Representation of Signals and Systems:

Fourier series, Fourier Transform, Properties of Fourier Transform,

Signal power and power spectral density, Signal energy and energy
spectral density, Dirac delta function and its applications

Elements of a Communication system, Block diagram of digital

communication system.
Signals: A signal is a set of information or data.
Examples:
• A telephone or a television signal,
• The monthly sales figures of a corporation or closing stock prices.
✓ the independent variable time.
• A signal may be → a voltage or a current.
The instantaneous power dissipated in the resistor is defined by
𝑣(𝑡) 2 2
𝑝= or 𝑝 = 𝑅 𝑖(𝑡)
𝑅
For resistor R=1 ohm,
power associated with the signal is 𝑝 = 𝑔(𝑡) 2
square of the amplitude of the signal.
Regardless of whether given signal 𝑔 𝑡 is a voltage or current
Systems: processes a set of signals (inputs) to yield another set of signals
(outputs).
• A system may be made up of physical components, as in electrical,
mechanical, or hydraulic systems (hardware realization),
• or it may be an algorithm that computes an output from an input signal
(software realization).
Linear System: A system is linear when it obeys the principle of
superposition.
➢ If 𝑦1 (𝑡) and 𝑦2 (𝑡)are the responses of a system to the input 𝑥1 (𝑡) and
𝑥2 (𝑡) respectively.
➢ Then the response of the composite signal 𝑎𝑥1 (𝑡) +b𝑥2 (𝑡) of a linear
system will be 𝑎𝑦1 (𝑡) +b𝑦2 (𝑡). Where a and b are arbitrary constants.
Time -invariant: If 𝑦 𝑡 is the response of a system to the input 𝑥 𝑡 ,
Then the response of the time invariant system for the time shifted input
𝑥 𝑡 − 𝑡0 is equal to 𝑦 𝑡 − 𝑡0 . Where 𝑡0 is a constant.
Depending on the feature we may have five classes of signals.

1. Continuous time and discrete time signals

2. Analog and digital signals
3. Periodic and aperiodic signals
4. Deterministic and probabilistic signals
5. Energy and power signals
Continuous time and discrete time signals
Periodic: A signal 𝑔(𝑡) is said to be periodic if there exists a positive constant
𝑇0 such that
𝑔 𝑡 = 𝑔 𝑡 + 𝑛𝑇0 𝑤ℎ𝑒𝑟𝑒 𝑛 = 0,1,2,3 … . . 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑣𝑎𝑙𝑢𝑒𝑠 𝑜𝑓 𝑡
The smallest value of 𝑻𝟎 that satisfies the periodicity condition is known as
the period of g(t).
Aperiodic: If there is no value of 𝑇0 that satisfy the above condition is
aperiodic signal.
• By definition, a periodic signal g (t) remains unchanged when time-shifted
by one period.
⇒ a periodic signal must start at 𝑡 = −∞ and continue forever, as
shown in the figure.

• Periodic g(t) can be generated by periodic extension of any segment of g(t)

of duration 𝑇𝑜 ( the period)

Segment of g(t)
starting at t=-1

Segment of g(t)
starting at t=0
Deterministic: A signal whose physical description is known
completely, either in a mathematical form or a graphical form is a
deterministic signal.

Probabilistic signal: A signal that is known only in terms of

probabilistic description, such as mean value, mean square value, and
distributions, rather than its full mathematical or graphical description
is a random signal.

• Noise signals

• Message signals
 ∞
2
𝐸𝑔 = න 𝑔 𝑡 𝑑𝑡
−∞

1 𝑇/2 2
𝑃𝑔 = lim න 𝑔 𝑡 𝑑𝑡
𝑇→∞ 𝑇 −𝑇/2
• Periodic signal → as an infinite sum of sine and cosine terms.
• Motivation: To represent the signal in frequency domain.
Let 𝑔𝑝 𝑡 = a periodic signal with period 𝑇0
2𝜋𝑛𝑡 2𝜋𝑛𝑡
Then 𝑔𝑝 𝑡 = 𝑎0 + 2 σ∞
𝑛=1 𝑎𝑛 𝑐𝑜𝑠 + 𝑏𝑛 𝑠𝑖𝑛 (1)
𝑇0 𝑇0

Where 𝑎𝑛 , 𝑏𝑛 = Unknown coefficient (amplitude) of sine and cosine terms

𝑛 1
= nth harmonic of the fundamental frequency
𝑇0 𝑇0
2𝜋𝑛𝑡 2𝜋𝑛𝑡
Each of the term 𝑐𝑜𝑠 𝑎𝑛𝑑 𝑠𝑖𝑛 = basis function
𝑇0 𝑇0

The basis function forms an orthogonal set over the interval 𝑇0

Basis Function: No one of them can be expressed as a sum of other N-1 terms.
 𝑻𝟎
𝟐
𝟐𝝅𝒎𝒕 𝟐𝝅𝒏𝒕 𝑻𝟎
න 𝒄𝒐𝒔 . 𝒄𝒐𝒔 𝒅𝒕 = ቐ 𝟐 , 𝒎=𝒏
𝑻𝟎 𝑻𝟎
𝑻𝟎 𝟎, 𝒎≠𝒏
−𝟐

𝑻𝟎
𝟐
𝟐𝝅𝒎𝒕 𝟐𝝅𝒏𝒕 𝑻𝟎
න 𝒔𝒊𝒏 . 𝒔𝒊𝒏 𝒅𝒕 = ቐ 𝟐 , 𝒎=𝒏
𝑻𝟎 𝑻𝟎
𝑻𝟎 𝟎, 𝒎≠𝒏
−𝟐

𝑻𝟎
𝟐
𝟐𝝅𝒎𝒕 𝟐𝝅𝒏𝒕
න 𝒄𝒐𝒔 . 𝒔𝒊𝒏 𝒅𝒕 = 𝟎 𝒇𝒐𝒓 𝒂𝒍𝒍 𝒗𝒂𝒍𝒖𝒆𝒔 𝒐𝒇 𝒎 𝒂𝒏𝒅 𝒏
𝑻𝟎 𝑻𝟎
𝑻
− 𝟐𝟎

The Fourier coefficient 𝑎0 , 𝑎𝑛 , 𝑏𝑛 are to be determined

To determine 𝒂𝟎 , integrate both side of the equation (1)
𝑎0 = mean value of the periodic signal 𝑔𝑝 𝑡 over one period = dc value.
𝑻𝟎
𝟐
𝟏
𝒂𝟎 = න 𝒈𝒑 𝒕 𝒅𝒕
𝑻𝟎
𝑻
− 𝟐𝟎

To determine 𝒂𝒏
2𝜋𝑛𝑡
Multiply both side of the equation (1) by 𝑐𝑜𝑠 and integrate over time period.
𝑇0

𝑻𝟎
𝟐
𝟏 𝟐𝝅𝒏𝒕
𝒂𝒏 = න 𝒈𝒑 𝒕 𝒄𝒐𝒔 𝒅𝒕 𝒇𝒐𝒓 𝒏 = 𝟏, 𝟐, 𝟑 …
𝑻𝟎 𝑻𝟎
𝑻
− 𝟐𝟎

Similarly, for 𝒃𝒏
𝑻𝟎
𝟐
𝟏 𝟐𝝅𝒏𝒕
𝒃𝒏 = න 𝒈𝒑 𝒕 𝒔𝒊𝒏 𝒅𝒕 𝒇𝒐𝒓 𝒏 = 𝟏, 𝟐, 𝟑 …
𝑻𝟎 𝑻𝟎
𝑻
− 𝟐𝟎
To apply Fourier series the following condition must be satisfied:
✓ The function 𝑔𝑝 𝑡 is single valued within the interval 𝑇0
✓ The function 𝑔𝑝 𝑡 has finite number of discontinuities in the interval 𝑇0
✓ The function 𝑔𝑝 𝑡 has finite number of maxima and minima in the
interval 𝑇0
✓ The function 𝑔𝑝 𝑡 is absolutely integrable in the interval 𝑇0 ie
𝑇0
2
න 𝑔𝑝 𝑡 𝑑𝑡 < ∞
𝑇
− 20

These conditions are known as Dirichlet’s conditions.

They are satisfied by the periodic signal usually encountered in
communication system.
 2𝜋𝑛𝑡 2𝜋𝑛𝑡
𝑐𝑜𝑠 𝑎𝑛𝑑 𝑠𝑖𝑛 can be written in the complex exponential form as
𝑇0 𝑇0

2𝜋𝑛𝑡 1 𝑗2𝜋𝑛𝑡 −𝑗2𝜋𝑛𝑡

𝑐𝑜𝑠 = 𝑒 𝑇0 + 𝑒 𝑇0
𝑇0 2
𝑗2𝜋𝑛𝑡 −𝑗2𝜋𝑛𝑡
2𝜋𝑛𝑡 1
𝑠𝑖𝑛 = 𝑒 𝑇0 −𝑒 𝑇0
𝑇0 2𝑗

𝑗2𝜋𝑛𝑡 −𝑗2𝜋𝑛𝑡 𝑗2𝜋𝑛𝑡 −𝑗2𝜋𝑛𝑡

𝑔𝑝 𝑡 = 𝑎0 + σ∞
𝑛=1 𝑎𝑛 𝑒
𝑇0 + 𝑎𝑛 𝑒 𝑇0 − 𝑗𝑏𝑛 𝑒 𝑇0 + 𝑗𝑏𝑛 𝑒 𝑇0

∞
𝑗2𝜋𝑛𝑡 −𝑗2𝜋𝑛𝑡
𝑔𝑝 𝑡 = 𝑎0 + ෍ (𝑎𝑛 −𝑗𝑏𝑛 ) 𝑒 𝑇0 + (𝑎𝑛 +𝑗𝑏𝑛 ) 𝑒 𝑇0
𝑛=1
 ∞
𝑗2𝜋𝑛𝑡 −𝑗2𝜋𝑛𝑡
𝑔𝑝 𝑡 = 𝑎0 + ෍ (𝑎𝑛 −𝑗𝑏𝑛 ) 𝑒 𝑇0 + (𝑎𝑛 +𝑗𝑏𝑛 ) 𝑒 𝑇0
𝑛=1

Let 𝐶𝑛 = (𝑎𝑛 −𝑗𝑏𝑛 ) for n > 0

𝑎0 for n = 0

(𝑎𝑛 +𝑗𝑏𝑛 ) for n < 0

𝒋𝟐𝝅𝒏𝒕
 𝒈𝒑 𝒕 = σ∞
−∞ 𝑪𝒏 𝒆
𝑻𝟎
Complex exponential Fourier series
𝑇0 −𝑗2𝜋𝑛𝑡
1
 𝐶𝑛 = ‫׬‬ 2
𝑔𝑝 𝑡 𝑒 𝑇0 𝑑𝑡 𝑓𝑜𝑟 𝑛 = ±1, ±2, ±3 …
𝑇0 −𝑇0
2
=Complex Fourier coefficient
• The representation of a periodic signal → using various harmonic components of
frequencies 0, ±𝑓0 , ±2𝑓0 , ±3𝑓0 ……...
1
where 𝑓0 = is the fundamental frequency.
𝑇0

➢ Frequency domain representation of the signal is called as the spectrum.

Thus, a periodic signal 𝑔𝑝 𝑡 can be represented in two ways:

• Time domain representation: where 𝑔𝑝 𝑡 is defined as a function of time.

• Frequency domain representation: where 𝑔𝑝 𝑡 is defined in terms of its

spectrum.

The Fourier coefficient 𝑪𝒏 is a complex number and can be expressed as

𝑪𝒏 = 𝑪𝒏 𝒆𝒋𝜽𝒏

Where 𝐶𝑛 = amplitude of the nth harmonic component of the signal 𝑔𝑝 𝑡

A plot of 𝐶𝑛 vs frequency yields discrete amplitude spectrum of the signal.

➢ A plot of arg (𝐶𝑛 ) 𝑤ℎ𝑖𝑐ℎ 𝑖𝑠 𝜃 vs frequency yields discrete phase spectrum of the
signal

➢ 𝐶−𝑛 = 𝐶𝑛 ∗ where 𝐶𝑛 ∗ is the complex conjugate of 𝐶𝑛

𝐶−𝑛 = 𝐶𝑛

and arg 𝐶−𝑛 = − arg( 𝐶𝑛 )

➢ Thus, the amplitude spectrum of a real valued periodic signal is symmetric (an even
function of n)

➢ Whereas the phase spectrum is antisymmetric (an odd function of n) about the
vertical axis passing through the origin.
What if the signal is non-periodic.
To represent non-periodic signal, we have to first construct a periodic signal
𝑔𝑝 𝑡 of period 𝑇0 in such a way that 𝑔(𝑡) defines one cycle of this periodic
function.
Let 𝑇0 is infinite ⇒ lim 𝑔𝑝 𝑡 = 𝑔 𝑡
𝑇0 →∞
𝑗2𝜋𝑛𝑡
From Fourier series we have 𝑔𝑝 𝑡 = σ∞
−∞ 𝐶𝑛 𝑒
𝑇0

𝑇0 −𝑗2𝜋𝑛𝑡
1
Where 𝐶𝑛 = ‫׬‬ 2
𝑔 𝑡 𝑒 𝑇0 𝑑𝑡
𝑇0 −𝑇0 𝑝
2
1 𝑛
Let ∆𝑓 = ; 𝑓𝑛 = and 𝐺 𝑓𝑛 = 𝐶𝑛 𝑇0
𝑇0 𝑇0
𝐺 𝑓𝑛
Then 𝑔𝑝 𝑡 = σ∞ 𝐶
−∞ 𝑛 𝑒 𝑗2𝜋𝑓𝑛 𝑡 = σ∞
−∞ 𝑒 𝑗2𝜋𝑓𝑛 𝑡 = σ∞
−∞ 𝐺 𝑓𝑛 𝑒 𝑗2𝜋𝑓𝑛 𝑡 ∆𝑓
𝑇0
𝑇0
2

wℎ𝑒𝑟𝑒 𝐺 𝑓𝑛 = න 𝑔𝑝 𝑡 𝑒 −𝑗2𝜋𝑓𝑛𝑡 𝑑𝑡
𝑇
− 0
2
 1
Now let 𝑇0 → ∞ => ∆𝑓 = →0
𝑇0
𝑛
∆𝑓 → 0 => discrete frequency 𝑓𝑛 (= ) approaches the continuous
𝑇0
frequency variable f and
Discrete sum of 𝑔𝑝 𝑡 = σ∞ −∞ 𝐺 𝑓𝑛 𝑒
𝑗2𝜋𝑓𝑛 𝑡
∆𝑓 becomes an integral
defining the areas under a continuous function of frequency f namely
𝐺(𝑓)𝑒 𝑗2𝜋𝑓𝑛 𝑡
Also, 𝑇0 → ∞ => 𝑔𝑝 𝑡 → 𝑔 𝑡 { from σ∞
−∞ 𝐺 𝑓𝑛 𝑒 𝑗2𝜋𝑓𝑛 𝑡 ∆𝑓 }
∞

𝒈 𝒕 = න 𝑮(𝒇) 𝒆𝒋𝟐𝝅𝒇𝒕 𝒅𝒇
−∞
𝑇0
Where,{𝐺 𝑓𝑛 = ‫׬‬2
𝑇 𝑔𝑝 𝑡 𝑒 −𝑗2𝜋𝑓𝑛 𝑡 𝑑𝑡}
− 20
∞

𝑮(𝒇) = න 𝒈(𝒕) 𝒆−𝒋𝟐𝝅𝒇𝒕 𝒅𝒕

−∞
➢ 𝐺(𝑓) can be thought of as a transformed version of 𝑔(𝑡) and is
referred to as the Fourier transform of 𝑔(𝑡)
➢ 𝑔 𝑡 is referred to as the inverse Fourier transform of 𝐺(𝑓)
➢ 𝑔(𝑡) and 𝐺(𝑓) constitute a Fourier transform pair (𝑔(𝑡) 𝐺(𝑓))
𝑭 𝒈(𝒕) = 𝑮(𝒇) and 𝑭−𝟏 𝑮(𝒇) = 𝒈(𝒕)
For a signal 𝑔(𝑡) to be Fourier transformable 𝑔(𝑡) should satisfy
Dirichlet’s Conditions.
The function 𝑔(𝑡) is single valued with finite number of maxima and
minima and a finite number of discontinuities in any finite time
interval.
∞
The function 𝑔(𝑡) is absolutely integrable ie ‫׬‬−∞ 𝑔(𝑡) 𝑑𝑡 < ∞

These conditions include all energy signals for which

∞

න 𝑔(𝑡) 2 𝑑𝑡 < ∞
−∞
The Fourier transform 𝐺(𝑓) is a complex function of frequency f and can be
expressed as
𝐺(𝑓) = 𝐺(𝑓) 𝑒 𝑗𝜃(𝑓)
Where 𝐺(𝑓) = Continuous amplitude of the 𝑔(𝑡) and 𝜃(𝑓) Continuous phase
of the 𝑔(𝑡)
Continuous: because 𝑮(𝒇) is defined for all values of 𝒇
For a real valued periodic signal 𝑔 𝑡
𝐺(𝑓) = 𝐺 ∗ (−𝑓)
⇒ 𝐺 −𝑓 = 𝐺(𝑓) and 𝜃 −𝑓 = −𝜃(𝑓)
Thus, the amplitude spectrum of a real valued signal is an even function of 𝑓
Whereas the phase spectrum 𝜃(𝑓)is an odd function of 𝑓
PROPERTY 1 Linearity (Superposition)

If 𝒈𝟏 𝒕 𝑮𝟏 𝒇 and 𝒈𝟐 𝒕 𝑮𝟐 𝒇
Then for all constants a and b and we have
𝒂𝒈𝟏 𝒕 + 𝒃𝒈𝟐 𝒕 𝒂𝑮𝟏 𝒇 + 𝒃𝑮𝟐 𝒇

∞
Proof: 𝐹 𝑎𝑔1 𝑡 + 𝑏𝑔2 𝑡 = ‫׬‬−∞ 𝑎𝑔1 𝑡 + 𝑏𝑔2 𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
∞ ∞

= 𝑎 න 𝑔1 𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡 + 𝑏 න 𝑏𝑔2 𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡

−∞ −∞
= 𝑎𝐺1 𝑓 + 𝑏𝐺2 𝑓
PROPERTY 2 Time Scaling: If 𝒈(𝒕) 𝑮(𝒇)
𝟏 𝒇
Then 𝒈(𝒂𝒕) 𝑮
𝒂 𝒂

∞
Proof: 𝑭 𝑔(𝑎𝑡) = ‫׬‬−∞ 𝑔(𝑎𝑡) 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
𝜏 1
Let 𝜏 = 𝑎𝑡 => 𝑡 = => 𝑑𝑡 = 𝑑𝜏
𝑎 𝑎
𝜏
1 ∞ −𝑗2𝜋𝑓
If 𝑎 > 0 𝑭 𝑔(𝑎𝑡) = ‫)𝜏(𝑔 ׬‬ 𝑒 𝑎 𝑑𝜏
𝑎 −∞
1 𝑓
= 𝐺
𝑎 𝑎
𝜏 1
If 𝑎 < 0 𝜏 = −𝑎𝑡 => 𝑡 = − => 𝑑𝑡 = − 𝑑𝜏
𝑎 𝑎
And limit changes from ∞ 𝑡𝑜 − ∞
∞
1 𝜏
−𝑗2𝜋𝑓𝑎
𝑭 𝑔(−𝑎𝑡) = − න 𝑔(𝜏) 𝑒 𝑑𝜏
𝑎
−∞
1 𝑓
− 𝐺
𝑎 𝑎
𝑔 𝑎𝑡 → represents 𝑔(𝑡) compressed in time by factor a
𝑓
𝐺 → represents G(𝑓) expanded in frequency by the factor a
𝑎
PROPERTY 3 Duality: If 𝒈(𝒕) 𝑮(𝒇)
Then 𝑮(𝒕) 𝒈 −𝒇

Proof: Inverse Fourier transform

𝐹 −1 𝐺(𝑓) = 𝑔(𝑡)
∞

𝑔 𝑡 = න 𝐺(𝑓) 𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓
−∞
∞

𝑔 −𝑡 = න 𝐺(𝑓) 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑓
−∞
And then interchanging t and f we get
∞
𝑔 −𝑓 = ‫׬‬−∞ 𝐺(𝑡) 𝑒 −𝑗2𝜋𝑓𝑡 𝑑t
PROPERTY 4 Time Shifting: If 𝒈(𝒕) 𝑮(𝒇)
Then 𝒈(𝒕 − 𝒕𝟎 ) 𝑮(𝒇)𝒆−𝒋𝟐𝝅𝒇𝒕𝟎

∞
Proof: 𝐹 𝑔(𝑡 − 𝑡0 ) = ‫׬‬−∞ 𝑔(𝑡 − 𝑡0 ) 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
𝜏 = 𝑡 − 𝑡0 => 𝑑𝜏 = 𝑑𝑡 and 𝜏 + 𝑡0 = 𝑡
∞
𝐹 𝑔(𝑡 − 𝑡0 ) = ‫׬‬−∞ 𝑔(𝜏) 𝑒 −𝑗2𝜋𝑓𝜏 𝑒 −𝑗2𝜋𝑓𝑡0 𝑑𝜏
−𝑗2𝜋𝑓𝑡 ∞
= 𝑒 0 ‫׬‬−∞ 𝑔(𝜏) 𝑒 −𝑗2𝜋𝑓𝜏 𝑑𝜏
𝐺(𝑓)𝑒 −𝑗2𝜋𝑓𝑡0
If 𝑔(𝑡) is shifted in +ve direction by amount 𝑡0 , the effect is equivalent
to multiply its Fourier transform by 𝑒 −𝑗2𝜋𝑓𝑡0
Amplitude of G(f) is unaffected by the time shift but its phase is
changed by the amount −2𝜋𝑓𝑡0
PROPERTY 5 Frequency Shifting: (Modulation Theorem)
If 𝒈(𝒕) 𝑮(𝒇)
Then 𝐠(𝒕)𝒆𝒋𝟐𝝅𝒇𝑪𝒕 𝑮(𝒇 − 𝒇𝑪 ) Where 𝒇𝑪 = real constant

𝒋𝟐𝝅𝒇𝑪 𝒕 ∞
Proof: 𝐹 𝒈(𝒕)𝒆 = ‫׬‬−∞ 𝒈(𝒕)𝒆𝒋𝟐𝝅𝒇𝑪 𝒕 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
∞

= න 𝒈(𝒕)𝒆−𝒋𝟐𝝅(𝒇−𝒇𝑪)𝒕 𝑑𝑡
−∞
= 𝐺(𝒇 − 𝒇𝑪 )
• That is multiplication of a function 𝑔(𝑡) by the factor 𝑒 𝑗2𝜋𝑓𝐶 𝑡 is
• equivalent to shifting its Fourier transform 𝐺(𝑓) in the +ve direction
by the amount 𝒇𝑪
PROPERTY 6 Area Under g(t)
𝒈(𝒕) 𝑮(𝒇)
∞

න 𝒈(𝒕) 𝒅𝒕 = 𝑮(𝟎)
−∞
That is area under g(t) = the value of its Fourier transforms 𝑮(𝒇) at f=0
∞ ∞ ∞

𝐺 𝑓 = න 𝑔(𝑡) 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡 => 𝐺 0 = න 𝑔(𝑡) 𝑒 −𝑗2𝜋𝑡0 𝑑𝑡 = න 𝑔(𝑡) 𝑑𝑡

−∞ −∞ −∞
PROPERTY 7 Area under 𝑮(𝒇)
𝒈(𝒕) 𝑮(𝒇)
∞
g 𝟎 = ‫׬‬−∞ 𝑮(𝒇) 𝒅𝐟

That is area under G(f) = the value of its Fourier transforms 𝑔(𝒕) at t=0

Proof:
∞ ∞ ∞

𝑔 𝑡 = න 𝐺(𝑓) 𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓 => 𝑔 0 = න 𝐺(𝑓) 𝑒 𝑗2𝜋𝑓0 𝑑𝑓 = න 𝐺(𝑓) 𝑑𝑓

−∞ −∞ −∞
PROPERTY 8 Differentiation in time Domain
𝒈(𝒕) 𝑮(𝒇) and the 1st derivative of g(t) is Fourier transformable then
𝒅
𝒈(𝒕) 𝒋𝟐𝝅𝒇𝑮(𝒇)
𝒅𝒕
Ie differentiation of a time function 𝑔(𝑡) has the effect of multiplying its Fourier
transform 𝐺(𝑓) by factor 𝑗2𝜋𝑓
∞ 𝑑 𝑑 ∞
Proof: 𝑔 𝑡 = ‫׬‬−∞ 𝐺(𝑓) 𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓=> 𝑔 𝑡 = ‫׬‬ 𝐺(𝑓) 𝑒 𝑗2𝜋𝑓𝑡
𝑑𝑓
𝑑𝑡 𝑑𝑡 −∞

Interchanging integration and differentiation

∞ ∞
𝑑
𝑔 𝑡 = න 𝑗2𝜋𝑓𝐺(𝑓) 𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓 = 𝑗2𝜋𝑓 න 𝐺(𝑓) 𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓
𝑑𝑡
−∞ −∞
𝑑
𝐹 𝑔 𝑡 = 𝑗2𝜋𝑓𝐺(𝑓)
𝑑𝑡

𝒅𝒏
In generalized form 𝒈(𝒕) (𝒋𝟐𝝅𝒇)𝒏 𝑮(𝒇)
𝒅𝒕𝒏
PROPERTY 9 Integration in the time Domain
𝒈(𝒕) 𝑮(𝒇) and if G(0) =0 we have
𝒕
𝟏
න 𝒈(𝝉) 𝒅𝝉 𝑮(𝒇)
𝟐𝒋𝝅𝒇
−∞
ie integration of 𝑔(𝑡) has the effect of dividing the Fourier transform by 𝑗2𝜋𝑓 provided 𝐺(0) =
0
Expressing g(t) as
𝑡
𝑑
𝑔 𝑡 = න 𝑔(𝜏) 𝑑𝜏
𝑑𝑡
−∞
Applying differentiation property
𝑡

𝐺 𝑓 = 𝑗2𝜋𝑓 𝐹 න 𝑔(𝜏) 𝑑𝜏
−∞
𝑡
1
𝐹 න 𝑔(𝜏) 𝑑𝜏 = G(f)
𝑗2𝜋𝑓
−∞
PROPERTY 10 Conjugate Function
𝒈(𝒕) 𝑮(𝒇)
then for a complex valued time function g(t) we have
𝒈∗ (𝒕) 𝑮∗ (−𝒇)

Proof:
∞

𝑔 𝑡 = න 𝐺(𝑓) 𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓
−∞
Taking complex conjugate both side
∞

𝑔∗ 𝑡 = න 𝐺 ∗ (𝑓) 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑓
−∞
Replacing f with -f we have
−∞

𝑔∗ 𝑡 = − න 𝐺 ∗ (−𝑓) 𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓
∞
∞

න 𝐺 ∗ (−𝑓) 𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓
−∞
𝑔∗ 𝑡 𝐺 ∗ (−𝑓)
PROPERTY 11 Multiplication in Time Domain
If 𝒈𝟏 𝒕 𝑮𝟏 𝒇 and 𝒈𝟐 𝒕 𝑮𝟐 𝒇
∞
Then 𝒈𝟏 𝒕 𝒈𝟐 𝒕 𝑮𝟏 𝒇 ∗ 𝑮𝟐 𝒇 ‫׬‬−∞ 𝑮𝟏 𝝀 𝑮𝟏 𝒇 − 𝝀 𝒅𝝀
Proof: Let us represent Fourier transform of the product 𝑔1 𝑡 𝑔2 𝑡 by 𝐺12 𝑓 we
may write
𝑔1 𝑡 𝑔2 𝑡 𝐺12 𝑓
∞
𝐺12 𝑓 = ‫׬‬−∞ 𝑔1 𝑡 𝑔2 𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
∞ ′ 𝑗2𝜋𝑓 ′𝑡
For 𝑔2 𝑡 = ‫׬‬−∞ 𝐺2 (𝑓 ) 𝑒 𝑑𝑓 ′
+∞ ∞ ′
𝐺12 𝑓 = ‫׬‬−∞ ‫׬‬−∞ 𝑔1 𝑡 𝐺2 (𝑓 ′ ) 𝑒 −𝑗2𝜋(𝑓−𝑓 )𝑡 𝑑𝑓 ′ 𝑑𝑡

𝑓 − 𝑓 ′ = 𝜆 => 𝑓 ′ = 𝑓 − 𝜆
∞ ∞

𝐺12 𝑓 = න 𝑑𝜆 𝐺2 (𝑓 − 𝜆) න 𝑔1 𝑡 𝑒 −𝑗2𝜋𝜆𝑡 𝑑𝑡
−∞ −∞
∞
‫׬‬−∞ 𝐺1 (𝜆) 𝐺2 (𝑓 − 𝜆)𝑑𝜆
• Convolution is commutative that is 𝐺12 𝑓 = 𝐺21 𝑓 => 𝐺1 𝑓 ⨂ 𝐺2 𝑓 = 𝐺2 𝑓 ⨂ 𝐺1 𝑓
PROPERTY 12 Convolution in Time Domain
If 𝒈𝟏 𝒕 𝑮𝟏 𝒇 and 𝒈𝟐 𝒕 𝑮𝟐 𝒇
∞

න 𝒈𝟏 𝝉 𝒈𝟐 𝒕 − 𝝉 𝒅𝝉 𝑮𝟏 (𝒇)𝑮𝟐 (𝒇)
−∞
Using Duality property and multiplication property Convolution of two signals in time domain
is transformed into the multiplication of this individual Fourier transforms in the frequency
domain.
𝑔1 𝑡 ⨂𝑔2 𝑡 𝐺1 𝑓 . 𝐺2 𝑓
∞ ∞

න න 𝑔1 𝜏 𝑔2 𝑡 − 𝜏 𝑑𝜏. 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
−∞ −∞

Let 𝑡 = 𝜏 + 𝜆 => 𝑑𝑡 = 𝑑𝜆
∞ ∞
=‫׬‬−∞ ‫׬‬−∞ 𝑔1 𝜏 𝑔2 𝜆 𝑑𝜏. 𝑒 −𝑗2𝜋𝑓(𝜏+𝜆) 𝑑𝜆
∞ ∞

= න 𝑔1 𝜏 𝑒 −𝑗2𝜋𝑓𝜏 𝑑𝜏. න 𝑔2 𝜆 𝑒 −𝑗2𝜋𝑓𝜆 𝑑𝜆

−∞ −∞

=𝐺1 (𝑓)𝐺2 (𝑓)

A periodic signal 𝑔𝑝 𝑡 with a time period of the signal = 𝑇0
is represented in terms of the complex exponential Fourier series
𝑗2𝜋𝑛𝑡
𝑔𝑝 𝑡 = σ∞
−∞ 𝐶𝑛 𝑒𝑥𝑝 𝑇0
𝑇0
1 −𝑗2𝜋𝑛𝑡
Where 𝐶𝑛 = complex Fourier coefficient 𝐶𝑛 = ‫׬‬ 2
𝑔 𝑡 𝑒𝑥𝑝 𝑑𝑡
𝑇0 −𝑇0 𝑝 𝑇0
2

If 𝑔 𝑡 is a pulse like function, which equals 𝑔𝑝 𝑡 over one period and is

𝑇0 𝑇0
𝑔𝑝 𝑡 , − ≤𝑡≤
zero elsewhere; that is, 𝑔 𝑡 =൝ 2 2
0, 𝑒𝑙𝑠𝑒𝑤ℎ𝑒𝑟𝑒
Then periodic signal 𝑔𝑝 𝑡 may now be expressed in terms of the function
𝑔 𝑡 as the infinite summation.
∞

𝑔𝑝 𝑡 = ෍ 𝑔 𝑡 − 𝑚𝑇0
−∞
Thus, 𝑔 𝑡 may be viewed as a generating function, which generates a
periodic signal 𝑔𝑝 𝑡 .
The function 𝑔 𝑡 is Fourier transformable so we can rewrite 𝐶𝑛 as
1 𝑛
𝐶𝑛 = 𝐺
𝑇0 𝑇0

𝑛 𝑛
Where 𝐺 = Fourier transform of 𝑔(𝑡) evaluated at frequency
𝑇0 𝑇0

We may rewrite the equation of 𝑔𝑝 𝑡 as

𝑗2𝜋𝑛𝑡 1 ∞ 𝑛 𝑗2𝜋𝑛𝑡
𝑔𝑝 𝑡 = σ∞
−∞ 𝐶𝑛 𝑒𝑥𝑝 = σ 𝐺 𝑒𝑥𝑝
𝑇0 𝑇0 −∞ 𝑇0 𝑇0
 ∞ ∞
1 𝑛 𝑗2𝜋𝑛𝑡
𝑔𝑝 𝑡 = ෍ 𝑔 𝑡 − 𝑚𝑇0 = ෍𝐺 𝑒𝑥𝑝
𝑇0 𝑇0 𝑇0
−∞ −∞

which defines the Poisson’s sum formula.

Using Fourier transform of a complex exponential function
1 ∞ 𝑛 𝑛
𝑔𝑝 𝑡 = σ∞
−∞ 𝑔 𝑡 − 𝑚𝑇0 σ−∞ 𝐺 𝛿 𝑓−
𝑇0 𝑇0 𝑇0

➢ Signal consists of delta functions occurring at integer multiples of the

1
fundamental frequency , including the origin
𝑇0

𝑛
➢ Each delta function is weighted by a factor equal to 𝐺
𝑇0

➢ Periodicity in the time domain has the effect of changing the spectrum of
a pulse like signal into a discrete form defined at integer multiples of the
fundamental frequency, and vice versa.
Textbook:
1. Communication Systems, by Simon Haykin, John Wiley and Sons,
Delhi.

2. Modern Digital and Analog Communication System by B. P. Lathi,

Zhi Ding, Oxford University Press