Chapter 3: Difference Equation
Redzuan Abdul Manap
April 2025
Chapter Outline
3.1 LTI System Properties
3.2 Difference Equation
3.2.1 Non-recursive difference equation
3.2.2 Recursive difference equation
3.3 Signal Flow Graph
3.3.1 Signal Flow Graph FIR
3.3.2 Signal Flow Graph IIR
3.4 FIR Calculation
3.5 IIR Calculation
2
3.1 Linear time invariant (LTI) system
• By the principle of superposition, the response of a discrete-time LTI
system is the sum of the responses to the individual shifted impulses
making up the input signal .
x[n] h[n] y[n]
• Few important properties of the system include linearity, time invariant,
causality and stability.
3.1 System property - Linearity
• The linearity means that the system obeys the superposition
principle
𝐻 (𝛼 𝑥 1 [𝑛]+ 𝛽 𝑥2 [𝑛])=𝛼 𝐻 (𝑥 1 [𝑛])+ 𝛽 𝐻 (𝑥 2 [𝑛])
3.1 System property – Time invariant
• A time-invariant system (shift invariant system) is one in which the
internal parameters do not vary with time.
3.1 System property - Causality
• A system is said to be causal, if its output at time instant n depends
only on the present and past input values, but not on the future
input values.
• It implies that for every choice of ; if ] for , then for .
• A causal system is the one in which the output at time depends
only on the current input at time , and its past input sample values
such as , ,..
• Otherwise, if a system output depends on future input values such
as , ,., the system is noncausal. The noncausal system cannot be
realized in real time.
3.1 System property - Stability
• A system is said to be stable, if and only if every bounded input
sequence produces a bounded output sequence.
• This type of stability is called bounded-input bounded-output (BIBO)
stability.
• The input is bounded if there exists a fixed positive finite value such
that
• Similarly, the output is bounded if there exists a fixed positive finite
value such that
or
3.1 System property - LTI
LTI system must be linear and time invariant
It can either be BIBO stable or unstable, and it
can also be causal or non-causal
3.1 System property - Example
• Check for stability, causality, linearity, and time-invariance of the system
described by
Solution:
• This transformation outputs the current value of ] multiplied by either
±1.
• It is BIBO stable, since it does not change the magnitude of and hence
satisfies the conditions for bounded-input bounded-output stability
3.1 System property - Example
• It is causal, because each output depends only on the current value of .
• It is linear because:
3.1 System property - Example
• It is time varying:
If the input is shifted by 1, the output is
If the output is shifted by 1,
By comparison
3.1 System property – Exercise (Ulearn)
• Determine whether the given systems are linear, time
invariant, causal and stable:
3.2 Differential and Difference Equation
• The difference?
– Continuous-time system can be represented by
using differential equations
– Discrete-time system can be represented by using
difference equations
13
3.2 Difference Equation - Categories
• An important class of Linear Time Invariant (LTI) of discrete-
time systems is one that is characterized by a linear
difference equation with constant coefficients.
• Difference equation can be divided into two different types:
– Non-recursive difference equation
– Recursive difference equation
14
3.2 Difference Equation – General form
• A causal LTI system can be described by a difference equation
of the following form:
(3.1)
– and are a non zero system coefficient.
– Notice that is the current output, which depends on the past output
samples , the current input sample , and the past input samples .
3.2 Difference Equation – Example 1
• Find the non-zero system coefficients given the following
difference equation:
• Solution:
– Using equation (3.1), we can see that
3.2 Difference Equation – System Representation
• An LTI system can be completely described by its unit-impulse
response, which is defined as the system response due to the
impulse input with zero initial conditions.
• I.e. when ,
3.2 Difference Equation – Example 2
• Assume we have a LTI system with an initial conditon .
a. Determine the unit-impulse response .
b. Draw the system block diagram.
c. Write the output using the obtained impulse response.
3.2 Difference Equation – Example 2
• The unit-impulse response can be obtained by replacing with .
Thus
I.e. we have an impulse response of
3.2 Difference Equation – Example 2
• The block diagram can be drawn as follow:
• The output can be written in terms of impulse response as:
3.2.1 Non-recursive difference equation
• A non-recursive LTI discrete-time system can be characterized by a linear
constant coefficient difference equation of the form below (for causal and
non recursive):
(3.2)
with if and for
• Thus a LTI, causal, and non-recursive system can be characterized by an
Nth-order linear non-recursive difference equation.
• The Nth-order non-recursive difference equation has a finite impulse
response (FIR).
3.2.1 Non-recursive difference equation – Example
• Write the equation for a second order FIR
2
𝑦 [𝑛]= ∑ 𝑏 𝑚 𝑥 [𝑛− 𝑚]
𝑚=0
22
3.2.2 Recursive difference equation
• The response of a discrete-time system depends on the present and
previous values of the input as well as the previous values of the output.
• A LTI, causal, and recursive discrete-time system can be represented by
the following Nth-order linear recursive difference equation:
(3.3)
where and are constants.
• The Nth-order recursive difference equation has an infinite impulse
response (IIR).
• Hence, an IIR filter is characterized by a recursive difference equation.
3.2.2 Recursive difference equation - Example
• Write the equation for a second order IIR.
24
3.3 Signal Flow Graph
• Signal flow graph (SFG) represents digital system signal
processing routine in the illustration/graphical form.
• There are basic components generally used to illustrate the
SFG which are
Multiplier Adder Unit Delay
x(n) k kx(n) x(n) y(n)=x(n)+w(n) x(n) -1
x(n-1) y(n)=x(n)+w(n) x(n)
Z
x(n) k kx(n) x(n) y(n)=x(n)+w(n) x(n) Z
-1
x(n-1) x(n) k kx(n) x(n) Z
-1
x(n-1)
w(n) w(n)
w(n)
Signal Flow Graph
25
3.3.1 Signal Flow Graph-FIR
• A 2nd order FIR can be represented as:
2
y[ n] b[ ]x[ n ]
0
b[0] x[ n] b[1] x[ n 1] b[ 2] x[ n 2]
• The signal flow graph describing this equation is:
z 1 z 1
x[n]
b[0] b[1] b[2]
y[n]
26
3.3.2 Signal Flow Graph-IIR
• A 2nd order IIR can be represented as:
2 2
y[ n] a[ ]y[ n ] b[ ]x[ n ]
1 0
y[n] a[1] y[n 1] a[2] y[n 2] b[0]x[n] b[1]x[n 1] b[2]x[n 2]
• The Type I (Direct Form I) signal flow graph for this is:
x[n] z 1 b[0] z 1 y[n]
z 1 b[1] a[1] z 1
b[2] a[2]
27
3.3.2 Signal Flow Graph-IIR
• A 2nd order of IIR can be represented as:
2 2
y[ n] a[ ]y[ n ] b[ ]x[ n ]
1 0
y[n] a[1] y[n 1] a[2] y[n 2] b[0]x[n] b[1]x[n 1] b[2]x[n 2]
• The Type II (Direct Form II) signal flow graph for this is:
x[n] 1 b[0] y[n]
z
a[1] z 1 b[1]
a[2] b[2]
28
3.3.2 Signal Flow Graph-IIR
A digital filter is canonic if the number of delays in the block
diagram representation is equal to the order of the difference
equation.
Otherwise it is referred as non-canonic structure.
p0
x[n] + y[n]
z-1 z-1 First-order LTI digital filter
p1 -d1
y[n] d1 y[n 1] p0 x[n] p1 x[n 1]
The above structure is a non-canonic since it employs two delays to realize
the first-order difference equation.
29
3.4 FIR calculation - Example
A discrete-time system is a 2nd order FIR defined as
y[n] 0.5 x[n] 1x[n 1] 0.5 x[n 2]
It is assumed that the input is valid for n 0, and zero for
n<0. The input is an impulse function
Calculate the output of the system.
30
3.4 FIR calculation - Example
y[n] 0.5 x[n] 1x[n 1] 0.5 x[n 2]
𝑥 [ 𝑛 ] =[1 0 0 0 0 0 0 …]
The computation for the output is shown in the following
table.
Note that the output is zero for n 3.
31
3.4 IIR calculation - Example
A discrete-time system is 1st order IIR defined as
y[n] 0.5 y[n 1] x[n]
It is assumed that the input and output are valid for n 0,
and zero for n<0. The input is an impulse function
Calculate the output of the system.
32
3.4 IIR calculation - Example
y[n] 0.5 y[n 1] x[n]
𝑥 [ 𝑛 ] =[1 0 0 0 0 0 0 …]
The computation for the output is shown in the following
table.
Note that the output does not reach zero but approaches a small value as n
increases
33
Past Exam Examples .
A discrete-time system is defined as:
It is assumed that the input x[n] and output y[n] is valid for n≥0, and zero for n<0.
The input is an impulse function.
Determine whether the system is an FIR or IIR system. Justify your answer.
Then calculate the output y(n) for and determine the system impulse response
Determine whether the system is causal and stable.
34
Past Exam Examples .
A discrete-time system is defined as:
It is assumed that the input x[n] and output y[n] is valid for n≥0, and zero for n<0.
The input is an impulse function.
Determine whether the system is an FIR or IIR system. Justify your answer.
IIR System – The system output y(n) depends on previous output y(n-1)
35
Past Exam Examples .
A discrete-time system is defined as:
Calculate the output y(n) for and determine the system impulse response
n x(n) y(n) = x(n-1) + 0.5y(n-1)
0 1 y(0) = x(-1) + 0.5y(-1) = 0 + 0.5(0) = 0
1 0 y(1) = x(0) + 0.5y(0) = 1 + 0.5(0) = 1
2 0 y(2) = x(1) + 0.5y(1) = 0 + 0.5(1) = 0.5
3 0 y(3) = x(2) + 0.5y(2) = 0 + 0.5(0.5) = 0.25
4 0 y(4) = x(3) + 0.5y(3) = 0 + 0.5(0.25) = 0.125
5 0 y(5) = x(4) + 0.5y(4) = 0 + 0.5(0.125) = 0.0625
The system impulse response when input = delta function
36
.
Past Exam Examples
A discrete-time system is defined as:
It is assumed that the input x[n] and output y[n] is valid for n≥0, and zero for n<0.
The input is an impulse function.
Determine whether the system is causal and stable.
The system is valid for n≥0, and zero for n<0, system is causal.
, so the system is stable.
37
