EEE332

NOTE 2
Signals
• Introduction
• Analogy between vectors and signals
• Orthogonal functions
• Fourier series of periodic function
• Fourier transform
• Convolution
 Introduction
• A signal is a formal description of a phenomenon evolving over time and
space
• It describes a time varying physical phenomenon which is intended to convey
information
• So it is a function of one or more variables that conveys information about some
(usually physical) phenomena
• Mathematically, signals are represented as functions of one or more
independent variables
• 𝑥 𝑡 = 𝐴 sin 𝜔𝑡 + 𝜙

• Categories:
• Analog Signal
• Digital Signal
Analog Signals
• An analog signal is a smoothly and
continuously varying voltage or current
• Examples:
• A sine wave is a single-frequency analog
signal
• Voice and video voltages
• Vary in accordance with the sound or light
variations that are analogous to the
information being transmitted
Digital Signals
• Digital signals, in contrast to
analog signals, do not vary
continuously, but change in steps
or in discrete increments.
• Most digital signals use binary or
two-state codes
• Examples:
• Telegraph (Morse code)
• Continuous-wave (CW) code
• Serial binary code
Analog and Digital Signals in Communication
• Digital Signals provide a very convenient abstract representation
which is both simple and powerful, yet the real world is best
modelled by the analog paradigm.
• Many transmissions are of signals that originate in digital form but
that must be converted to analog form to match the transmission
medium
• Analog signals can also be transmitted digitally
• Communication system is an example of coordinated interplay
between the digital and the analog world
Analogy between vectors and signals
• There is a perfect analogy between signals and vectors
• A vector can be represented as a sum of its components in various
ways
• A signal can as well be represented as a sum of its components in
various ways
Components of a Vector
• A vector is usually specified by its magnitude and direction.
• Given two vectors 𝑽1 and 𝑽2 , let the component of 𝑽1 along 𝑽2 be
denoted by C12 𝑽2 .
• To determine this component, a perpendicular line is drawn from the end
of 𝑽1 until it gets to vector 𝑽2

• The vector 𝑽1 can be expressed in terms of vector 𝑽2 as:

𝑽1 = C12 𝑽2 + 𝑽𝑒
• Where 𝑽𝑒 is the error vector
• There are many other possible ways of expressing 𝑽1 in terms of 𝑽2
and each possibility comes with different error vector

• 𝑽1 = C1 𝑽2 + 𝑽𝑒1 ; 𝑽1 = C2 𝑽2 + 𝑽𝑒2
• If the error vector tends to zero, the vector 𝑽1 approximates to C12 𝑽2
𝑽1 ≃ C12 𝑽2
• 𝑽𝒆 = 𝑽1 − C12 𝑽2
• To minimize 𝑽𝑒 , the factor C12 must increase
• The length of the component 𝑽1 along 𝑽2 is 𝑽1 cos 𝜃
• ∴ C12 𝑽2 = 𝑽1 cos 𝜃
• Multiplying both sides by 𝑽2 :
• C12 𝑽2 2 = 𝑽1 𝑽2 cos 𝜃 = 𝑽1 ∙ 𝑽2 (dot product)
𝑽1 ∙𝑽2
• C12 =
𝑽2 2

• NOTE: Similar analogy could be drawn for the signals using two orthogonal signals
Dot (Inner or scalar) Product of two vectors
• Given two vectors x and y, their dot product is given as:

𝐱 ∙ 𝐲 = 𝐱 𝒚 cos 𝜃
• Where 𝜃 is the angle between vectors x and y
• Similarly,
𝐱 ∙ 𝐱 = 𝐱 𝐱 cos 0

2
∴𝐱∙𝐱= 𝐱
• When 𝐱 ∙ 𝐲 = 𝟎, it means that cos 𝜃 = 0 which implies that θ = 90°
• Thus, x and y are said to be orthogonal
Inner Product of two signals
• Given two signals 𝑦(𝑡) and 𝑥(𝑡), their inner product is defined as the
area under the product of the two signals.
∞
𝑦 𝑡 ,𝑥 𝑡 = න 𝑦 𝑡 𝑥 𝑡 𝑑𝑡
−∞
• For complex signals,
∞
𝑦 𝑡 ,𝑥 𝑡 = න 𝑦 𝑡 𝑥 ∗ 𝑡 𝑑𝑡
−∞
• Two signals are said to be orthogonal if
∞
𝑦 𝑡 ,𝑥 𝑡 = න 𝑦 𝑡 𝑥 𝑡 𝑑𝑡 = 0
−∞
 Energy of orthogonal signals
• If vectors x and y are orthogonal, and if z = x + y, then from Pythagorean theorem,
2 2 2
𝒛 = 𝐱 + 𝒚
• Similarly, If signals 𝑦(𝑡) and 𝑥 𝑡 are orthogonal, and 𝑧 𝑡 = 𝑥 𝑡 + 𝑦(𝑡), then

𝐸𝑧 = 𝐸𝑥 + 𝐸𝑦
• Where 𝐸𝑧 , 𝐸𝑥 , 𝑎𝑛𝑑 𝐸𝑦 are energy signals
Correlation
• Given two vectors x and y, the correlation measure is given as:
𝐱∙𝐲
𝒄𝒏 = cos 𝜃 =
𝐱 𝒚
• Obviously, −1 ≤ 𝑐𝑛 ≤ 1
• Similarly, for energy signals,
∞
𝟏
𝒄𝒏 = න 𝑦 𝑡 𝑥 𝑡 𝑑𝑡
𝐸𝑥 𝐸𝑦 −∞

• Also, −1 ≤ 𝑐𝑛 ≤ 1
Line Spectra and Fourier Series
• A signal physically exists in the time domain, but it could also be
represented in the frequency domain where it is viewed as consisting
of sinusoidal components at various frequencies
• This frequency-domain description is called the spectrum
• Spectral analysis, using the Fourier series and transform, is one of the
fundamental methods of communication engineering
• It allows an entire class of signals with similar properties in the frequency
domain to be treated as a whole, rather than getting bogged down in detailed
time-domain analysis of individual signals
• It provides valuable insight for communication system design when coupled
with the frequency-response characteristics of filters and other system
components
Phasors
• Sinusoids are, by convention, expressed in terms of the cosine function:
𝑣 𝑡 = 𝐴 cos 𝑤𝑜 𝑡 + 𝜙
• Where 𝐴 is the amplitude (peak value), 𝑤𝑜 𝑡 is the radian frequency and 𝜙 is the phase angle
which represents the fact that the peak has been shifted away from the time origin
2𝜋
• Period, 𝑇𝑜 =
𝑤𝑜
1 𝑤𝑜
• Cyclical frequency, 𝑓𝑜 ≜ =
𝑇𝑜 2𝜋
• Phasor representation of a sinusoidal signal is given by Euler’s theorem:
𝑒 ±𝑗𝜃 = cos 𝜃 ± 𝑗 sin 𝜃
• Substituting 𝜃 with 𝑤𝑜 𝑡 + 𝜙, we obtain
𝑒 ±𝑗 𝑤𝑜 𝑡+𝜙 = cos 𝑤𝑜 𝑡 + 𝜙 ± 𝑗 sin 𝑤𝑜 𝑡 + 𝜙
• Writing the sinusoid as the real part of a complex exponential we have:
𝐴 cos 𝑤𝑜 𝑡 + 𝜙 = 𝐴 𝑅𝑒 𝑒 ±𝑗 𝑤𝑜 𝑡+𝜙
= 𝑅𝑒 𝐴𝑒 𝑗𝜙 𝑒 𝑗𝑤𝑜𝑡
Periodic Signals
• Sinusoids and phasors are members of the general class of periodic
signals
• These signals obey the relationship
𝑣 𝑡 ± 𝑚𝑇𝑜 = 𝑣 𝑡 −∞<𝑡 <∞
• Where 𝑚 is any integer and 𝑇𝑜 is the fundamental signal period.
• Shifting the signal by an integer number of periods to the left or right
leaves the waveform unchanged
• Consequently, a periodic signal is fully described by specifying its
behavior over any one period
• The frequency-domain representation of a periodic signal is a line
spectrum obtained by Fourier series expansion which requires that
the signal must have finite average power.
Average value of a time function
• Given any time function v(t), its average value over all time is defined
as:
1 𝑇 Τ2
𝑣(𝑡) ≜ lim න 𝑣 𝑡 𝑑𝑡
𝑇→∞ 𝑇 −𝑇Τ2

• If the function is a periodic signal, the equation above reduces to the

average over any interval of duration 𝑇𝑜

1 𝑡1 +𝑇𝑜 1
𝑣(𝑡) = න 𝑣 𝑡 𝑑𝑡 = න 𝑣 𝑡 𝑑𝑡
𝑇𝑜 𝑡1 𝑇𝑜 𝑇𝑜
• Where ‫𝑇׬‬ implies integration from any time 𝑡1 to 𝑡1 + 𝑇𝑜
𝑜
• Given that 𝑣(𝑡) is the voltage across a resistance 𝑅, the instantaneous
power is given by:
𝑖 𝑡 𝑣 𝑡 = 𝑖 𝑡 × 𝑖 𝑡 𝑅 = 𝑖2 𝑡 𝑅
• Or
𝑣(𝑡) 𝑣 2 (𝑡)
𝑖 𝑡 𝑣 𝑡 = ×𝑣 𝑡 =
𝑅 𝑅
• Normalizing the power by assuming 𝑅 = 1Ω, 𝑝 = 𝑖(𝑡) 2 = 𝑣(𝑡) 2

• Average Power is given by:

2
1
𝑝 ≜ 𝑣(𝑡) = න 𝑣(𝑡) 2 𝑑𝑡
𝑇𝑜 𝑇𝑜
• The average value of a power signal may be positive, negative, or zero.
Fourier Theory
• The mathematical analysis of the modulation and multiplexing
methods used in communication systems assumes sine wave carriers
and information signals
• This simplifies the analysis and makes operation predictable
• In real world, not all information signals are sinusoidal
• They are typically complex signals that are essentially composites of sine
waves of many frequencies and amplitudes
• But can take on an infinite number of waves such as rectangular, triangular,
sawtooth waves, and other nonsinusoidal forms.
• Such signals require that a non–sine wave approach be taken to determine
the characteristics and performance of any communication circuit or system.
• Fourier analysis provides a means to accurately analyze the content of
most complex nonsinusoidal signals
 Fourier Theorem
• Fourier Theorem states that any practical periodic function of frequency
𝑤𝑜 can be expressed as an infinite sum of sine or cosine functions that are
integral multiples of 𝑤𝑜 .

• Where:
• 𝑎0 is the dc component or the average value of 𝑓 𝑡
• sin 𝑛𝜔0 and cos 𝑛𝜔0 are the 𝑛𝑡ℎ harmonics
• 𝑎𝑛 and 𝑏𝑛 are the amplitudes of the sinusoids in the ac component.
Sum of Sinusoids
• Given a continuous-time signal:
𝑥 𝑡 = 𝐴1 cos(𝑡) + 𝐴2 cos 2𝑡 + 𝜋Τ6 + 𝐴3 cos 8𝑡 + 𝜋Τ3
• The signal has
• 3 Frequency components at 1,2, and 8 rad/sec
• 3 Amplitudes: 𝐴1 , 𝐴2 , 𝑎𝑛𝑑 𝐴3
𝜋 𝜋
• 3 Phases: 0, , and
6 3
• The shape of the signal x(t) depends on the relative magnitudes of
the frequency components, specified in terms of the amplitudes 𝐴1 ,
𝐴2 , and 𝐴3
Fourier Series
• The Fourier series of a periodic function 𝑓(𝑡) is a representation that
resolves 𝑓(𝑡) into a dc component and an ac component comprising
an infinite series of harmonic sinusoids.
• Fourier series exists if Dirichlet’s conditions are satisfied
• 𝑓 𝑡 must be single-valued everywhere
• 𝑓 𝑡 must be absolutely integrable over the interval T seconds
𝑡 +𝑇
‫ 𝑡׬‬0 𝑓(𝑡) 𝑑𝑡 < ∞ for any 𝑡0
0
• 𝑓 𝑡 must have finite number of discontinuities in any one period
• 𝑓 𝑡 must remain finite and must have only finite number of maxima and
minima in any one period
Determination of Fourier Coefficients
• To find 𝑎0 integrate the expression for Fourier series:
𝑇 𝑇 ∞

න 𝑓 𝑡 𝑑𝑡 = න 𝑎0 + ෍ 𝑎𝑛 cos 𝑛𝜔0 t + 𝑏𝑛 sin 𝑛𝜔0 t 𝑑𝑡

0 0 𝑛=1

𝑇 𝑇 𝑇
= ‫׬‬0 𝑎0 𝑑𝑡 + σ∞
𝑛=1 ‫׬‬0 𝑎𝑛 cos 𝑛𝜔0 t 𝑑𝑡 + ‫׬‬0 𝑏𝑛 sin 𝑛𝜔0 t 𝑑𝑡 𝑑𝑡

𝑇 𝑇
∴ න 𝑓 𝑡 𝑑𝑡 = න 𝑎0 𝑑𝑡 = 𝑎0 𝑇
0 0

1 𝑇
∴ 𝑎0 = න 𝑓 𝑡 𝑑𝑡
𝑇 0
 Fourier Coefficient: 𝑎𝑛
• To find 𝑎𝑛 multiply both sides of the expression for Fourier series by cos 𝑚𝜔0 t and
then integrate over one period :

𝑇 𝑇 ∞

න 𝑓 𝑡 cos 𝑚𝜔0 t 𝑑𝑡 = න 𝑎0 + ෍ 𝑎𝑛 cos 𝑛𝜔0 t + 𝑏𝑛 sin 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡

0 0 𝑛=1

𝑇 𝑇 𝑇
= ‫׬‬0 𝑎0 cos 𝑚𝜔0 t 𝑑𝑡 + σ∞
𝑛=1 ‫׬‬0 𝑎𝑛 cos 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡 + ‫׬‬0 𝑏𝑛 sin 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡

𝑇
But from trigonometric integrals, ‫׬‬0 cos 𝑚𝜔0 t 𝑑𝑡 = 0

𝑇 ∞ 𝑇 𝑇
∴ න 𝑓 𝑡 cos 𝑚𝜔0 t 𝑑𝑡 = ෍ න 𝑎𝑛 cos 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡 + න 𝑏𝑛 sin 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡
0 𝑛=1 0 0
 𝑇
Also from trigonometric integrals,‫׬‬0 sin 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡 = 0 while
𝑇
‫׬‬0 cos 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡 = 0 𝑜𝑛𝑙𝑦 𝑤ℎ𝑒𝑛 𝑛 ≠ 𝑚
𝑇 ∞ 𝑇
∴ න 𝑓 𝑡 cos 𝑚𝜔0 t 𝑑𝑡 = ෍ න 𝑎𝑛 cos 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡
0 𝑛=1 0
When 𝑛 = 𝑚,
𝑇 𝑇
𝑇
න 𝑎𝑛 cos 𝑛𝜔0 t cos 𝑚𝜔0 t 𝑑𝑡 = 𝑎𝑛 න 𝑐𝑜𝑠 2 𝑛𝜔0 𝑡 𝑑𝑡 = 𝑎𝑛
0 0 2
𝑇
𝑇
⇒ න 𝑓 𝑡 cos 𝑛𝜔0 t 𝑑𝑡 = 𝑎𝑛
0 2

2 𝑇
∴ 𝑎𝑛 = න 𝑓 𝑡 cos 𝑛𝜔0 t 𝑑𝑡
𝑇 0
Fourier Coefficient: 𝑏𝑛
• To find 𝑏𝑛 multiply both sides of the expression for Fourier series by 𝑠𝑖𝑛 𝑚𝜔0 t
and then integrate over one period :

𝑇 𝑇 ∞

න 𝑓 𝑡 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 = න 𝑎0 + ෍ 𝑎𝑛 cos 𝑛𝜔0 t + 𝑏𝑛 sin 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡

0 0 𝑛=1

𝑇 ∞ 𝑇 𝑇
= න 𝑎0 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 + ෍ න 𝑎𝑛 cos 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 + න 𝑏𝑛 sin 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡
0 𝑛=1 0 0

𝑇
But from trigonometric integrals, ‫׬‬0 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 = 0

𝑇 ∞ 𝑇 𝑇
∴ න 𝑓 𝑡 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 = ෍ න 𝑎𝑛 cos 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 + න 𝑏𝑛 sin 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡
0 𝑛=1 0 0
 𝑇
Also from trigonometric integrals,‫׬‬0 cos 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 = 0 while
𝑇
‫׬‬0 sin 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 = 0 𝑜𝑛𝑙𝑦 𝑤ℎ𝑒𝑛 𝑛 ≠ 𝑚
𝑇 ∞ 𝑇
∴ න 𝑓 𝑡 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 = ෍ න 𝑏𝑛 sin 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡
0 𝑛=1 0
When 𝑛 = 𝑚,
𝑇 𝑇
𝑇
න 𝑏𝑛 sin 𝑛𝜔0 t 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 = 𝑏𝑛 න 𝑠𝑖𝑛2 𝑛𝜔0 𝑡 𝑑𝑡 = 𝑏𝑛
0 0 2
𝑇
𝑇
⇒ න 𝑓 𝑡 𝑠𝑖𝑛 𝑚𝜔0 t 𝑑𝑡 = 𝑏𝑛
0 2

2 𝑇
∴ 𝑏𝑛 = න 𝑓 𝑡 𝑠𝑖𝑛 𝑛𝜔0 t 𝑑𝑡
𝑇 0
Amplitude-Phase Fourier Representation
• The Fourier Series could be represented
∞
in amplitude-phase form thus:
𝑓 𝑡 = 𝑎0 + ෍ 𝐴𝑛 cos(𝑛𝜔0 t + 𝜙𝑛 )
𝑛=1
• Recall: cos 𝑎 + 𝑏 = cos 𝑎 cos 𝑏∞− sin 𝑎 sin 𝑏
∴ 𝑓 𝑡 = 𝑎0 + ෍ 𝐴𝑛 cos(𝑛𝜔0 t + 𝜙𝑛 )
∞ 𝑛=1

= 𝑎0 + ෍ (𝐴𝑛 cos 𝜙𝑛 ) cos 𝑛𝜔0 t − (𝐴𝑛 sin 𝜙𝑛 ) sin 𝑛𝜔0 t

𝑛=1
𝑏𝑛
• Where: 𝑎𝑛 = 𝐴𝑛 cos 𝜙𝑛 , 𝑏𝑛 = −𝐴𝑛 sin 𝜙𝑛 , 𝐴𝑛 = 𝑎𝑛2 + 𝑏𝑛2 , 𝜙𝑛 = − tan−1
𝑎𝑛
• The plot of the amplitude 𝐴𝑛 of the harmonics versus 𝑛𝜔0 is called the amplitude
spectrum of the 𝑓(𝑡); plot of the phase 𝜙𝑛 versus 𝑛𝜔0 is the phase spectrum of 𝑓 𝑡 ;
Both the amplitude and phase spectra form the frequency spectrum of 𝑓(𝑡).
Example 1: • For the ac coefficients:
Determine the Fourier series of the
waveform shown
Obtain the amplitude and phase spectra

• Solution:
∞

𝑓 𝑡 = 𝑎0 + ෍ 𝑎𝑛 cos 𝑛𝜔0 t + 𝑏𝑛 sin 𝑛𝜔0 t

𝑛=1
1, 0<𝑡<1 2𝜋
• From the waveform, 𝑓 𝑡 = ൜ , 𝑇 = 2, 𝜔0 = = 𝜋
0, 1<𝑡<2 𝑇

• For the dc coefficient:

• Therefore,

• The Fourier series contains only the dc Amplitude spectrum

component and sine terms with the
fundamental component and odd
harmonics

Phase spectrum
Example 2: Determine the Fourier Series
for the signal waveform shown.

• Solution:
Fourier Transform
• Fourier series is only applicable to periodic signals.
• There are many signals which are nonperiodic or aperiodic and as such
cannot be represented using Fourier Series.
• Aperiodic signals occupy a continuum of frequencies unlike the periodic
signals which have line spectrum
• Fourier transform allows us to extend the concept of frequency spectrum
to nonperiodic functions.
• Here, signals are transformed from time domain to frequency domain or
from frequency domain to time domain.
• If a nonperiodic signal has finite total energy, its frequency-domain
representation will be a continuous spectrum obtained from the Fourier
transform.
Nonperiodic Signals
• Fig a shows a rectangular pulse bounded
𝜏
along the t-axis by ± .
2
• The pulse has an amplitude value of 𝐴 and is
time-limited as the value of 𝑣(𝑡) remains
zero, outside the pulse.
• Fig b shows a signal which is
asymptomatically time-limited since 𝑣(𝑡) ⟶
0 as 𝑡 ⟶ ±∞
• Any attempt to average 𝑣(𝑡) or 𝑣(𝑡) 2 over
all time would yield zero averages
• Consequently, instead of talking about
average power, a more meaningful property
of a nonperiodic signal is its energy
Relationship between Periodic and
Nonperiodic signals
• p(t) is nonperiodic while f(t) is periodic
with period T
• Shape of f(t) over one period is same as
p(t)
• Letting the period 𝑇 → ∞ yields a single
pulse of width 𝜏 which now becomes
nonperiodic
• Adjacent pulses are moved to infinity
• ∴ 𝑓 𝑡 = 𝑝(𝑡) as period, 𝑇 → ∞
 Effects of increasing T on the spectrum
• The shape of the spectrum remains the same
• The frequency at which the envelope first becomes
zero remains the same
• Amplitude of the spectrum decreases
• Since the total energy of the components within a band
must remain unchanged, the amplitudes of the
harmonics must decrease as T increases
• Principle of energy conservation
• Spacing between adjacent components decreases
• Number of harmonics increases
• Over a range of frequencies, the sum of the
amplitudes of the harmonics remains almost
constant.
• As T increases, f decreases, and as such, the
discrete spectrum ultimately becomes continuous
Normalized Signal Energy
• If 𝑣(𝑡) is the voltage across a resistance, the total delivered energy
𝑣 2 (𝑡)
would be found by integrating the instantaneous power
𝑅
• Normalizing the power by assuming 𝑅 = 1, the normalized signal
energy is:
∞
• 𝐸 ≜ ‫׬‬−∞ 𝑣(𝑡) 2 𝑑𝑡
• This implies that E is just the total area under the curve of 𝑣(𝑡) 2
• Consider the exponential∞representation of Fourier series: • As 𝑇 ⟶ ∞, summation becomes integration,
incremental spacing ∆𝜔 becomes differential separation
𝑓 𝑡 = ෍ 𝑐𝑛 𝑒 𝑗𝑛𝜔0 𝑡 𝑑𝜔, and the discrete harmonic frequency 𝑛𝜔0 becomes
𝑛=−∞ a continuous frequency 𝜔:
• Where:
𝑇 Τ2 ∞ ∞
1 1
𝑐𝑛 = න 𝑓 𝑡 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡 ∴𝑓 𝑡 = න න 𝑓 𝑡 𝑒 −𝑗𝜔𝑡 𝑑𝑡 𝑒 𝑗𝜔𝑡 𝑑𝜔
𝑇 2𝜋
−𝑇 Τ2 −∞ −∞
• Spacing between adjacent harmonics is: • Fourier transform of 𝑓(𝑡) is given as:
2𝜋 ∆𝜔 1
∆𝜔 = 𝑛 + 1 𝜔0 − 𝑛𝜔0 = 𝜔0 = ⇒ =
𝑇 2𝜋 𝑇 ∞

∞ 𝑇 Τ2 ∴ 𝐹 𝜔 = න 𝑓 𝑡 𝑒 −𝑗𝜔𝑡 𝑑𝑡
1
𝑓 𝑡 = ෍ න 𝑓 𝑡 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡 𝑒 𝑗𝑛𝜔0 𝑡 −∞
𝑇 • The Fourier transform is an integral transformation of
𝑛=−∞ −𝑇 Τ2
𝑓(𝑡) from the time domain to the frequency domain
∞ 𝑇 Τ2
∆𝜔 • For a function to have Fourier Transform, it must be
= ෍ න 𝑓 𝑡 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡 𝑒 𝑗𝑛𝜔0 𝑡 completely integrable such that
2𝜋
𝑛=−∞ −𝑇 Τ2
∞
∞ 𝑇 Τ2
1 න 𝑓 𝑡 𝑑𝑡 < ∞
= ෍ න 𝑓 𝑡 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡 ∆𝜔𝑒 𝑗𝑛𝜔0 𝑡
2𝜋 −∞
𝑛=−∞ −𝑇 Τ2
• The magnitude of 𝐹(𝜔) is called the amplitude
spectrum while its phase is the phase spectrum
• Inverse Fourier Transform is given by:
Example 1: Find the Fourier Transform of an
impulse function 𝛿 𝑡 − 𝑡0
• Solution:
∞
• 𝐹 𝜔 = ℱ 𝛿 𝑡 − 𝑡0 = ‫׬‬−∞ 𝛿 𝑡 − 𝑡0 𝑒 −𝑗𝜔𝑡 𝑑𝑡
• From sifting property of the impulse function, when a
function is integrated with the impulse function, we
obtain the value of the function at the point where the
impulse occurs.
∞
• ∴ 𝐹 𝜔 = ‫׬‬−∞ 𝛿 𝑡 − 𝑡0 𝑒 −𝑗𝜔𝑡 𝑑𝑡 = 𝑒 −𝑗𝜔𝑡0
Example 2: Determine the Fourier Transform of the
function: cos 𝜔0 𝑡
• Solution
• 𝐹 𝜔 = ℱ cos 𝜔0 𝑡
𝑒 𝑗𝜃 +𝑒 −𝑗𝜃
• Recall: From Euler’s formula, cos 𝜃 =
2
𝑒 𝑗𝜔0𝑡 +𝑒 −𝑗𝜔0𝑡
• ∴ 𝐹 𝜔 = ℱ cos 𝜔0 𝑡 = ℱ
2
1 1
• = ℱ 𝑒 𝑗𝜔0 𝑡 + ℱ 𝑒 −𝑗𝜔0 𝑡
2 2
1 1
•= 2𝜋𝛿 𝜔 − 𝜔0 + 2𝜋𝛿 𝜔 + 𝜔0
2 2
• = 𝜋𝛿 𝜔 − 𝜔0 + 𝜋𝛿 𝜔 + 𝜔0
 Properties of Fourier Transform
• Linearity:

• Time scaling
Time expansion corresponds to
frequency compression and vice versa

• Time Shifting

• Delay in the time domain corresponds to a phase shift in the

frequency domain
• Frequency Shifting (Amplitude Modulation)

• A frequency shift in the frequency domain adds a phase shift

to the time function
• Time Differentiation

• The transform of the derivative of f(t) is got by multiplying the

transform of f(t) by 𝑗𝜔
• Time Integration:

• The transform of the integral of f(t) is got by dividing the transform of

f(t) by 𝑗𝜔 and adding the result to the impulse term that reflects the
dc component F(0)
• Reversal:

• States that reversing f(t) about the time axis reverses 𝐹 𝜔

about the frequency axis
• Duality: If 𝐹 𝜔 is the Fourier Transform of 𝑓(𝑡), then the
Fourier transform of 𝐹 𝑡 is 2π𝑓 −𝜔

• Convolution
• If 𝑥(𝑡) is the input excitation to a circuit with an impulse function of
ℎ(𝑡) then the output response is:
∞
• 𝑦 𝑡 = ℎ 𝑡 ∗ 𝑥 𝑡 = ‫׬‬−∞ ℎ 𝜆 𝑥 𝑡 − 𝜆 𝑑𝜆
• If 𝑋 𝜔 , 𝐻 𝜔 , and 𝑌(𝜔) are the Fourier transforms of x t , ℎ(𝑡)
and 𝑦(𝑡) respectively, then
• 𝑌 𝜔 =ℱ ℎ 𝑡 ∗𝑥 𝑡 =𝐻 𝜔 𝑋 𝜔
• Convolution in the time domain corresponds to multiplication in the
frequency domain
 Example 3: Find the Fourier
Transform of the function shown:
• Solution: {let us explore the derivative property}
• 1st derivative:

• 2nd derivative:

• Taking Fourier Transform of both sides:

 Example 4: Obtain the inverse
Fourier Transform of:
• Solution:
• Substitute 𝑗𝜔 with 𝑠 in the function given.

• This is to avoid complex algebra

• Applying Partial fraction expansion technique:

• Inverse FT:
Convolution
• The convolution of two signals consists of time-reversing one of the
signals, shifting it, and multiplying it point by point with the second
signal, and integrating the product
• Given two functions of the same variable, say 𝑣(𝑡) and 𝑤(𝑡), their
convolution is:
∞
𝑣(𝑡) ∗ 𝑤(𝑡) ≜ න 𝑣 𝜆 𝑤 𝑡 − 𝜆 𝑑𝜆
−∞
• Output is equal to the input convolved with the unit impulse response
• Where 𝜆 is a dummy variable and asterisk denotes convolution
Properties of convolution
• Commutative:
𝑦 𝑡 =𝑣 𝑡 ∗𝑤 𝑡 =𝑤 𝑡 ∗𝑣 𝑡
∞ ∞
𝑦 𝑡 = න 𝑣 𝜆 𝑤 𝑡 − 𝜆 𝑑𝜆 = න 𝑤 𝜆 𝑣 𝑡 − 𝜆 𝑑𝜆
−∞ −∞
• The order in which the two functions are convolved is of no consequence
• Associative:
𝑥 𝑡 ∗ 𝑣 𝑡 ∗𝑤 𝑡 = [𝑥 𝑡 ∗ 𝑣 𝑡 ] ∗ 𝑤 𝑡
• Distributive:
𝑥 𝑡 ∗ 𝑣 𝑡 +𝑤 𝑡 = 𝑥 𝑡 ∗𝑣 𝑡 + [𝑥(𝑡) ∗ 𝑤 𝑡 ]
Other properties
∞
𝑥 𝑡 ∗ 𝛿 𝑡 = න 𝑥 𝜆 𝛿 𝑡 − 𝜆 𝑑𝜆 = 𝑥(𝑡)
−∞

𝑥 𝑡 ∗ 𝛿 𝑡 − 𝑡0 = 𝑥 𝑡 − 𝑡0

∞
𝑥 𝑡 ∗ 𝛿 𝑡 = න 𝑥 𝜆 𝛿 ′ 𝑡 − 𝜆 𝑑𝜆 = 𝑥 ′ (𝑡)
′
−∞

∞ 𝑡
𝑥 𝑡 ∗ 𝑢 𝑡 = න 𝑥 𝜆 𝑢 𝑡 − 𝜆 𝑑𝜆 = න 𝑥 𝜆 𝑑𝜆
−∞ −∞
 Convolution • Proof:
Theorems ∞ ∞

• The Convolution ℱ 𝑣 𝑡 ∗𝑤 𝑡 = න න 𝑣 𝜆 𝑤 𝑡 − 𝜆 𝑑𝜆 𝑒 −𝑗𝜔𝑡 𝑑𝑡

theorems state that −∞ −∞

convolution in the time ∞ ∞

domain becomes ℱ 𝑣 𝑡 ∗𝑤 𝑡 = න𝑣 𝜆 න 𝑤 𝑡 − 𝜆 𝑒 −𝑗𝜔𝑡 𝑑𝑡 𝑑𝜆
multiplication in the
−∞ −∞
frequency domain, while
multiplication in the time ∞

domain becomes = න 𝑣 𝜆 𝑊 𝑓 𝑒 −𝑗𝜔𝜆 𝑑𝜆

convolution in the −∞
frequency domain ∞

= න 𝑣 𝜆 𝑒 −𝑗𝜔𝜆 𝑑𝜆 𝑊 𝑓
𝑣 𝑡 ∗𝑤 𝑡 ⟷𝑉 𝑓 𝑊 𝑓
−∞

𝑣 𝑡 𝑤 𝑡 ⟷𝑉 𝑓 ∗𝑊 𝑓