Signals and Systems

Lecture 1
Signals
A signal is a function of one or more variables that conveys information about some phenomenon.
𝑓(𝑡1 , 𝑡2 , … , 𝑡𝑛 ) each of {𝑡𝑘 } is called independent variable, and 𝑓 is the dependent variable
Signals are classified based on their dimensionality (how many independent variables they have) and
whether their variables are continuous or discrete.

Independent variable = time

Continuous time (a) vs discrete time (b)

*A discrete-time signal is sometimes called a sequence

Systems
A system is an entity that processes one or more input signals in order to produce one or more output
signals.

Systems are classified based on their number of inputs and outputs (single or multi) and the types of
signals they handle (continuous vs. discrete, analog vs. digital).

A signal can be thought of as a single function mapping inputs to outputs, while a system can be
thought of as an operator mapping an input function to an output function
Properties of Signals
Even and Odd Signals
Even
𝑥 (𝑡) = 𝑥 (−𝑡) for all 𝑡
Odd
𝑥 (𝑡) = −𝑥 (−𝑡) for all 𝑡
Geometrically even signals are symmetric about the y-axis, while odd signals are antisymmetric about
the x-axis.

*notice the
antisymmetry here

Operations between even and odd functions:

- the sum of functions of the same symmetry also has the same symmetry (e.g.
even function + even function = even function)
- The product of functions of the same symmetry is even (e.g. even × even = even
odd × odd = even )
- The product of functions of different symmetry is odd (e.g. even × odd = odd )
Periodic Signals
𝑥 = 𝑥 (𝑡 + 𝑇) for all 𝑡 and some constant 𝑇, 𝑇 > 0
i.e. the function is repeating for all 𝑡 for some constant interval (period) 𝑇

A periodic signal has frequency 𝑓 and an angular frequency 𝜔

1 2𝜋
𝑓= 𝜔 = 2𝜋𝑓 =
𝑇 𝑇

* A signal that is not periodic is said to be aperiodic

The period of a periodic signal is not unique (e.g. sine wave has several periods 2𝜋, 4𝜋, 6𝜋, …). The
smallest period with which a signal is periodic is called the fundamental period.
Operations between periodic functions:

- The sum of two periodic signals 𝑥1 (𝑡), 𝑥2 (𝑡) is periodic if and only if the ratio of their
𝑇1
fundamental periods is a rational number.
𝑇2

Lecture 2
Properties of Signals cont.
Right-Sided Signals
𝑥 (𝑡) = 0 for all 𝑡 < 𝑡0 𝑡0 ∈ ℝ

*if 𝑡0 = 0 then the signal is said to be casual

Left-Sided Signals
𝑥 (𝑡) = 0 for all 𝑡 > 𝑡0 𝑡0 ∈ ℝ
*if 𝑡0 = 0 then the signal is said to be anticasual

Finite Duration and Two-Sided Signals

A signal that is both left and right sides is said to be finite duration or time limited.
A signal that is neither left or right sided is said to be two-sided.
Bounded Signals
|𝑥(𝑡)| ≤ 𝐴 < ∞ for all 𝑡 𝐴 > 0, 𝐴 ∈ ℝ

A bounded signal is finite for all 𝑡 (i.e. it exists only within a certain range) ex. 𝑐𝑜𝑠 , 𝑠𝑖𝑛

Signal Energy and Power

Energy
∞
𝐸 = ∫ |𝑥(𝑡)|2 𝑑𝑡
−∞

A signal with finite energy is said to be an energy signal

*an energy signal must be limited in time
Power
𝑇
1 2
𝑃 = lim ∫ |𝑥 (𝑡)|2 𝑑𝑡
𝑇→∞ 𝑇 −𝑇
2

A signal with nonzero finite average power is said to be a power signal

*a power signal is not limited in time (has infinite length)

*the power of a energy signal is 0

Energy is a transfer of work (i.e. how much work is done), while power is the rate of transfer of
energy (i.e. how quickly the work is done).

This is why in the energy function an integral ∫ is involved (we are measuring the amount of work),
while in the power function the integral is taken over a limit lim∫ (we are measuring the rate of
change in area over the entire domain).
Independent- and Dependent-Variable Transformations
Transformations on a signal can be done on the time axis (x-axis) or the amplitude axis (y-axis). The
transformation itself is either a scaling or a shifting operation or a combination of the two.
Time Scaling

𝑦(𝑡) = 𝑥 (𝑎𝑡) 𝑎∈ℝ

The signal is scaled based on the value of 𝑎

- If |𝑎| < 1, the signal is stretched/expanded

- If |𝑎| > 1, the signal is compressed
- If 𝑎 < 0, the signal is also reflected about the vertical line x-axis
Time Shifting
𝑦(𝑡) = 𝑥 (𝑡 − 𝑏) 𝑏∈ℝ
The shift operation depends on the value of 𝑏

- If 𝑏 > 0, the shift operation is to the right (the signal is delayed)

- If 𝑏 < 0, the shift operation is the left (the signal is advanced)
Combined Time Shifting and Scaling
𝑦(𝑡) = 𝑥 (𝑎𝑡 − 𝑏) 𝑎, 𝑏 ∈ ℝ 𝑎≠0
Time scaling and shifting operations are not commutative (i.e. the order they are performed in
matters)

*intuitively the shifting

operation should be
done first, then the
scaling operation

Amplitude Scaling
𝑦(𝑡) = 𝑎𝑥(𝑡) 𝑎∈ℝ
Amplitude Shifting
𝑦 = 𝑥 (𝑡) + 𝑏 𝑏∈ℝ
Combined Amplitude Scaling and shifting
𝑦(𝑡) = 𝑎𝑥 (𝑡) + 𝑏 𝑎, 𝑏 ∈ ℝ
Amplitude and time transformations have an opposite effect to each other, meaning that for example a
positive time scale 𝑦(𝑡) = 𝑥(2𝑡) would compress the signal along the time axis, while a positive time
scale 𝑦(𝑡) = 2𝑥(𝑡) would stretch the signal along the amplitude axis.

Lecture 3
Elementary Signals
Real Sinusoid
𝑥(𝑡) = 𝐴 cos(𝜔𝑡 + 𝜃) , 𝐴, 𝜔. 𝜃 ∈ ℝ
Real sinusoids are periodic with fundamental period and frequency
2𝜋
𝑇= 𝑓 = |𝜔 |
|𝜔 |

The 𝐴 constant affects the

amplitude/height of the wave.

The constant 𝜔 affects the

frequency of the wave

The constant 𝜃 affects the phase of

the wave (movement along the x-
axis)
Complex Exponentials

𝑥 (𝑡) = 𝐴𝑒 𝜆𝑡 𝐴, 𝜆 ∈ ℂ

Depending on the values of 𝐴, 𝜆 a complex exponential can exhibit different modes of behavior.

Think of how the shape of the

complex exponential would
project on it’s real and
imaginary parts.

*in this example the shape is

oscillating but the shape
could also be exponential or
s-shaped with different
parameters

In the general case, we can rewrite

𝑥 (𝑡) = |𝐴|𝑒 𝜎𝑡 cos(𝜔𝑡 + 𝜃) + 𝑗 |𝐴|sin(𝜔𝑡 + 𝜃)
We can observe 3 modes of behavior based on the value of 𝜎

Illustrations of both real and

imaginary parts

Product of sinusoid and Sinusoid Product of sinusoid and

growth decay

Real Exponentials

𝑥(𝑡) = 𝐴𝑒 𝜆𝑡 𝐴, 𝜆 ∈ ℝ
Has 3 distinct modes of behavior

Growth Decay
 Complex Sinusoids

𝑥(𝑡) = 𝐴𝑒 𝑗𝜔𝑡 𝐴 ∈ ℂ, 𝜔 ∈ ℝ

Im (Imaginary) part
Re + Im 𝑗|𝐴|sin(𝜔𝑡 + 𝜃)
|𝐴|cos(𝜔𝑡 + 𝜃) + 𝑗|𝐴|sin(𝜔𝑡 + 𝜃)

Re (Real) part
|𝐴|cos(𝜔𝑡 + 𝜃)

Based on Euler’s relation

𝑒 𝑖𝜃 = cos 𝜃 + 𝑖 sin 𝜃
The same expression can be formed for complex sinusoids

𝐴𝑒 𝑗𝜃 = 𝐴 cos 𝜃 + 𝑗𝐴 sin 𝜃
As well as an expression for real sinusoids
𝐴 𝑗(𝜔𝑡+𝜃)
A cos(𝜔𝑡 + 𝜃) = [𝑒 + 𝑒 −𝑗(𝜔𝑡+𝜃) ]
2
𝐴
A sin(𝜔𝑡 + 𝜃) = [𝑒 𝑗(𝜔𝑡+𝜃) − 𝑒 −𝑗(𝜔𝑡+𝜃) ]
2𝑗
Unit-Step Function
1 if 𝑡 ≥ 0 *value of 𝑢(0) doesn’t
𝑢(𝑡) = { have to be 1
0 otherwise

Rectangular Function
1 1
1 if − ≤𝑡≤ 1
*values other than ± 2
rect(𝑡) = { 2 2
can be used
0 otherwise

Triangular Function
1 − |𝑥| if |𝑥| < 1
𝑡𝑟𝑖 (𝑡) = {
0 otherwise

Cardinal Sine Function

sin 𝑡
sinc(𝑡) =
𝑡

sinc(0) = 1
Unit-Impulse/Delta Function
𝛿(𝑡) = 0 for 𝑡 ≠ 0
∞
∫ 𝛿(𝑡) 𝑑𝑡 = 1
−∞

The 𝛿 function can be viewed as a limiting case of a rectangular pulse

1 1 1
if − ≤𝑡≤ *the area under 𝑔𝜖 for any value
𝑔𝜖 (𝑡) = { 𝜖 𝜖 𝜖
of 𝜖 will always be 1
0 otherwise ∞
∫ 𝑔𝜖 (𝑡) 𝑑𝑡 = 1
−∞

𝛿(𝑡) = lim 𝑔𝜖 (𝑡)

𝜖→0

𝛿 is a limit of the rectangular function 𝑔𝜖 where the pulse width becomes infinitesimally small and the
pulse height becomes infinitely large, such that the integral of the function results in 1
Delta function properties

- Equivalence
𝑥 (𝑡)𝛿(𝑡 − 𝑡0 ) = 𝑥(𝑡0 )𝛿(𝑡 − 𝑡0 )

𝑥 (𝑡)𝛿 (𝑡) = 𝑥 (0)𝛿(𝑡)

- Sifting
∞
∫ 𝑥 (𝑡)𝛿(𝑡 − 𝑡0 ) 𝑑𝑡 = 𝑥 (𝑡0 )
−∞

The integral of any function 𝑥 with 𝛿 equals the value of 𝑥 at 0

∞
∫ 𝑥(𝑡)𝛿(𝑡) 𝑑𝑡 = 𝑥 (0)
−∞
- Even
𝛿(𝑡) = 𝛿(−𝑡)

- Time Scaling
1
𝛿 (𝑎𝑡) = 𝛿(𝑡)
|𝑎 |
A rectangular step function can be represented using delta functions
1 if 𝑎 ≤ 𝑡 < 𝑎
x(𝑡) = {
0 otherwise
𝑥 (𝑡) = 𝛿(𝑡 − 𝑎) − 𝛿(𝑡 − 𝑏)

This way a formula with multiple cases can be collapsed into a single expression

Continuous-Time Systems
A system is described by

𝑦(𝑡) = 𝒯{𝑥(𝑡)}
𝑥 (𝑡) ⟶𝒯 𝑦(𝑡)

A system is represented in a block diagram as a single block

In block diagrams several systems may be connected to each other, there are two ways this can be
done

𝑦(𝑡) = 𝒯2 {𝒯1 {𝑥(𝑡)}}

𝑦(𝑡) = 𝒯2 {𝑥(𝑡)} + 𝒯1 {𝑥(𝑡)}
Properties of Continuous-Time Systems
Memory and Causality

A system has memory if 𝑦(𝑡) at 𝑡0 is dependent on the value of 𝑥(𝑡) at any other time. i.e. if the
output depends on values of input from past or future
A system is causal if 𝑦(𝑡) at 𝑡0 is dependent on the value of 𝑥(𝑡) at 𝑡 ≤ 𝑡0 . i.e. if the output depends
on past input values
Invertibility
𝒯 −1

𝒯 −1 {𝒯{𝑥 (𝑡)} = 𝑥(𝑡)

A system is invertible if its input can be uniquely determined from its output. i.e. there is only one
possible input for every output (no two inputs give the same output)
Bounded-Input Bounded-Output (BIBO) Stability
A system is BIBO stable if
|𝑥(𝑡)| ≤ 𝐴 < ∞ ⇒ |𝑦(𝑡)| ≤ 𝐴 < ∞ for all 𝑡
In a BIBO stable system every bounded input leads to a bounded output i.e. a finite input always
results in a finite output
Time Invariance

𝑥 (𝑡 − 𝑡0 ) ⟶𝒯 𝑦(𝑡 − 𝑡0 )
In time invariant systems, a time shift in input results in an identical time shift in the output. The
system’s behavior is unchanged with respect to time.
Linearity
Linearity means the relationship between input and output is a linear mapping

𝑎𝑥 (𝑡) ⟶𝒯 𝑎𝑦(𝑡) 𝑎∈ℝ *if these conditions apply for all

values of 𝑎, 𝑥(𝑡), 𝑥′(𝑡) the
𝑥 ′ (𝑡) ⟶𝒯 𝑦 ′ (𝑡) system possesses the
𝑥 (𝑡) + 𝑥 ′ (𝑡) ⟶𝒯 𝑦(𝑡) + 𝑦′(𝑡) superposition property

Additivity and Homogeneity

Additivity
𝒯{𝑥1 (𝑡) + 𝑥2 (𝑡)} = 𝒯{𝑥1 (𝑡)} + 𝒯{𝑥2 (𝑡)}
Homogeneity
𝑎𝒯{𝑥 (𝑡)} = 𝒯{𝑎𝑥(𝑡)} 𝑎∈ℝ
A system that is both additive and homogenous is also linear, using this we can rewrite the condition
for linearity as

𝒯{𝑎1 𝑥1 (𝑡) + 𝑎2 𝑥2 (𝑡)} = 𝑎1 𝒯{𝑥1 (𝑡)} + 𝑎2 𝒯{𝑥2 (𝑡)} ∀ 𝑎𝑖 , 𝑥𝑖

Chapter 2 Problems *page 41 in textbook
2.1 Identify the time and/or amplitude transformations that must be applied to the signal 𝑥(𝑡) in order
to obtain each of the signals specified below. Choose the transformations such that time shifting
precedes time scaling and amplitude scaling precedes amplitude shifting. Be sure to clearly indicate
the order in which the transformations are to be applied.

(a) 𝑦(𝑡) = 𝑥(2𝑡 − 1);

1
(c) 𝑦(𝑡) = 2𝑥(− 2 𝑡 + 1) + 3;
(e) 𝑦(𝑡) = −3𝑥(2[𝑡 − 1]) − 1.
Answer

(a) 𝑥 (𝑡) → 𝑥 (𝑡 − 1) → 𝑥 (2(𝑡) − 1) time shift followed by time scale

or
1
𝑥 (𝑡) → 𝑥 (2𝑡) → 𝑥 (2 (𝑡 − 2)) → 𝑥 (2𝑡 − 1) time scale followed by time shift
1 1
(c) 𝑥 (𝑡) → 𝑥 (𝑡 + 1) → 𝑥 (− 2 𝑡 + 1) → 2𝑥 (− 2 𝑡 + 1) + 3

time shift followed by time scale followed by amplitude operations

(e) 𝑥 (𝑡) → 𝑥 (2𝑡) → 𝑥 (2[𝑡 − 1]) → −3𝑥 (2[𝑡 − 1]) − 1
time scale followed by time shift followed by amplitude operations

2.3 Given the signal x(t) shown in the figure below, plot and label each of the following signals:

(a) 𝑥(𝑡 − 1); (d) 𝑥(2𝑡 + 1);

(b) 𝑥(2𝑡); (e)
1 1
𝑥 (− 𝑡 + 1) −
1
4 2 2
(c) 𝑥(−𝑡);
Answer
2.4 Determine whether each of the following functions is even, odd, or neither even nor odd:
(a) 𝑥(𝑡) = 𝑡 3 ;
(b) 𝑥(𝑡) = 𝑡 3 |𝑡|;
(c) 𝑥(𝑡) = |𝑡 3 |;
(d) 𝑥(𝑡) = (cos 2𝜋𝑡)(sin 2𝜋𝑡 );
(e) 𝑥(𝑡) = 𝑒 𝑗2𝜋𝑡 ; and
1
(f) 𝑥 (𝑡) = [𝑒 𝑡 + 𝑒 −𝑡 ].
2

Answer
(a) odd
𝑡 3 = −(−𝑡)3
(b) odd
𝑡 3 is odd, |𝑡| is even
even × odd = odd
(c) even
|𝑥| 𝑥∈ℝ is even ∴ |𝑡 3 | is even
(d) odd
sin 𝑥 𝑥 ∈ ℝ is odd, cos 𝑥 𝑥 ∈ ℝ is even
even × odd = odd
(e) neither even nor odd
(f) even
1 𝑡 1
[𝑒 + 𝑒 −𝑡 ] = [𝑒 −𝑡 + 𝑒 𝑡 ]
2 2
2.9 Determine whether each of the signals given below is periodic. If the signal is periodic, find its
fundamental period.
(a) 𝑥 (𝑡) = cos 2𝜋𝑡 + sin 5𝑡 ;
𝜋 2
(b) 𝑥(𝑡) = [cos (4𝑡 − 3 )] ;

(c) 𝑥(𝑡) = 𝑒 𝑗2𝜋𝑡 + 𝑒 𝑗3𝜋𝑡 ; and

(d) 𝑥(𝑡) = 1 + cos 2𝑡 + 𝑒 𝑗5𝑡 .

Answer
2𝜋
(a) 𝑇1 of cos 2𝜋𝑡 = 1 𝑇2 of sin 5𝑡 = 5
𝑇1 1 5
= =
𝑇2 2𝜋 2𝜋
5
5
∉ ℚ (ℚ = the set of rational numbers)
2𝜋
∴ cos 2𝜋𝑡 + sin 5𝑡 is not periodic
(b) 𝑇 of cos(𝑡)2 = 𝜋
𝜋
𝑇 of cos(4𝑡 + 𝜃)2 =
4
𝜋 2 𝜋
[cos (4𝑡 − )] is periodic with fundamental period 𝑇 =
3 4
2.10 Evaluate the following integrals:
∞ 𝜋
(a) ∫−∞ sin (2𝑡 + ) 𝛿(𝑡) 𝑑𝑡;
4
𝑡
(b) ∫−∞[cos 𝜏]𝛿 (𝜏 + 𝜋) 𝑑𝜏;
𝑡
(e)∫−∞ 𝛿 (𝜏) 𝑑𝜏.

Answer
(a) using the rule:
∞
∫ 𝑥 (𝑡)𝛿 (𝑡) 𝑑𝑡 = 𝑥 (0)
−∞
∞
𝜋 𝜋
∫ sin (2𝑡 + ) 𝛿(𝑡) 𝑑𝑡 = sin (2(0) + )
−∞ 4 4

∞
𝜋 𝜋 1
∫ sin (2𝑡 + ) 𝛿 (𝑡) 𝑑𝑡 = sin =
−∞ 4 4 √2
(b) using the rule:
∞
∫ 𝑥 (𝑡)𝛿(𝑡 − 𝑡0 ) 𝑑𝑡 = 𝑥 (𝑡0 )
−∞
𝑡
∫ [cos 𝜏]𝛿(𝜏 + 𝜋) 𝑑𝜏 = cos 𝜋
−∞
𝑡
∫ [cos 𝜏]𝛿 (𝜏 + 𝜋) 𝑑𝜏 = −1 𝑡 > −𝜋
−∞
(e) using the definition of 𝛿
1 1 1
if − ≤𝑡≤
𝑔𝜖 (𝑡) = { 𝜖 𝜖 𝜖
0 otherwise
𝛿(𝑡) = lim 𝑔𝜖 (𝑡)
𝜖→0

And the rule:

∞
∫ 𝛿(𝑡) 𝑑𝑡 = 1
−∞
if 𝜏 > 0 ∫ 𝛿 would equal 0 since it wouldn't contain the 'impulse'
if 𝜏 ≤ 0 ∫ 𝛿 would equal 1
𝑡 0 if 𝜏 > 0
∫ 𝛿(𝜏) 𝑑𝜏 = {
−∞ 1 if 𝜏 ≤ 0
2.12 Determine whether the system with input 𝑥(𝑡) and output 𝑦(𝑡) defined by each of the following
equations is linear:
𝑡+1
(a) 𝑦(𝑡) = ∫𝑡−1 𝑥 (𝜏) 𝑑𝜏;
(d) 𝑦(𝑡) = 𝑥 2 (𝑡).
Answer
(a) using the definition for linearity:

𝒯{𝑎1 𝑥1 (𝑡) + 𝑎2 𝑥2 (𝑡)} = 𝑎1 𝒯{𝑥1 (𝑡)} + 𝑎2 𝒯{𝑥2 (𝑡)} ∀ 𝑎𝑖 , 𝑥𝑖

𝑡+1
𝑦1 (𝑡) = ∫ 𝑥1 (𝜏) 𝑑𝜏
𝑡−1
𝑡+1
𝑦2 (𝑡) = ∫ 𝑥2 (𝜏) 𝑑𝜏
𝑡−1
𝑡+1 𝑡+1
𝑦3 (𝑡) = ∫ 𝑎1 𝑥1 (𝜏) 𝑑𝜏 + ∫ 𝑎2 𝑥2 (𝜏) 𝑑𝜏
𝑡−1 𝑡−1
𝑡+1 𝑡+1
= 𝑎1 ∫ 𝑥1 (𝜏) 𝑑𝜏 + 𝑎2 ∫ 𝑥2 (𝜏) 𝑑𝜏
𝑡−1 𝑡−1
= 𝑎1 𝑦1 (𝑡) + 𝑎2 𝑦2 (𝑡)
𝑡+1
∴ 𝑦(𝑡) = ∫ 𝑥 (𝜏) 𝑑𝜏 is linear
𝑡−1

(d) using the definition for linearity:

𝒯{𝑎1 𝑥1 (𝑡) + 𝑎2 𝑥2 (𝑡)} = 𝑎1 𝒯{𝑥1 (𝑡)} + 𝑎2 𝒯{𝑥2 (𝑡)} ∀ 𝑎𝑖 , 𝑥𝑖

𝑦1 (𝑡) = 𝑥12 (𝑡)

𝑦2 (𝑡) = 𝑥22 (𝑡)
𝑦3 (𝑡) = [𝑎1 𝑥1 (𝑡) + 𝑎2 𝑥2 (𝑡)]2
= 𝑎12 𝑥12 (𝑡) + 2𝑎1 𝑎2 𝑥1 (𝑡)𝑥2 (𝑡) + 𝑎22 𝑥22 (𝑡)
≠ 𝑎1 𝑦1 (𝑡) + 𝑎2 𝑦2 (𝑡)
∴ 𝑦(𝑡) = 𝑥 2 (𝑡) is not linear
2.16 Determine whether the system with input x(t) and output y(t) defined by each of the equations
given below is BIBO stable.
𝑡+1
(a) 𝑦(𝑡) = ∫𝑡 𝑥 (𝜏) 𝑑𝜏;
1
(c) 𝑦(𝑡) = 𝑥(𝑡).

Answer
(a) using the definition for stability:
|𝑥(𝑡)| ≤ 𝐴 < ∞ ⇒ |𝑦(𝑡)| ≤ 𝐴 < ∞ for all 𝑡
suppose |𝑥(𝜏)| ≤ 𝐴 < ∞
𝑡+1
|∫ 𝑥 (𝜏) 𝑑𝜏| ≤ 𝐴 < ∞
𝑡
 𝑏 𝑏
since |∫ 𝑓(𝑥) 𝑑𝑥 | ≤ ∫ |𝑓(𝑥)| 𝑑𝑥
𝑎 𝑎
𝑡+1 𝑡+1
|∫ 𝑥 (𝜏) 𝑑𝜏| ≤ ∫ |𝑥 (𝜏)| 𝑑𝜏 < ∞
𝑡 𝑡

𝑡+1
∴ 𝑦(𝑡) = ∫ 𝑥 (𝜏) 𝑑𝜏 is BIBO stable
𝑡

(c) using the definition for stability:

|𝑥(𝑡)| ≤ 𝐴 < ∞ ⇒ |𝑦(𝑡)| ≤ 𝐴 < ∞ for all 𝑡

suppose |𝑥(𝑡)| ≤ 𝐴 < ∞

consider 𝑥 (𝑡) = 0
1
𝑦(𝑡) = =∞
0
1
∴ 𝑦(𝑡) = is not BIBO stable
𝑥(𝑡)

2.19 Given the signals 𝑥1 (𝑡) and 𝑥2 (𝑡) shown in the figures below, express 𝑥2 (𝑡) in terms of 𝑥1 (𝑡).

Answer
𝑥2 (𝑡) = 𝑥1 (−4𝑡 − 1)

2.21 For the system shown in the figure below, express the output 𝑦(𝑡) in terms of the input 𝑥(𝑡) and
the transformations 𝑇1 , 𝑇2 , . . . , 𝑇5 .

Answer

𝑦(𝑡) = 𝑇1 (𝑇2 (𝑇3 ) + 𝑇4 )(𝑇2 +𝑇5 )

Lecture 4
Continuous-Time Convolution
The convolutions of 2 functions is defined as
∞
𝑥 (𝑡) ∗ ℎ(𝑡) = ∫ 𝑥 (𝜏)ℎ(𝑡 − 𝜏) 𝑑𝜏
−∞

Computing Convolutions
∞
𝑦(𝑡) = 𝑥 (𝑡) ∗ ℎ(𝑡) = ∫ 𝑤1 (𝜏)𝑑𝜏
−∞

𝑤1 (𝜏) = 𝑥 (𝜏)ℎ(𝑡 − 𝜏)
1. Plot 𝑥 (𝜏) and ℎ(𝑡 − 𝜏) as a function of 𝜏
2. Consider an arbitrarily large negative value for t. This will result in ℎ(𝑡 − 𝜏) being shifted
very far to the left
3. Write a mathematical expression for 𝑤1 (𝜏)
4. Increase 𝑡 until 𝑤1 (𝜏) changes form, keeping track of the interval where the expression was
valid
5. Repeat steps 3 and 4 until 𝑡 is an arbitrarily large positive number
6. For each interval defined, integrate 𝑤1 (𝜏) to find an expression for 𝑦(𝑡) in that interval
7. Combine the intervals to obtain a value for 𝑦(𝑡)

𝑥(𝑡) ∗ ℎ(𝑡)

Convolutions are commutative, associative, and distributive

Periodic Convolutions
Periodic convolution is the convolution of two periodic functions that have the same period.

𝑥(𝑡) ⊛ ℎ(𝑡) = ∫ 𝑥(𝜏)ℎ(1 − 𝜏) 𝑑𝜏

𝑇

If we express the periodic functions as summations of aperiodic components

∞

𝑥(𝑡) ≜ ∑ 𝑥0 (𝑡 − 𝑘𝑇)
𝑘=−∞

We can define periodic convolution as

𝑥(𝑡) ⊛ ℎ(𝑡) = 𝑥0 ∗ ℎ(𝑡)
Representation Using Impulses
∞
𝑥 (𝑡) = ∫ 𝑥 (𝜏)𝛿 (𝑡 − 𝜏) 𝑑𝜏 = 𝑥 (𝑡) ∗ 𝛿(𝑡)
−∞

Using this property we can express any signal 𝑥(𝑡) using an expression containing the impulse/delta
function

LTI Systems
Impulse Response
Impulse response is a systems response to the unit-impulse function as input
ℎ(𝑡) = 𝒯{𝛿(𝑡)]
The behavior of an LTI system is completely characterized by its impulse response, if the impulse
response of a system is known, we van determine the response of the system to any input.
𝑦(𝑡) = 𝒯{𝑥(𝑡)}
𝑦(𝑡) = 𝑥 (𝑡) ∗ ℎ(𝑡)
Step Response
Step response is the response of a system to a unit-step function as input
𝑠(𝑡) = 𝑢 (𝑡) ∗ ℎ(𝑡)
𝑡
𝑠(𝑡) = ∫ ℎ(𝜏) 𝑑𝜏
−∞

The derivative of the step response gives us the impulse response

𝑑𝑠(𝑡)
= ℎ(𝑡)
𝑑𝑡
Block Diagram Representation
LTI systems are represented using their impulse responses