ELT2035 Signals & Systems

Lesson 3: Representations of CT LTI

systems in time domain

Hoang Gia Hung

Faculty of Electronics and Telecommunications
University of Engineering and Technology, VNU Hanoi
 Last lecture review
❑ Fundamental concepts of system
➢ System model: mathematical description of the rules of operation
➢ System analysis vs system synthesis (design)

❑ Classification of system
➢ By configurations (SISO, MIMO, …)
➢ By properties of the system (CT/DT, Instantaneous/dynamic,
causal/noncausal, time varying/time invariant, …)

❑ Properties of systems
➢ Stability: BIBO criterion
➢ Memory/Memoryless
➢ Causality: cannot “see” future inputs
➢ Invertibility: ability to restore the input from the output
➢ Time invariance: system behaviours do not change with time
➢ Linearity: superposition + homogeneity

❑ Today’s lecture: Modelling of CT LTI systems in time domain

 Introduction

Input signal Output signal

(excitation) System (response)
x(t) H[.] y(t)

❑ Objectives: design/synthesis a system which will produce the

desired output for the given input. How?
➢ The first step is the construction of a system model (i.e. determine an
operator H[.] that satisfactorily approximates the behaviour of the system).
➢ Next: use the constructed system to analyze the system rules of operation

❑ System modelling techniques

➢ Internal description approach: H[.] is derived from a knowledge of the
internal structure of the system
➢ External description (“black box”) approach: H[.] is obtained from the
measurements of the system’s input and the corresponding output
 A circuit modelling example
❑ Input: 𝑥(𝑡), output: 𝑦 𝑡 , what’s the system’s model?
➢ Model the resistor (Ohm’s law): 𝑣𝑅 𝑡 = 𝑅𝑦(𝑡)
𝑑𝑦(𝑡)
➢ Model the inductor (the 𝑖 − 𝑣 eq.): 𝑣𝐿 𝑡 = 𝐿
𝑑𝑡
➢ Model the circuit (Kirchoff’s valtage law):
𝑑𝑦(𝑡)
𝑅𝑦 𝑡 + 𝐿 = 𝑥(𝑡) (1)
𝑑𝑡
➢ Eq. (1): linear constant-coefficient ordinary differential equation (LCCODE)

❑ Generalization: a CT LTI system is described by the LCCODE

𝑛 𝑑 𝑖 𝑦(𝑡) 𝑚 𝑑 𝑗 𝑥(𝑡)
σ𝑖=0 𝑎𝑖 = σ𝑗=0 𝑏𝑗 , (2)
𝑑𝑡 𝑖 𝑑𝑡 𝑗
where 𝑎𝑖 and 𝑏𝑗 are constants.
𝑑
➢ Denote 𝐷 = , eq. (2) is rewritten as: 𝐷𝑛 + 𝑎𝑛−1 𝐷𝑛−1 + ⋯ + 𝑎0 𝑦 𝑡 =
𝑑𝑡
𝑏𝑚 𝐷 + 𝑏𝑚−1 𝐷𝑚−1 + ⋯ + 𝑏0 𝑥 𝑡 , or
𝑚

𝑄 𝐷 𝑦 𝑡 =𝑃 𝐷 𝑥 𝑡 . (3)
➢ Eq. (3) is obtained by applying natural laws (e.g. Ohm/Kirchoff laws) to the
internal structure of the system (e.g. components and topology in a circuit).
 Solving for LCCODE – the calculus
approach
❑ System analysis: use the model to determine the response of
the system to any given input. Two methods:
➢ Solution to linear constant-coefficients differential equations (today)
➢ Laplace transform (later in the course)
𝑑𝑦(𝑡)
❑ The R-L circuit example: 𝑅𝑦 𝑡 + 𝐿 = 𝑥(𝑡). Given: 𝐿 =
𝑑𝑡
1 𝐻 , 𝑅 = 1 [Ω], 𝑥 𝑡 = cos(𝑡), 𝑦 0 = 2 [𝐴], find 𝑦 𝑡 .
𝑑𝑦𝑛 (𝑡)
➢ Homogeneous solution (natural response): 𝑦𝑛 𝑡 + = 0 → 𝑦𝑛 𝑡 =
𝑑𝑡
𝐴𝑛 𝑒 𝜆𝑡 →𝐴𝑛 𝑒 𝜆𝑡 1 + 𝜆 = 0 → 𝜆 = −1 → 𝑦𝑛 (𝑡) = 𝐴𝑛 𝑒 . −𝑡

𝑑𝑦𝑓 (𝑡)
➢ Particular solution (forced response): 𝑦𝑓 𝑡 + = cos(𝑡) → 𝑦𝑓 𝑡 =
𝑑𝑡
1
𝐵 cos(𝑡 + 𝐶) → 𝐵 cos 𝑡 + 𝐶 − 𝐵 sin 𝑡 + 𝐶 = cos(𝑡) → 𝑦𝑓 𝑡 = cos(𝑡) +
2
1
sin(𝑡).
2
1
➢ General solution (total response): 𝑦 𝑡 = 𝑦𝑛 𝑡 + 𝑦𝑓 𝑡 = 𝐴𝑛 𝑒 −𝑡 + cos(𝑡) +
2
1 1 3 3 1 1
sin(𝑡) → 𝑦 0 = 𝐴𝑛 + → 𝐴𝑛 = , 𝑦 𝑡 = 𝑒 −𝑡 + cos(𝑡) + sin(𝑡).
2 2 2 2 2 2
❖ Note: the initial condition is imposed to the total response, which begins right
after the input is applied at 𝑡 = 0 → the i.c. is required at 𝑡 = 0+ .
Solving for LCCODE – the engineering
approach
❑ In practice, it is difficult to measure initial conditions at 𝑡 = 0+ .
➢ It’s much easier to determine 𝑦 0− .
➢ 𝑦 0− = internal condition (e.g. 𝑡=0
𝑦 0−
energy storage) of the system prior
to the application of the input.

❑ With respect to initial conditions at 𝑡 = 0− , the total response is

the sum of two independent components:
Total response = zero-input response + zero-state response

➢ Z-I response is due to 𝑦 0− alone & independent of external input: 𝑦𝑍𝐼 𝑡 +

𝑑𝑦𝑍𝐼 (𝑡)
= 0 → 𝑦𝑍𝐼 (𝑡) = 𝐴𝑍𝐼 𝑒 −𝑡 → 𝑦𝑍𝐼 0 = 2 → 𝐴𝑍𝐼 = 2, 𝑦𝑍𝐼 (𝑡) = 2𝑒 −𝑡 .
𝑑𝑡
➢ Z-S response is caused by external input only, i.e. the internal condition is 0:
𝑑𝑦 (𝑡) 1 1
𝑦𝑍𝑆 𝑡 + 𝑍𝑆 = cos(𝑡) → 𝑦𝑍𝑆 𝑡 = 𝐴𝑍𝑆 𝑒 −𝑡 + cos(𝑡) + sin(𝑡) → 𝑦 0 =
𝑑𝑡 2 2
1 1 1 1
𝐴𝑍𝑆 + = 0 → 𝑦𝑍𝑆 𝑡 = − 𝑒 −𝑡 + cos(𝑡)
+ sin(𝑡).
2 2 2 2
3 1 1
➢ Total response: 𝑦 𝑡 = 𝑦𝑍𝐼 𝑡 + 𝑦𝑍𝑆 𝑡 = 𝑒 −𝑡 + cos(𝑡) + sin(𝑡).
2 2 2
Relations between the two approaches
❑ Natural response vs zero-input response:
➢ Both are solutions to 𝑄 𝐷 𝑦 𝑡 = 0 (with different initial conditions) → same
form, differs in constant coefficient of 𝑒 −𝑡 .
➢ natural response = zero-input response + all characteristic modes (i.e. 𝑒 −𝑡
term in this case) from zero-state response.

❑ Forced response vs. zero-state response:

➢ Both are solutions to 𝑄 𝐷 𝑦 𝑡 = 𝑃 𝐷 𝑥 𝑡 (different initial conditions) →
same form, differs in constant coefficient of 𝑒 −𝑡 .
➢ forced response = zero-state response - characteristic mode terms

𝟑 −𝑡
𝑦𝑍𝐼 𝑡 = 𝟐𝑒 −𝑡 𝑦𝑛 𝑡 = 𝑒
𝟐
𝟏 1 1 1 1
𝑦𝑍𝑆 𝑡 = − 𝑒 −𝑡 + cos(𝑡) + sin(𝑡) 𝑦𝑓 𝑡 = 𝟎. 𝑒 −𝑡 + cos(𝑡) + sin(𝑡)
𝟐 2 2 2 2
Zero-input response of CT-LTI systems
❑ Solution to 𝐷 𝑛 + 𝑎𝑛−1 𝐷 𝑛−1 + ⋯ + 𝑎0 𝑦0 𝑡 = 0 (4)
➢ Linear combination of 𝑦0 𝑡 and its 𝑛 successive derivatives are zero for all
t → 𝑦0 𝑡 and its derivatives must have the same form → 𝑦0 𝑡 = 𝑐𝑒 𝜆𝑡 .
➢ Substitute 𝑐𝑒 𝜆𝑡 for 𝑦0 𝑡 yields: 𝑐 𝜆𝑛 + 𝑎𝑛−1 𝜆𝑛−1 + ⋯ + 𝑎0 𝑒 𝜆𝑡 = 0 ∀𝑡 →
nontrivial solution satisfies 𝜆𝑛 + 𝑎𝑛−1 𝜆𝑛−1 + ⋯ + 𝑎0 = 0 or 𝑄 𝜆 = 0.
➢ Eq. (4) has 𝑛 possible solutions 𝑐1𝑒 𝜆1𝑡 , … , 𝑐𝑛𝑒 𝜆𝑛𝑡 → the general solution is
𝑦0 𝑡 = 𝑐1 𝑒 𝜆1 𝑡 + ⋯ + 𝑐𝑛 𝑒 𝜆𝑛 𝑡 , where 𝑐1:𝑛 are defined by 𝑛 initial conditions.

❑ 𝑄 𝜆 is referred to as the characteristic polynomial of the

system, while 𝑄 𝜆 = 0 is called the characteristic equation.
➢ 𝜆1:𝑛 are called characteristic roots, or eigenvalues, or natural frequencies.
➢ 𝑒 𝜆𝑖𝑡 are called characteristic/natural modes of the system.

❑ The zero-input response is a linear combination of the

system’s characteristic modes.
➢ Non-characteristic modes are lumped together in the forced response.
Zero-state response of CT LTI systems
to the unit impulse input
h(t) = H[δ(t)]

System
H[.]

❑ When 𝛿(𝑡) enters the input of the system:

➢ All the initial conditions are 0 at 𝑡 = 0− , then at 𝑡 = 0 the impulse creates
nonzero initial conditions at 𝑡 = 0+ → ℎ 𝑡 must consists of the system’s
characteristic modes for 𝑡 ≥ 0+ .
➢ 𝛿(𝑡) vanishes for 𝑡 > 0, but what happens on the system’s output at the
single moment 𝑡 = 0? There can at most be an impulse.

❑ For LTI systems: ℎ 𝑡 = 𝑏𝑛 𝛿 𝑡 + 𝑃 𝐷 𝑦𝑛 𝑡 𝑢(𝑡).

➢ 𝑦𝑛 𝑡 is a linear combination of the characteristic modes that satisfies:
𝑛 = 1: 𝑦𝑛 0 = 1
𝑛 = 2: 𝑦𝑛 0 = 0 and 𝑦ሶ𝑛 0 = 1
𝑛 = 3: 𝑦𝑛 0 = 𝑦ሶ𝑛 0 = 0 and 𝑦ሷ𝑛 0 = 1
𝑛 = 4: 𝑦𝑛 0 = 𝑦ሶ𝑛 0 = 𝑦ሷ𝑛 0 = 0 and 𝑦ഺ𝑛 (0) = 1
➢ ℎ 𝑡 is a descriptor of the system and is called the impulse response.
 Impulse response example
Determine the unit impulse response ℎ(𝑡) for a system specified
by the equation 𝐷 2 + 3𝐷 + 2 𝑦 𝑡 = 𝐷𝑓(𝑡).
=?
f(t) = δ(t)

𝐷2 + 3𝐷 + 2 𝑦 𝑡 = 𝐷𝑓(𝑡)

Solution
➢ 𝑄 𝐷 = 𝐷2 + 3𝐷 + 2, 𝑃 𝐷 = 𝐷 ⟹ 𝑏2 = 0 ⟹ ℎ 𝑡 = 𝑏2 𝛿 𝑡 +
𝑃 𝐷 𝑦𝑛 𝑡 𝑢 𝑡 = 𝑦ሶ𝑛 𝑡 𝑢(𝑡).
➢ Characteristic polynomial: 𝜆2 + 3𝜆 + 2 = 0 ⟹ eigenvalues 𝜆1 =
− 1, 𝜆2 = −2 ⟹ 𝑦𝑛 𝑡 = 𝑐1 𝑒 −𝑡 + 𝑐2 𝑒 −2𝑡 .
➢ Since 𝑛 = 2, the initial conditions are: 𝑦ሶ𝑛 0 = 1, 𝑦𝑛 0 = 0.
➢ Differentiate 𝑦𝑛 𝑡 and substitute to the initial conditions yields
0 = 𝑐1 + 𝑐2
ቊ ⟹ 𝑐1 = 1, 𝑐2 = −1 ⟹ 𝑦𝑛 𝑡 = 𝑒 −𝑡 − 𝑒 −2𝑡 .
1 = −𝑐1 − 2𝑐2
➢ Hence, ℎ 𝑡 = 𝐷𝑦𝑛 𝑡 𝑢 𝑡 = −𝑒 −𝑡 + 2𝑒 −2𝑡 𝑢(𝑡).
Zero-state response of CT LTI systems
to an arbitrary input
❑ Given ℎ(𝑡), what’s the Z-S response to an arbitrary input 𝑓(𝑡)?
System
𝛿(𝑡) → ℎ(𝑡)

❑ Approximate 𝑓(𝑡) as a sum of narrow rectangular pulses

1
𝛿መ 𝑡 = ቐ∆𝜏 0 ≤ 𝑡 < ∆𝜏
0 otherwise

෠
ℎ(𝑡)
𝛿መ 𝑡

Impulse approximation

෠ − 𝑛∆𝜏)
𝑓(𝑛∆𝜏)∆𝜏 ℎ(𝑡
መ − 𝑛∆𝜏)
𝑓(𝑛∆𝜏)∆𝜏 𝛿(𝑡

LTI properties
Zero-state response of CT LTI systems
መ
𝑓 0 ∆𝜏 𝛿(𝑡) ෠
𝑓 0 ∆𝜏 ℎ(𝑡)
System
H[.]

መ − ∆𝜏) ෠ − ∆𝜏)
𝑓 ∆𝜏 ∆𝜏 ℎ(𝑡
𝑓 ∆𝜏 ∆𝜏 𝛿(𝑡

መ − 𝑛∆𝜏)
𝑓(𝑛∆𝜏)∆𝜏 𝛿(𝑡 ෠ − 𝑛∆𝜏)
𝑓(𝑛∆𝜏)∆𝜏 ℎ(𝑡

Superposition

መ
𝑓(𝑡)
≈ 𝑓(𝑡)
𝑦(𝑡)
ො
≈ 𝑦(𝑡)
Convolution

❑ In the limit, approximations become exact representations

➢ The operator that performs superposition of the scaled, shifted impulse
responses to produce the system’s output is called convolution.
 Derivation of the convolution integral

System
H[.]

❑ Applying linearity and time-invariant properties:

መ
𝛿(𝑡) → ෠
ℎ(𝑡)
መ − 𝑛∆𝜏)
𝑓(𝑛∆𝜏)∆𝜏 𝛿(𝑡 → ෠ − 𝑛∆𝜏)
𝑓(𝑛∆𝜏)∆𝜏 ℎ(𝑡
∞ ∞
መ − 𝑛∆𝜏) →
෍ 𝑓(𝑛∆𝜏)∆𝜏 𝛿(𝑡 ෠ − 𝑛∆𝜏)
෍ 𝑓(𝑛∆𝜏)∆𝜏 ℎ(𝑡
𝑛=−∞ 𝑛=−∞
መ
𝑓(𝑡)≈ 𝑓(𝑡) ො
𝑦(𝑡)≈ 𝑦(𝑡)

❑ Taking the limit:

∞ ∞
መ − 𝑛∆𝜏)∆𝜏
lim ෍ 𝑓(𝑛∆𝜏)𝛿(𝑡 → ෠ − 𝑛∆𝜏)∆𝜏
lim ෍ 𝑓 𝑛∆𝜏 ℎ(𝑡
∆𝜏→0 ∆𝜏→0
𝑛=−∞ 𝑛=−∞
∞ ∞
න 𝑓 𝜏 𝛿 𝑡 − 𝜏 𝑑𝜏 → න 𝑓 𝜏 ℎ 𝑡 − 𝜏 𝑑𝜏
−∞ −∞
= 𝑓(𝑡) = 𝑦(𝑡)
∞
❑ Convolution: 𝑦 𝑡 = ‫׬‬−∞ 𝑓 𝜏 ℎ 𝑡 − 𝜏 𝑑𝜏 .
 Convolution integral evaluation
❑ Find 𝑦 𝑡 = 𝑥1 𝑡 ∗ 𝑥2 𝑡 ➢ Step 1
∞
= ‫׬‬−∞ 𝑥1 𝜆 𝑥2 𝑡 − 𝜆 𝑑𝜆.

➢ Steps 2 → 4 for 2 ≤ 𝑡 < 3

❑ Steps in convolution operation

1. Fold/Reflect 𝑥2 𝑡 → 𝑥2 (−𝜆).
2. Shift/Slide 𝑥2 (−𝜆) by 𝒕 (seen as a
➢ Steps 2 → 4 for 3 ≤ 𝑡 < 4
constant inside the integration).
3. Find 𝑔𝑡 𝜆 = 𝑥1 𝜆 𝑥2 𝑡 − 𝜆 .
4. Integrate 𝑔𝑡 𝜆 over 𝝀.
5. Repeat 2 → 4 for all values of 𝑡.
❑ Result:
0, 𝑡<2 ➢ Steps 2 → 4 for 4 ≤ 𝑡 ≤ 5
𝑡−2 2≤𝑡 <3
𝑦 𝑡 = 1 3≤𝑡<4
5−𝑡 4≤𝑡 <5
0 𝑡≥5
 Impulse response - a descriptor of LTI
systems
❑ Given ℎ(𝑡), we can determine the system’s output 𝑦(𝑡) for any
input 𝑥(𝑡) by performing the convolution 𝑦 𝑡 = ℎ(𝑡) ∗ 𝑥(𝑡).
❑ Both impulse response and differential equation are used to
describe the input-output relationship of CT LTI systems, but:
➢ The impulse response description applies only to systems that are initially
at rest → can determine the system’s zero−state response only.
➢ Differential equation representation is more flexible: it applies to systems
which are either at rest or with non-zero initial conditions.

❑ Relationship between the responses from each description:

Total response = σ𝑛𝑗=1 𝑐𝑗 𝑒 𝜆𝑗𝑡 + ℎ 𝑡 ∗ 𝑥(𝑡)
zero−input response zero−state response

❑ Both descriptions do NOT provide insight into the system’s

internal mechanism to produce the output from the input.
➢ They are declarative (i.e. tell rules that the system obeys), while the block
diagram is imperative (i.e. tells what the system does with signals).
 Properties of the convolution integral
(proofs: homework)
❑ Commutative:
𝑓1 𝑡 ∗ 𝑓2 𝑡 = 𝑓2 𝑡 ∗ 𝑓1 𝑡

❑ Distributive:
𝑓1 𝑡 ∗ 𝑓2 𝑡 + 𝑓3 𝑡 = 𝑓1 𝑡 ∗ 𝑓2 𝑡 + 𝑓1 𝑡 ∗ 𝑓3 𝑡

❑ Associative:
𝑓1 𝑡 ∗ 𝑓2 𝑡 ∗ 𝑓3 𝑡 = 𝑓1 𝑡 ∗ 𝑓2 𝑡 ∗ 𝑓3 𝑡

❑ The shift property: if 𝑓1 𝑡 ∗ 𝑓2 𝑡 = 𝑐(𝑡) then

𝑓1 𝑡 ∗ 𝑓2 𝑡 − 𝑇 = 𝑐 𝑡 − 𝑇 , and
𝑓1 𝑡 − 𝑇 ∗ 𝑓2 𝑡 = 𝑐 𝑡 − 𝑇 , and
𝑓1 𝑡 − 𝑇1 ∗ 𝑓2 𝑡 − 𝑇2 = 𝑐 𝑡 − 𝑇1 − 𝑇2 .

❑ Convolution with an impulse

𝑓 𝑡 ∗ 𝛿 𝑡 = 𝑓(𝑡)
 Impulse response as an indicator of
system’s properties
∞
❑ The output of a CT LTI system: 𝑦 𝑡 = ‫׬‬−∞ ℎ 𝜏 𝑥 𝑡 − 𝜏 𝑑𝜏
➢ Memoryless:
ℎ 𝑡 = 𝑐𝛿 𝑡 , where c is an arbitrary constant
➢ Causal:
ℎ 𝑡 = 0 for all 𝑡 < 0.
➢ Stable: ℎ 𝑡 is absolutely integrable, i.e.
∞
‫׬‬−∞ ℎ 𝜏 𝑑𝜏 < ∞.

❑ Invertibility and deconvolution

➢ A system is invertible if the input can be recovered from the output except
for a constant scale factor.
➢ An inverve system (of an invertible system) recovers 𝑥 𝑡 from ℎ(𝑡) ∗ 𝑥(𝑡)
→ an inverse system performs deconvolution: ℎ 𝑡 ∗ ℎinv 𝑡 = 𝛿(𝑡).
➢ Real-life application: equalizers are inverse systems that remove signal
distortion in communication modems or noise-cancelling headphones.
 Recap: block diagram representation
❑ Use basic network elements to represent basic operations
performed along the signal flow paths.
𝑡
න 𝑦 𝑡 = න 𝑥 𝑡 𝑑𝑡
−∞

❑ Allows explicit representation of the system.

➢ Impulse responses or differential equations are mathematically compact
while block diagrams are more “natural” as they illustrate signal flow paths.

❑ Note: the block diagram description of a system is not unique.

❑ Example: find a block diagram representation for the system
described by the following differential equation
𝑑
𝑦 𝑡 + 10𝑦 𝑡 = 2𝑥 𝑡
𝑑𝑡
 Example solution
𝑑 1 𝑑 1
❑ Form I: 𝑑𝑡 𝑦 𝑡 + 10𝑦 𝑡 = 2𝑥 𝑡 ⟺ 2 𝑑𝑡 𝑦 𝑡 = 𝑥 𝑡 − 10 𝑦 𝑡
2

𝑑
❑ Form II: 𝑑𝑡 𝑦 𝑡 + 10𝑦 𝑡 = 2𝑥 𝑡 ⟺ 𝑦 𝑡 = 2 ‫ 𝑡𝑑 𝑡 𝑥 ׬‬− 10 ‫𝑡𝑑 𝑡 𝑦 ׬‬
 State variable description of LTI
systems

න න

❑ The state of a system is a minimal set of signals that

represents the system’s entire memory of the past.
➢ Each signal in the selected set is called a state variables & the selection of
the state of a system is not unique.
➢ A convenient way to select state variables are picking the outputs of the
integrators in the block diagram.

❑ An LTI system can be represented by a set of 1st-order

differential equations describing operations in its block diagram
➢ The state equations characterize the dynamics of the system’s state.
➢ The output equation relates the output to the system’s state and the input.
 Example

න න

4. State variable description is often

1. Select state variables: 𝑞1 𝑡 , 𝑞2 𝑡 written in matrix form as follows
➢ The inputs to the integrators are the 𝒒ሶ 𝑡 = 𝑨𝒒 𝑡 + 𝒃𝑥 𝑡
derivatives of the state variables 𝑦 𝑡 = 𝒄𝒒 𝑡 + 𝐷𝑥(𝑡)
2. Write down the state equations 5. The state equation in matrix form
𝑑
𝑞 𝑡 = 2𝑞1 𝑡 − 𝑞2 𝑡 + 𝑥(𝑡) 𝑞ሶ 1 (𝑡) 2 −1 𝑞1 𝑡 1
𝑑𝑡 1 = + 𝑥(𝑡)
𝑑 𝑞ሶ 2 (𝑡) 1 0 𝑞2 𝑡 0
𝑞 𝑡 = 𝑞1 𝑡
𝑑𝑡 2
6. The output equation in matrix form
3. Write down the output equation
𝑞 𝑡
𝑦 𝑡 = 3𝑞1 𝑡 + 𝑞2 𝑡 𝑦 𝑡 = 3 1 1 + 0 𝑥(𝑡)
𝑞2 𝑡
 Exploring concepts with Octave
❑ Simulating CT convolution
Ts=0.001;
x=[ones(1/Ts,1);zeros(3/Ts,1)];
y=[zeros(4/Ts,1);ones(2/Ts,1)];
z=conv(x,y)*Ts;

∗
 Exploring concepts with Octave
❑ Impulse response
➢ E.g. 𝐷2 + 3𝐷 + 2 𝑦 𝑡 = 𝐷𝑓(𝑡)
pkg load control;
sys=tf([1 0],[1 3 2]);
impulse(sys);
 Summary
❑ Modelling of CT LTI systems
➢ Internal description approach: applying natural laws to the internal
structure of the system to obtain LCCODE.
➢ External description (“black box”) approach: measure the response of the
system to the unit impulse response.
➢ Other models: block diagram & state-variable representations.

❑ System analysis: given the system & input, what’s the output?
➢ Solving LCCODE: total response = natural response + forced response =
zero-input response + zero-state response.
∞
➢ Convolution of the input and impulse response: 𝑦 𝑡 = ‫׬‬−∞ 𝑓 𝜏 ℎ 𝑡 − 𝜏 𝑑𝜏.

❑ System’s properties can be inferred from its impulse response

➢ Memoryless if ℎ(𝑡)=𝑐𝛿(𝑡), where 𝑐 is an arbitrary constant.
➢ Causal if ℎ 𝑡 = 0 for all 𝑡 < 0.
∞
➢ Stable if ‫׬‬−∞ ℎ 𝜏 𝑑𝜏 < ∞.

❑ Next lecture: Modelling of DT LTI systems in time domain