. The system is NOT invertible because it is not possible to determine x(t) from y(t). All
positive values of x(t) produce the same output of +1, while all negative values of x(t) produce the
same output of 1.
t0

(ix)

x()d + 2 x(t )

y (t ) =

t0

The system is NOT memoryless as it is not causal. The system is also invertible.
t0

(x)

dx

x()d + dt

y (t ) =

The system is NOT memoryless as it is not causal.

The system is invertible.
(xi)

d4y
d3y
d2y
dy
d 2x
+
3
+
5
+
3
+
y
(
t
)
=
+ 2 x(t ) + 1
dt
dt 4
dt 3
dt 2
dt 2
The output y(t) at t = t0 is given by
t0

t0

y (t ) t =t = 3 y ( )d 5
0

t0

+5

t0

y ( ) d d 3

y ( )d d d

t0

x( )d d + 2

y ( )d d d +

t0

y( )d d d d

t0

d d d d

The system has memory since the past values of the input x(t), for t t0, are needed to
calculate the output y(t) at t = t0.
Invertiblity: The system is invertible.

Supplement to Problem 2.10 Determine if the systems are memoryless and invertible.

(i)

y[k ] = ax[k ] + b
Since the output requires only the current value of input, therefore, the system is both causal and
memoryless.

76

Chapter 2
Further, the system is invertible with the inverse system given by
x[k ] =

(ii)

1
a

( y[k ] + b ) .

y[k ] = 5 x[3k 2]
The system is NOT memoryless as it is not causal. Further, the system is invertible with the inverse
system given by
x[k ] =

(iii)

1
5

[k +3 2 ] .

y[k ] = 2 x[ k ]
The system is memoryless as well as causal since the output does not depend on the past or future
values of the input. Furthermore, the system is invertible with the inverse system given by

x[k ] = log 2 (x[k ]) .

(iv)

y[k ] =

x[m]

m =

The system is NOT memoryless as the output requires past values of the input. Furthermore, the
system is invertible with the inverse system given by
x[k ] = y[k ] y[k 1] .
(v)

y[k ] =

k +2

x[m] 2 | x[k ] |

m=k 2

The system is NOT memoryless as it is not causal. Further, the system is NOT invertible because of
the |x[k]| term.
(vi)

y[k ] + 5 y[ k 1] + 9 y[ k 2] + 5 y[k 3] + y[k 4] = 2 x[k ] + 4 x[k 1] + 2 x[k 2]

The system is NOT memoryless as it requires some past values of the input. Furthermore, the
system is invertible.

(vii) y[k ] = 0.5 x[6k 2] + 0.5 x[6k + 2]

The system is NOT memoryless as it is not causal. Furthermore, the system is invertible.