Chapter Three

ANALYSIS OF LINEAR TIME-INVARIANT SYSTEMS

3.1. Introduction
In this chapter, we will focus on understanding the behaviour of linear time-invariant (LTI)
systems. As we have established in chapter 2, linearity and time-invariance properties provide
a nice way to understand the input-output relationship of system. Detail description of these
systems (both discrete time and continuous time) is done in this chapter. We will begin by
describing discrete-time LTI system and then finally conclude with continuous-time LTI
systems.

3.2. Discrete-Time LTI Systems

Consider a discrete-time system with input x[n] and output y[n]. First, define the impulse
response of the system to be the output when x[n] = δ[n] (i.e., the input is an impulse function).
We denote this impulse response by the signal h[n].

Considering an arbitrary signal x[n]. Recall from the shifting property of the impulse function
that.
∞

𝑥[𝑛] = ∑ 𝑥[𝑘]δ[𝑛 − 𝑘]
𝑘=−∞

i.e., x[n] can be written as a superposition of scaled and shifted impulse functions. Since the
system is time-invariant, the response of the system to the input δ[𝑛 − 𝑘] is h[𝑛 − 𝑘] . yy
linearity (and specifically the scaling property), the response to 𝑥[𝑘]δ[𝑛 − 𝑘] is 𝑥[𝑘]δ[𝑛 − 𝑘].
yy the additivity property, the response to:

∑ 𝑥[𝑘]δ[𝑛 − 𝑘]
𝑘=−∞

Is then,
∞

𝑦[𝑛] = ∑ 𝑥[𝑘]h[𝑛 − 𝑘]
𝑘=−∞
The above is called the convolution sum: the convolution of the signals x[n] and h[n] is
denoted by:
∞

𝑥[𝑛] ∗ 𝑦[𝑛] = ∑ 𝑥[𝑘]h[𝑛 − 𝑘]

𝑘=−∞

Thus, we have the following very important property of discrete-time LTI systems:

If x[n] is the input signal to an LTI system, and h[n] is the impulse response of the system, then
the output of the system is given as:

y[n] = x[n] * h[n]

Example:
Consider an LTI system with impulse response:
1 𝑖𝑓 0 ≤ 𝑛 ≤ 3
ℎ [𝑛 ] = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Suppose the input signal is:
1 𝑖𝑓 0 ≤ 𝑛 ≤ 3
𝑥 [𝑛 ] = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Then we have:
∞

𝑦[𝑛] = 𝑥[𝑛] ∗ 𝑦[𝑛] = ∑ 𝑥[𝑘]h[𝑛 − 𝑘]

𝑘=−∞

Since both x[k] = 0 for k < 0 and h[n – k] = 0 for k > n,

n

𝑦[𝑛] = ∑ 𝑥[𝑘]h[𝑛 − 𝑘]
𝑘=0

Thus y[n] = 0 for n<0.

When n=0, we have:
0

𝑦[0] = ∑ 𝑥[𝑘]h[−𝑘] = 𝑥 [0]ℎ[0] = 1

𝑘=0

When n = 1, we have:
 1

𝑦[1] = ∑ 𝑥[𝑘]h[1 − 𝑘] = 𝑥[0]ℎ[1] + 𝑥[1]ℎ[0] = 2

𝑘=0

Similarly:
y[2] = 3; y[3] = 4; y[4] = 3; y[5] = 2; y[6] = 1; and y[n] = 0 for n ≥ 7

Example:
Consider an LTI system with unique response h[n] = u[n]. Suppose the input signal is x[n] =
αnu[n] with 0 < α < 1. Then we have:
∞

𝑦[𝑛] = 𝑥[𝑛] ∗ 𝑦[𝑛] = ∑ 𝑥[𝑘]h[𝑛 − 𝑘]

𝑘=−∞

Since both x[k] = 0 for k < 0 and h[n – k]n= 0 for k > n, we have:

n n
1 − α𝑛+1
𝑦[𝑛] = ∑ 𝑥[𝑘]h[𝑛 − 𝑘] = ∑ α𝑘 =
1−α
𝑘=0 𝑘=0

For n ≥ 0, and y[n] = 0 for n < 0.

3.3. Continuous – Time LTI Systems

The analysis we developed for discrete-time LTI systems will be applied for continuous-time
LTI systems. Recall that for any signal x(t), we can write:

ꝏ
𝑦(𝑡) = ∫ 𝑥(𝜏)𝛿(𝑡 − 𝜏)𝑑𝜏
−ꝏ

The expression on the right-hand side is a superposition of scaled and shifted impulse functions. Thus,
when this signal is applied to an LTI system, the output will be a superposition of scaled and shifted
impulse responses. More specifically, if h(t) is the output of the system when the input is x(t) = 𝛿 (t),
then the output for a general input x(t) is given by:
 ꝏ
𝑦(𝑡) = ∫ 𝑥(𝜏)ℎ(𝑡 − 𝜏)𝑑𝜏
−ꝏ

This is the convolution integral and is denoted as:

y(t) = x(t) * h(t)

Example 1
Suppose 𝑥(𝑡) = 𝑒 −αt u(t) with α ∈ R >0 and ℎ(𝑡) = 𝑢(𝑡) hhnn thn uutuut u thn Th
systnm with imuulsn rnsuunsn h(t) is givnn as:
ꝏ
𝑦(𝑡) = 𝑥 (𝑡) ∗ ℎ(𝑡) = ∫ 𝑥 (𝜏)ℎ(𝑡 − 𝜏)𝑑𝜏
−ꝏ
ꝏ
= ∫ 𝑒−α𝜏 u(𝜏)ℎ(𝑡 − 𝜏)𝑑𝜏
−ꝏ
𝑡
= ∫ 𝑒−α𝜏 𝑑𝜏
0

t > 0 and y(t) = 0 uthnrwisn Evaluating thn abuvn nxurnssiun, wn havn:

1
(1 − 𝑒 −α𝑡 ) 𝑖𝑓 𝑡 ≥ 0
𝑦(𝑡) = { α
0 𝑖𝑓 𝑡 < 0
Example 2
Cunsidnr thn signals:
1 𝑖𝑓 0 < 𝑡 < 𝑇 𝑡 𝑖𝑓 0 < 𝑡 < 2𝑇
𝑥(𝑡) = { , ℎ(𝑡) = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

Where T > 0 is some constant. The convolution of these signals is easiest to do graphically
and by considering different regions of the variable t. The results are:

0 𝑡<0
1 2
𝑡 0<𝑡<𝑇
𝑦(𝑡) = 2
1
𝑇𝑡 − 𝑇 2 𝑇 < 𝑡 < 2𝑇
{ 2
 1 3
− 𝑡 2 + 𝑇𝑡 + 2 𝑇 2 2𝑇 < 𝑡 < 3𝑇
𝑦(𝑡) = { 2
0 𝑡 > 3𝑇

3.4. Properties of Linear Time-Invariant Systems

3.4.1. The Commutative Property
The first useful property of convolution is that it is commutative:
𝑥[𝑛] ∗ ℎ[𝑛] = ℎ[𝑛] ∗ 𝑥[𝑛]

𝑥 (𝑡) ∗ ℎ(𝑡) = ℎ(𝑡) ∗ 𝑥 (𝑡)

In order to proof this, we start with the definition of convolution and perform a change of variable by
setting:

𝑟 = 𝑛−𝑘
This gives:
∞ ∞

𝑥[𝑛] ∗ ℎ[𝑛] = ∑ 𝑥[𝑘]ℎ[𝑛 − 𝑘] = ∑ 𝑥[𝑛 − 𝑟]ℎ[𝑟] = ℎ[𝑛] ∗ 𝑥[𝑛]

𝑘=−∞ 𝑟=−∞

The same holds for continuous-time convolution. Thus, it does not matter which of the
signals we choose to flip and shift in the convolution operation.

3.4.2. The Distributive Property

The second useful property of convolution is that it is distributive:
𝑥[𝑛] ∗ (ℎ1 [𝑛] + ℎ2 [𝑛]) = 𝑥[𝑛] ∗ ℎ1 [𝑛] + 𝑥[𝑛] ∗ ℎ2 [𝑛]

𝑥(𝑡) ∗ (ℎ1 (𝑡) + ℎ2 (𝑡)) = 𝑥(𝑡) ∗ ℎ1 (𝑡) + 𝑥 (𝑡) ∗ ℎ2 (𝑡)

This property is easy to verify:

∞
𝑥[𝑛] ∗ (ℎ1 [𝑛] + ℎ2 [𝑛]) = ∑ 𝑥[𝑘](ℎ1 [𝑛 − 𝑘] + ℎ2 [𝑛 − 𝑘])
𝑘=−∞
∞ ∞
= ∑ 𝑥[𝑘]ℎ1 [𝑛 − 𝑘]+= ∑ 𝑥[𝑘] ℎ2 [𝑛 − 𝑘])
𝑘=−∞ 𝑘=−∞

= 𝑥[𝑛] ∗ ℎ1 [𝑛] + 𝑥[𝑛] ∗ ℎ2 [𝑛]

The distributive property has implications for LTI systems connected in parallel:
 Figure 3.1(a). Parallel Connection (distributive property)

Let h1[n] be the impulse response of system 1, and let h2[n] be the impulse response of system 2.
Then we have:

𝑦1 [𝑛] = 𝑥[𝑛] ∗ ℎ1 [𝑛]

𝑦2 [𝑛] = 𝑥[𝑛] ∗ ℎ2 [𝑛]
Thus:
𝑦[𝑛] = 𝑦1 [𝑛] + 𝑦2 [𝑛] = 𝑥[𝑛] ∗ ℎ1 [𝑛] + 𝑥[𝑛] ∗ ℎ2 [𝑛] = 𝑥[𝑛] ∗ (ℎ1 [𝑛] + ℎ2 [𝑛])

The above expression indicates that the parallel interconnection can equivalently be viewed as x[n]
passing through a single system whose impulse response is ℎ1 [𝑛] + ℎ2 [𝑛]

Figure 3.1(b). Parallel Connection (distributive property)

3.4.3. The Associative Property

It is useful property of convolution and it is mathematical expression is given as:
𝑥[𝑛] ∗ (ℎ1 [𝑛] ∗ ℎ2 [𝑛]) = (𝑥[𝑛] ∗ ℎ1 [𝑛]) ∗ ℎ2 [𝑛]

𝑥(𝑡) ∗ (ℎ1 (𝑡) ∗ ℎ2 (𝑡)) = (𝑥(𝑡) ∗ ℎ1 (𝑡)) ∗ ℎ2 (𝑡)

In other words, it does not matter which order we do the convolutions. The above relationships can be
proved by manipulating the summations (or integrals).

Figure 3.2(a). Series Connection (Associative property)

Just as the distributive property had implications for parallel interconnections of systems, the associative
property has implications for serios interconnections of systems. Specifically, consider the series
interconnection in figure 3.2, we have:

𝑦[𝑛] = 𝑦1 [𝑛] ∗ ℎ2 [𝑛] = (𝑥[𝑛] ∗ ℎ1 [𝑛]) ∗ ℎ2 [𝑛] = 𝑥[𝑛] ∗ (ℎ1 [𝑛] ∗ ℎ2 [𝑛])

Thus, the series interconnection is equivalent to a single system with impulse response ℎ1 [𝑛] ∗ ℎ2 [𝑛],
as shown in figure 3.2(b)

Figure 3.3.(b) Equivalent representation of series interconnection (associative property)

In addition, note that since ℎ1 [𝑛] ∗ ℎ2 [𝑛] = ℎ2 [𝑛] ∗ ℎ1 [𝑛] , we can also interchange the order of the
systems in the series interconnection as shown in figure 3.3.(c), without changing the overall input-
output relationship between x[n] and y[n].

Figure 3.4.(c) Equivalent representation of series interconnection (associative property)

3.4.4. Memoryless Property

Let h[n] (or h(t)) be the impulse response of a given LTI system. Since we have:
∞ ∞

𝑦[𝑛] = ∑ 𝑥 [𝑘]ℎ[𝑛 − 𝑘] = ∑ 𝑥[𝑛 − 𝑘]ℎ[𝑘]

𝑘=−∞ 𝑘=−∞

We see that y[n] will depend on a value of the input signal other than at time-step n unless h[k] = 0 for
k ≠ 0. In order words, for an LTI system to be memoryless, we require h[n] = K δ[n] for some constant
K. Similarly, a continuous-time LTI system is memoryless if and only if h(t) = K δ(t)

For some constant K.

3.4.5. Invertibility Property
Consider an LTI system with impulse response h[n] (or h(t)). Recall (in chapter 2), a system is said to
be invertible if the output of the system uniquely specifies the input. If a system is invertible, there is
another system (known as the inverse system) that takes the output of the original system and outputs
the input to the original system as shown in figure 3.4.

Figure 3.4. A system in series with its inverse

Suppose the second system is LTI and has impulse response hI[n]. Then, by the associative property, we
see that the series interconnection of the system with its invers is equivalent to a single system with
impulse response ℎ[𝑛] ∗ ℎ𝐼 [𝑛]. In particular, we require:

𝑥[𝑛] = 𝑥[𝑛] ∗ (ℎ[𝑛] ∗ ℎ𝐼 [𝑛])

For all input signals x[n], from which we have:

ℎ[𝑛] ∗ ℎ𝐼 [𝑛] = δ[n]

In other words, if we have an LTI system with impulse response h[n], and another LTI system with impulse
response hI[n] such that ℎ[𝑛] ∗ ℎ𝐼 [𝑛] = δ[n], then those systems are inverse of each other. This analogous
statement holds in continuous-time as well.

Example

Consider the LTI system with impulse response ℎ[𝑛] = αn u[n]. nne can verify that this impulse
response corresponds to the system:

𝑦[𝑛] = ∑ 𝑥[𝑘]αn−k = α𝑦[𝑛 − 1] + 𝑥[𝑛]

𝑘=−∞

Now consider the system:

𝑦1 [𝑛] = 𝑥1 [𝑛] − α𝑥1 [𝑛 − 1] with an inuut signal x1[n] and uutuut signal y1[n] hhn imuulsn
rnsuunsn u this systnm is:
 ℎ𝐼 [𝑛] = δ[n] − αδ[n − 1]
We have:
ℎ[𝑛] ∗ ℎ𝐼 [𝑛] = αn u[n] ∗ (δ[n] − αδ[n − 1])

= αn u[n] ∗ δ[n] − (αn u[n]) ∗ (αδ[n − 1])

= αn u[n] − α(αn−1 u[n − 1])

= αn (u[n] − u[n − 1])

= αn δ[n]

= δ[n]

Thus, the system with impulse response hI[n] is the inverse of the system with impulse response h[n]

3.4.6. Causality Property

3.4.7. Stability Property
3.4.8. Step Response of LTI Systems