The Fourier Transform

Lecture 5
 ?Why Fourier Transform
f (t )
ω0, ω1, ω2……

F(ω)
numbers
t Correlation
similarity measuring(
)device
0 ω0 ω1 ω2 ω(Hz)
“spectrum analyzer”
(DC)

 For a continuous-time, periodic signal, the frequency content

(spectrum) is displayed by computing the Fourier series.
 The Fourier series for periodic functions of time produce
discrete line spectra with non-zero values only at specific
frequencies referred to as harmonics.
 ?Why Fourier Transform
ω0, ω1, ω2……

f (t )
F(ω)
Correlation
similarity-measuring(
)device

t “spectrum analyzer” 0 ω (Hz)

 The spectrum of a non-periodic, continuous-time

signal is displayed by computing the continuous-
time Fourier transform (FT).
 The frequency spectra for these functions are
continuous.
?Why Fourier Transform
 The Fourier transform is a mathematical
autocorrelation function (similarity-
measuring) that compares an input time-
domain signal to all frequencies and
generates the complex equivalent of the
level and phase for each frequency,
creating a frequency-domain
representation of the input time-domain
signal.
?Why Fourier Transform
 In general, the Laplace and Fourier
transforms are used to convert certain difficult
mathematical operations into easier ones:
 Applied to linear differential equations,
transform them into algebraic equations.

 When we do Fourier transform, we convert a

signal into its frequency components,
so that we can have a better analysis of that
signal such as noise removal….
Definition
 For a non-periodic signal, the Fourier transform or the
Fourier integral is defined as

 The Inverse Fourier transform is defined as

 The following notation is used to express the Fourier

transform and its inverse
Definition
 The Fourier transform is, in general, a complex
function. It can be expressed as the sum of its
real and imaginary components, or in exponential
form as:

 |F (ω)|: tells us how much the frequency component

ω is present over the signal,
 ɸ(ω): tells us how much the contribution of this
frequency component is phase-shifted.
 What is Phase Shift?

If two identical waveforms are not aligned to each

other along the x-axis, then there is a phase shift.
Properties and Theorems of the Fourier
Transform
Special Forms of the Fourier
Transform
 The time function is, in general, complex. It
can be expressed as the sum of its real and
imaginary parts

 The real and imaginary parts of F(ω) are

 The real and imaginary parts of f(t) are

Special Forms of the Fourier
Transform

 The table shows that if f) t ) is real: F(ω) is complex.

The real part of F(ω) is even, and the imaginary part is
odd.

 It follows that:
Special Forms of the Fourier
Transform
 The previous table states that:
Fourier Transform Pairs of
Common Functions
The delta function:

Proof:
The sifting theorem of the delta function states that:

At t0 = 0

From Fourier transform definition:

Fourier Transform Pairs of
Common Functions
The time-shifted delta function:

Proof:
Similarly, from the sifting theorem

Therefore,


F     t  t 0 e  j t
dt e  j t 0


Fourier Transform Pairs of
Common Functions
The unity function:

Proof:
Recalling that:

Applying the symmetry property:

Since δ(-ω)=δ(ω) (an even function), therefore:

Also, using the frequency shifting property:

Fourier Transform Pairs of
Common Functions
The cosine function:

Proof:
Since and e j 0t
 2   0 

By adding both terms and dividing by 2, we get the transform.

 Fourier Transform Pairs of
Common Functions
The sine function:

sinω0t = (ejω0t−e−jω0t)/2j  −jπ[δ(ω−ω0)−δ(ω+ω0)]

Proof:
Similar to the proof of the cosine function.
Fourier Transform Pairs of
Common Functions
The signum function:

Proof:
We express the sgn(t) as an exponential that approaches a
limit

Therefore,

-
Fourier Transform Pairs of
Common Functions
The unit step function:

Proof:
The sgn(t) can be expressed as

Therefore, u0(t) is expressed as

Recalling that and

Therefore,
Fourier Transform Pairs of
Common Functions
One-sided exponential:

Proof:
From the Fourier transform of the unit step function,

and the frequency shifting property,

we reach the above relation.

Fourier Transform Pairs of
Common Functions
One-sided cosine function:

Proof:
The cosine function is first expressed as:

Using

and
-
We reach the above relation.
Fourier Transform Pairs of
Common Functions
One-sided sine function:
-

Proof:
The sine function is expressed as:

The proof then continues similar to the one-sided cosine

function.
Fourier Transform Pairs of Common
Functions w
Summary: Fourier Transform of
Common Functions
 The delta Function

 The unity function

 Summary: Fourier Transform of
Common Functions
 The cosine and sine functions
real and even real and even

real and odd Imaginary and

odd


Summary: Fourier Transform of
Common Functions
 The signum function

 The unit-step function

Fourier Transform of
Common Waveforms
Derive the Fourier transform of the rectangular pulse

Solution:

sinc function
 Fourier Transform of
Common Waveforms
Derive the Fourier transform of the cosine-modulated
pulse
Solution:
Using the frequency shifting property:

since and

Therefore, the transform of the cosine and sine-modulated

signals is

Thus
Fourier Transform of
Common Waveforms
Derive the Fourier transform of the rectangular pulse

Solution:

Using the time-shifting property:

And from the previous example:

 Fourier Transform of
Common Waveforms
Derive the Fourier transform of the function

f2
Solution:

f1
This is precisely the sum of the waveforms of the two previous
examples
 Fourier Transform of
Common Waveforms
Derive the Fourier transform of the periodic train
function

Solution:
 Fourier transform of a periodic train of equidistant delta
functions in the time domain, is a periodic train of equally
spaced delta functions in the frequency domain.

where ωo = 1/T
Finding the Fourier Transform
from Laplace Transform
 The Fourier transform is nothing but
the Laplace transform evaluated on
the imaginary axis,
i.e. s = σ + j ω  At σ = 0 s = jω.
Finding the Fourier Transform
from Laplace Transform
 Non-integrable functions are difficult to compute
its Fourier transform
e-σt e-σt

sin function
Unit-step

 F(s) is equivalent to computing Fourier transform

of f(t)e- σt u(t).
 Thus, Laplace transform is a generalized
Fourier transform for one-sided, exponentially
weighted functions.
 Finding the Fourier Transform
from Laplace Transform
 If a time function f(t) is zero for t ≤ 0, we can obtain
its Fourier transform from the one-sided Laplace
transform by substitution of s with jω .

 If the time function f(t) = 0 for t ≥ 0, and f(t) ≠ 0 for t

< 0 , we use the substitution

**Condition imaginary axis (σ = 0) lies within the Laplace transform region of

convergence (ROC).
Finding the Fourier Transform
from Laplace Transform
Example:

Solution:
Finding the Fourier Transform
from Laplace Transform
Example:

Solution:

t
|𝒕 |= {𝒕 ,𝒕≥𝟎
− 𝒕 , 𝒕 <𝟎 }
Finding the Fourier Transform
from Laplace Transform
Solution, cont.

f (t) f (-t)
The System Function
 The system function and the impulse response
form the Fourier transform pair

 If we know the impulse response h(t), we can

compute the response g(t) of any input f(t), by
multiplication of the Fourier transforms F(ω)
and H(ω) to obtain G(ω) . Then, we take the
Inverse Fourier transform of G(ω) to obtain the
response g(t).
The System Function

Example:
 In a linear system with an impulse response

h(t), use the Fourier transform method to

compute the response g(t) when the input f(t) is
as shown.
The System Function
Solution:
The System Function
.Solution, cont
 The System Function
Solution, cont.
 The System Function
Solution, cont.

x(t) y(t)
LTI
x(t - a) y(t - a)

Replacing t with (t-3)

The overall response is

The System Function
Example:
For the linear circuit shown, the input-output
relation is

Use the Fourier transform method to find the

output voltage.
 The System Function

Solution:
Taking the Fourier transform of both sides of the equation and
recalling that

Therefore

The Fourier transform of the input is

The System Function

Solution, cont.
The Fourier transform of the output voltage is

By partial fractions expansion

The output response is

Application: EMG Signal
Analysis
 Any real continuous signal can be
expressed as an infinite sum of
weighted sinusoids. This set of
sinusoids is called a Fourier Series.
Application: EMG Signal
Analysis
 Consider a surface recorded MUAP (Motor Unit
Action Potential), shown in the red trace at the
bottom of figure.
 This trace can be decomposed into a series of
sinusoids derived from the Fourier Series.
 The first 10 sinusoids from the resulting series
are shown. The summation of these 10
sinusoids is depicted in the blue trace at the
bottom of the figure.
 It is clear from comparing this trace to the
original red one that a faithful recreation of the
signal can be made with only 10 sinusoids.
Application: EMG Signal
Analysis
Application: EMG Signal
Analysis
 This figure describes the
same signal in the
“Frequency Domain”,
since it describes the
amplitude of the
frequencies contained in
it.

 This type of graph is commonly called a “Frequency

Spectrum” or “Power Spectrum”.