LTI Systems

2.6 Interconnection of LTI Systems

2.6.1 Parallel Connection of LTI Systems Figure 2.18 (p. 128)
1. Two LTI systems: Fig. 2.18(a). Interconnection of two LTI systems.
(a) Parallel connection of two
Signals and Systems systems. (b) Equivalent system.

Chapter 2—Part II
Jau-Jr Lin 2. Output: y (t )  y (t )  y (t )
1 2

2024 Spring  x(t )  h1 (t )  x(t )  h2 (t )

 
y (t )   x( )h1 (t   )d   x( )h2 (t   )d
 

y (t )   x( )h1 (t   )  h2 (t   ) d where h(t) = h1(t) + h2(t)


  x( )h(t   )d  x(t )  h(t ) Fig. 2.18(b) 2


LTI Systems LTI Systems

2. The output is expressed in terms of z(t) as
 Distributive property for Continuous-time case:
y(t)  z(t)  h2 (t) (2.17)
x(t)  h1(t)  x(t)  h2 (t)  x(t)  {h1(t)  h2 (t)} (2.15) 

 Distributive property for Discrete-time case: y(t)   z( )h2 (t   )d (2.18)
-

x[n]  h1[n]  x[n]  h2 [n]  x[n]  {h1[n]  h2 [n]} (2.16) Since z(t) is the output of the first system, so it can be expressed as

2.6.2 Cascade Connection of LTI Systems z( )  x( )  h1( )   x( )h1(   )d (2.19)
-
1. Two LTI systems: Fig. 2.19(a). Substituting Eq. (2.19) for z(t) into Eq. (2.18) gives
 
y (t )   Change of variable
   x( v )h1 (  v )h2 (t   )dvd
=
y(t)   x( )   h1( )h2 (t     )d d
 

-    (2.20)
Define h(t) = h1(t)  h2(t), then

h(t  v)   h1 ( )h2 (t  v   )d


Figure 2.19 (p. 128) y(t)   x( )h(t   )d  x(t)  h(t) (2.21) Fig. 2.19(b).
Interconnection of two LTI systems. (a) Cascade connection of two systems. -

(b) Equivalent system. (c) Equivalent system: Interchange system order. 3. Associative property for continuous-time case:
3 4
 LTI Systems LTI Systems
{x(t)  h1(t)}  h2 (t)  x(t)  {h1(t)  h2 (t)} (2.22) Example 2.11 Equivalent System to Four Interconnected Systems
4. Commutative property: Consider the interconnection of four LTI systems, as depicted in Fig. 2.20. The
impulse responses of the systems are
Write h(t) = h1(t)  h2(t) as the integral
h1[n]  u[n], h2 [n]  u[n  2]  u[n], h3 [n]   [n  2], and h4 [n]   nu[n].
 Change of variable
h(t )   h1 ( )h2 (t   )d =t Find the impulse response h[n] of the overall system.


h(t)   h1(t   )h2 ( )d  h2 (t)  h1(t) (2.23) Fig. 2.19(c).
-

Interchanging the order of the LTI systems in the cascade without affecting
the result:
x(t )  h1 (t )  h2 (t )  x(t )  h2 (t )  h1 (t ) ,
Figure 2.20
Commutative property for continuous-time case: (p. 131)
h1(t)  h2 (t)  h2 (t)  h1(t) (2.24) Interconnection of
systems for
5. Associative property for discrete-time case:
Example 2.11.
{x[n]  h1[n]}  h2 [n]  x[n]  {h1[n]  h2 [n]} (2.25) <Sol.>
Commutative property for discrete-time case: 1. Parallel combination of h1[n] and h2[n]:
h1[n]  h2 [n]  h2 [n]  h1[n] (2.26) h12[n] = h1[n] + h2[n] Fig. 2.21 (a).
5 6

LTI Systems LTI Systems

2. h12[n] is in series with h3[n]:
Figure 2.21
h123[n] = h12[n]  h3[n]
(p. 131)
(a) Reduction of h123[n] = (h1[n] + h2[n])  h3[n] Fig. 2.21 (b).
parallel combination 3. h123[n] is in parallel with h4[n]:
of LTI systems in
upper branch of Fig. h[n] = h123[n]  h4[n]
2.20. (b) Reduction h[n]  (h1[n]  h2 [n])  h3 [n]  h4 [n], Fig. 2.21 (c).
of cascade of
systems in upper Thus, substitute the specified forms of h1[n] and h2[n] to obtain
branch of Fig. h12 [n]  u[n]  u[n  2]  u[n]
2.21(a).  u[n  2]
(c) Reduction of
parallel combination Convolving h12[n] with h3[n] gives
of systems in Fig. h123[n]  u[n  2]   [n  2]
2.21(b) to obtain an
equivalent  u[n]
system for Fig. 2.20.
 
h[n]  1   n u[n].

7 8
 LTI Systems LTI Systems
2.7 Relation Between LTI System Properties and the
Impulse Response
2.7.1 Memoryless LTI Systems
1. The output of a discrete-time LTI system:

y[n]  h[n]  x[n]   h[k ]x[n  k ]
k 

y[n]    h[ 2]x[n  2]  h[ 1]x[n  1]  h[0]x[n]  h[1]x[n  1]  h[2]x[n  2]  

(2.27)
 Table 2.1 summarizes the interconnection properties presented in this section.
2. To be memoryless, y[n] must depend only on x[n] and therefore cannot
depend on x[n  k] for k  0.
A discrete-time LTI system is memoryless if and only if
h[k ]  c [k ] c is an arbitrary constant

9 10

LTI Systems LTI Systems

 Continuous-time system: 3. Convolution sum in new form:
1. Output: 

 y[n]   h[k ]x[n  k ].

y (t )   h( )x(t   )d , k 0

 Continuous-time system:
2. A continuous-time LTI system is memoryless if and only if
1. Convolution integral: 3. Convolution integral in new form:
h ( )  c ( ) c is an arbitrary constant  
y (t )   h( ) x(t   )d . y(t)   h( )x(t  )d .

2.7.2 Causal LTI Systems 0

The output of a causal LTI system depends only on past or present values 2. For a continuous-time causal LTI system,
of the input. h( )  0 for   0
 Discrete-time system:
1. Convolution sum: y[n]    h[2] x[n  2]  h[ 1]x[ n  1]  h[0] x[ n] 2.7.3 Stable LTI Systems
A system is BIBO stable if the output is guaranteed to be bounded for every
 h[1]x[n  1]  h[2]x[n  2]  . bounded input.
2. For a discrete-time causal LTI system,  Discrete-time case: Input x[n]  M x   Output: y[n]  M y  

h[k ]  0 for k  0
11 12
 LTI Systems 
LTI Systems
1. The magnitude of output:

 h[k ]  .
k 
y[n]  h[n]  x[n]   h[k ]x[n  k ]
k 
 Continuous-time case:

Condition for impulse response of a stable continuous-time LTI system:
ab  a  b
y[n]   h[k ]x[n  k ] 
k  0
h( ) d  .

ab  a b
y[n]   h[k ] x[n  k ] Example 2.12 Properties of the First-Order Recursive System
k  The first-order system is described by the difference equation
2. Assume that the input is bounded, i.e., y[n]   y[n  1]  x[n]
x[n]  M x   x[n  k ]  M x and has the impulse response
and it follows that h[n]   nu[n]

Is this system causal, memoryless, and BIBO stable?
y[n]  Mx  h[k] (2.28) <Sol.>
k 
1. The system is causal, since h[n] = 0 for n < 0.
Hence, the output is bounded, or y[n] ≤  for all n, provided that the impulse 2. The system is not memoryless, since h[n]  0 for n > 0.
response of the system is absolutely summable. 3. Stability: Checking whether the impulse response is absolutely summable?
3. Condition for impulse response of a stable discrete-time LTI system: 13 14

  
LTI Systems LTI Systems
 h[k ]    k      2.7.4 Invertible Systems and Deconvolution
k
if and only if  < 1
k  k 0 k 0 A system is invertible if the input to the system can be recovered from the
◆ Special case: output except for a constant scale factor.
A system can be unstable even though the impulse response has a finite value. 1. h(t) = impulse response of LTI system,
Fig. 2.24.
1. Ideal integrator: 2. hinv(t) = impulse response of LTI inverse system
t
y(t)   x( )d (2.29)


Input: x() =  (), then the output is y(t) = h(t) = u(t).

h(t) is not absolutely integrable Figure 2.24 (p. 137)
Ideal integrator is not stable! Cascade of LTI system with impulse response h(t) and inverse system with
impulse response h-1(t).
2. Ideal accumulator:
n 3. The process of recovering x(t) from h(t)  x(t) is termed deconvolution.
y[n]   x[k ]
k 
4. An inverse system performs deconvolution.
x(t )  (h(t )  hin v (t ))  x(t ). Continuous-time
Impulse response: h[n] = u[n]
case
h[n] is not absolutely summable h(t )  h (t )   (t )
inv
(2.30)
Ideal accumulator is not stable! 5. Discrete-time case: h[n]  hinv [n]   [n] (2.31)
15 16
 LTI Systems LTI Systems
Example 2.13 Multipath Communication Channels: Compensation by means of 2) For n = 0, [n] = 1, and eq. (2.32) implies that
an Inverse System
hinv [n]  ahinv [n  1]  0,
Consider designing a discrete-time inverse system to eliminate the distortion
associated with multipath propagation in a data transmission problem. Assume hinv [n]  ahinv [n  1] (2.33)
that a discrete-time model for a two-path communication channel is
3. Since = 1, Eq. (2.33) implies that hinv[1] =  a, hinv[2] = a2, hinv[3] =  a3,
hinv[0]
y[n]  x[n]  ax[n  1]. and so on.
Find a causal inverse system that recovers x[n] from y[n]. Check whether this The inverse system has the impulse response
inverse system is stable. hinv [n]  (a ) n u[n]
<Sol.>
1. Impulse response: 4. To check for stability, we determine whether hinv[n] is absolutely summable,
which will be the case if
1, n0  

 hinv [k ]  
k
h[n]  a, n 1 a is finite.
 0, otherwise k  k 
  For a < 1, the system is stable.
2. The inverse system hinv[n] must satisfy h[n]  hinv[n] = [n]. ★ Table 2.2 summarizes the relation between LTI system properties and
hinv [n]  ahinv [n  1]   [n]. (2.32) impulse response characteristics.
1) For n < 0, we must have hinv[n] = 0 in order to obtain a causal inverse
system 17 18

LTI Systems LTI Systems

2.8 Step Response
1. The step response is defined as the output due to a unit step input signal.
2. Discrete-time LTI system:
Let h[n] = impulse response and s[n] = step response.

s[n]  h[n]* u[n]   h[k ]u[n  k ].
k 

3. Since u[n  k] = 0 for k > n and u[n  k] = 1 for k ≤ n, we have

n
s[n]   h[k ].
k 

The step response is the running sum of the impulse response.

 Continuous-time LTI system:
t
s(t)   h( )d (2.34)


The step response s(t) is the running integral of the impulse response h(t).
19 20
 LTI Systems LTI Systems
1. Step response:
◆ Express the impulse response in terms of the step response as 
1  RC
t
h[n]  s[n]  s[n  1] and h(t ) 
d
s (t ) s (t )   e u ( )d .
 RC
dt
Example 2.14 RC Circuit: Step Response  0, t0

The impulse response of the RC circuit depicted in Fig. 2.12 is s (t )   1 t  
 RC 
 e RC u ( )d t0
1  RCt
h(t )  e u (t )
RC  0, t0
Find the step response of the circuit. 
s (t )   1 t  
 RC 0
<Sol.>  e RC d t  0
Figure 2.12 (p. 119)  0, t0
RC circuit system with the voltage source x(t)
 
t
as input and the voltage measured across
the capacitor y(t), as output. 1  e RC , t  0 Figure 2.25 (p. 140)
RC circuit step response for RC = 1 s.
Fig. 2.25

21 22

Problem 2.50

23 24
 Homework #2
• Due on April 1 (Monday)
– 2.32 (a)(b), 2.33 (a)(b), 2.34 (a)(b), 2.38 (a)(b), 2.39
(a)(b), 2.40 (a)(b) , 2.49 (a)(c)(e)(h)

25