Topic 5 System Properties and Convolution Sum
Topic 5 System Properties and Convolution Sum
Topic 5 System Properties and Convolution Sum
Now, let’s try to characterize some input-output properties of a CT LTI system with respect
to their impulse response h(t). The basic results will be similar in the discrete-time case.
h(t) = 0 ∀ t < 0,
The unit-impulse input is nonzero only at t = 0, and in particular is zero for t < 0. Up to
the time t = 0 a causal system does not know whether the input signal continues to be zero,
or takes a nonzero value at t = 0 and the response of an LTI system to a zero input signal is
zero. Thus, we can say that the causality is equivalent to h(t) = 0 for t < 0.
h(t) = 0 ∀ t ̸= 0.
1
It follows that y(t) can only depend on x(t) and system has no memory. (An LTI system
is called memoryless if the output signal value at any time t depends only on the input signal
value at that same time.)
Generally, a memoryless LTI system is characterized by an impulse response of the form
h(t) = Kδ(t) , where K is a complex constant.
Consequently, every memoryless LTI system with input x and output y is characterized by
an equation of the form
y = x ∗ Kδ = Kx
i.e. the system is an ideal amplifier.
Such an hI (t) may or may not exist, and if it does, it might be difficult to compute and the
above condition is often quite difficult to test.
Example: For an input x(t) = 2u(t) − u(t − 1), the response of the system will be:
Answer: Corresponding response
Z t Z t−1
y(t) = 2 h(τ )dτ − h(τ )dτ
−∞ −∞
2
2 Discrete time Convolution sum
If a discrete time sequence is given an input x[n] and its impulse response is h[n] then the
output sequence y[n] can be given as
∞
X ∞
X
y[n] = h[k]x[n − k] = x[k]h[n − k]
k=−∞ k=−∞
Note: For a discrete-time LTI system, sometimes called as LSI system, impulse response is
defined as system’s response when it’s input is an impulse δ[n].
Like the CT case, a DT system can be completely characterized by its impulse response.
In other words, the impulse response tells us everything we need to know about an LSI system
as far as signal processing is concerned.
Example: : Find the output of the system with impulse response h[n] = u[n] if the input
is given as x[n] = an u[n].
Solution: With the direct evaluation of the convolution sum we find the response of the
system
∞
X
y[n] = x[n] ∗ h[n] = x[k]h[n − k]
k=−∞
∞
X
y[n] = ak u[k]u[n − k]
k=−∞
n
X 1 − an+1
y[n] = ak = ∀ n≥0
k=0
1−a
1 − an+1
y[n] = u[n]
1−a
Example: : Find the output of the system with impulse response h[n] = u[−1, 2, 2] if the
input is given as x[n] = [1, 2, 3].
Solution:
3
Homework: Find the convolution between following two sequences. Verify if the convolu-
tion length is N1 + N2 − 1
• [2 + j, 0, 1] and [2 − j, 0, −1]
• [1, 1, 1] and [ 1, 1, 1, 1]
In continuous time domain, we defined periodic convolution of two signals x(t) and h(t),
each having period T, as Z T
x(t) ⊛ h(t) = x(τ )h(t − τ )dτ.
0
In similar way, we define circular convolution in discrete time case:
In discrete time, we can define periodic a.k.a. circular convolution of two sequences x[n]
and h[n], each having period N, as
N
X −1
x[n] ⊛ h[n] = x[k]h[n − k]
k=0
Note: In real life, if two sequences don’t have same length periods, we can add extra zeros
at the end of the shorter period sequence to make both sequences N-periodic. This is done
only for the purpose of estimating circular convolution.
4
Example: Find circular convolution of two 4 point periodic sequences given as: x[n] = [1,
1, 0, 3] and h[n] = [2, 3, 1, 1].
Solution: Use the following circulant matrix trick:
x[n] ⊛ h[n] =
1 3 0 1 2 12
1 1 3 0 3 8
=
0 1 1 3 1 7
3 0 1 1 1 8
Reference material
1. Textbook: Signals and Systems by Simon Haykin
2. Lecture notes are inspired from the course materials of JHU 520.214; MIT 6.003; Purdue
ECE-301; UVic ECE-260; and Imperial College E2.5
[Please report any typos in the notes by sending an email to the instructor.]