Unit 1 Introduction To Fourier Transform
Unit 1 Introduction To Fourier Transform
Taught by:
Dr. Prateek Verma
Assistant Professor
DYPIU Akurdi
1
Communication Systems
• Communication is the process of establishing connection or link between two points for
information exchange.
• The electronic equipments which are used for communication purpose are called as communication
equipments.
• Different electronic equipments when combine together form a communication system e.g. mobile
communication.
• It started since 1840s with line telegraphy, and with invention of transistors, Integrated circuits and
other semiconductor devices it became a popular field.
• With invention of satellites and fibre optics, it is reaching new heights and being used widespread.
2
Elements of a Communication System
• The communication process can be described as:
• Generation of thought pattern or image at the origin.
• The message is passed through the transducer to convert it into a time varying
electrical signal e.g. microphone
• The transmitter process the electrical signal from different aspects e.g. modulation.
3
Contd.
• Then the signal is passed through the channel or medium from transmitter to receiver.
The channel acts like a physical connection between transmitter and receiver.
• The main function of the receiver is to reproduce the message signal in electrical form
from distorted received signal i.e. demodulation.
• Destination is the final stage where the electrical signal is converted into original
form.
4
Analog Signal
• The signal which varies smoothly and continuously with time.
• These signals are defined for every value of time and they take on continuous values in a
given time interval.
• The vast majority of signals in the real world around us are analog in nature e.g.
temperature of a certain location may vary over a continuous range.
5
Digital Signal
• It is represented by a sequence of numbers where each number is representing the signal
magnitude at an instant of time. The resulting signal is called a digital signal.
• We can use any number system to represent a digital signal e.g. binary number system.
6
Need for Fourier Transform
• For transmission purpose, there is need of electrical signal because all equipments and
devices are electrical in nature.
• A signal is generally given in time domain, which tells about the amplitude of the signal at
a particular time instant.
• But if we want to know the frequency component present in the signal, we need some
mathematical tools.
• Fourier series and Fourier transform are such mathematical tools which are used to
analyze the signal (frequency domain signal).
7
Complex Numbers
• A complex number (a, b) or a + jb can be represented graphically by a point whose Cartesian
coordinates are (a, b) in a complex plane. Let us denote this complex number by z so that
𝒛 = 𝒂 + 𝒋𝒃
• The numbers a and b (the abscissa and the ordinate) of z are the real part and the imaginary part,
respectively, of z. They are also expressed as
Re z = a and Im z = b
8
Sinusoids
• Sinusoids are the time-varying sine or cosine functions (or signals).
9
Contd.
• Here, the signal is a periodic signal ranging from −∞ to ∞.
• If we shift this signal towards the left by an amount 𝑛𝑇 then new signal will be 𝑣(𝑡 + 𝑛𝑇), which will be
equal to 𝑣 𝑡 .
• If we shift this signal towards the right by an amount 𝑛𝑇 then new signal will be 𝑣(𝑡 − 𝑛𝑇), which will
be equal to 𝑣 𝑡 .
10
Contd.
• We know, 𝜔 = 2𝜋𝑓. Here, 𝜔 is the angular freq. (radians/sec) and 𝑓 is freq. in Hz.
• From here it can be said that, 𝑓 = 𝜔/2𝜋. Also, 𝑓 = 1/𝑇 where T is the time period of the signal.
• Therefore, when 𝑡 = 0; 𝜔𝑡 = 0.
𝑇 2𝜋 𝑇
• Now when 𝑡 = ; 𝜔𝑡 = ∗ =𝜋
2 𝑇 2
2𝜋
• Now when 𝑡 = 𝑇; 𝜔𝑡 = ∗ 𝑇 = 2𝜋
𝑇
11
Contd.
• Therefore, more general term for a sinusoid function can be 𝑣 𝑡 = 𝑉𝑚 sin(𝜔𝑡 ± ϕ).
• Here, 𝜔𝑡 ± ϕ is called the argument of the function and ϕ is called the phase shift (which may be leading or
lagging).
• ω governs the rate of oscillation of the signal—smaller values of ω mean that the signal is oscillating
more slowly, or that there is a longer time between peaks in the sinusoid.
13
Conditions required to compare two sinusoids
• There are three conditions which should be satisfied to compare two sinusoids:
2. They should have the same form of expressions i.e. either sine or cosine.
3. They should have the same sign in their amplitudes (if one is + and one is -, then there is a shift of
180° ).
14
Question
• Calculate the amplitude, phase, angular freq., period and freq. of the signal 𝒗(𝒕) =
𝟓𝐬𝐢𝐧(𝟒𝝅𝒕 − 𝟔𝟎°).
• We will compare this given signal with the standard form i.e. 𝑣 𝑡 = 𝑉𝑚 𝑠𝑖𝑛(𝜔𝑡 − ϕ)
• Therefore, 𝑉𝑚 = 5.
• Then, 𝜔 = 4𝜋 radians/sec.
• Phase =ϕ = 60°
2𝜋 2𝜋 1
• Time period = T = = 4𝜋 = 2 = 0.5 𝑠𝑒𝑐.
𝜔
1
• Freq. = 𝑓 = 𝑇 = 2𝐻𝑧.
15
Question
• A spring oscillates above a table as per the diagram and graph shown. Determine for
the graph:
(a) Amplitude (b) Period (c) Phase shift for sine and cosine wave (d) Vertical
Translation (e) Equation
(a) 25 cm
c. 0.8 sec towards right (cos wave) or 0.4 sec towards right (sine wave)
(b) 1.6 sec
d. 50 cm up
2𝜋
e. ℎ = 25 cos 1.6
𝑡 − 0.8 + 50 16
Question
• Calculate the phase angle between 𝑽𝟏 = −𝟏𝟎𝐜𝐨𝐬(𝝎𝒕 + 𝟓𝟎°) and 𝑽𝟐 = 𝟏𝟐𝐬𝐢𝐧(𝝎𝒕 −
𝟏𝟎°).
• Now we have to convert sine to cos or vice versa with same sign in amplitude.
• Therefore, we will perform a right shift of 90° e.g. 𝑣 𝑡 = 𝑉𝑚 𝑠𝑖𝑛𝜃 which means 𝑣 𝑡 = 𝑉𝑚 sin 𝜃 − 90° = −𝑉𝑚 𝑐𝑜𝑠𝜃
17
Signals
• A signal is a set of information or data.
• Some other examples of our daily life can be music, speech, video signals etc.
• In all these examples, the signals are functions of the independent variable time.
• In electrical signal, the signal can be voltage or current, as these are a function of time as an
independent variable.
18
Systems
• Signals may be processed further by systems, which may modify them or extract additional
information from them.
• For example, an antiaircraft gun operator may want to know the future location of a hostile moving
target that is being tracked by his radar. Knowing the radar signal, he knows the past location and
velocity of the target.
• By properly processing the radar signal (the input), he can approximately estimate the future
location of the target. Thus, a system is an entity that processes a set of signals (inputs) to yield
another set of signals (outputs).
19
Size of a Signal
• The size of any entity is a number that indicates the largeness or strength of that entity.
• How can a signal that exists over a certain time interval with varying amplitude be measured by
one number that will indicate the signal size or signal strength?
• Such a measure must consider not only the signal amplitude, but also its duration.
• For instance, if we are to devise a single number V as a measure of the size of a human being, we
must consider not only his or her width, but also the height.
20
Signal Energy
• Consider the area under a signal x(t) as a possible measure of its size, because it takes account not
only of the amplitude but also of the duration.
• However, this will be a defective measure because even for a large signal x(t), its positive and
negative areas could cancel each other, indicating a signal of small size.
• This difficulty can be corrected by defining the signal size as the area under 𝑥 2 (𝑡), which is always
positive. We call this measure the signal energy 𝐸𝑥 , defined (for a real signal) as
∞
𝐸𝑥 = 𝑥 2 𝑡 𝑑𝑡
−∞
• A necessary condition for the energy to be finite is that the signal amplitude → 0 as |t| → ∞.
Otherwise the integral will not converge.
• When the amplitude of x(t) does not → 0 as |t| → ∞, the signal energy is infinite. A more
meaningful measure of the signal size in such a case would be the time average of the energy, if it
exists. This measure is called the power of the signal.
• Indeed, the square root of 𝑃𝑥 is the familiar rms (root-mean-square) value of x(t).
• Generally, the mean of an entity averaged over a large time interval approaching infinity exists if
the entity either is periodic or has a statistical regularity.
• If such a condition is not satisfied, the average may not exist. For instance, a ramp signal x(t) = t
increases indefinitely as |t| → ∞, and neither the energy nor the power exists for this signal.
• However, the unit step function, which is not periodic nor has statistical regularity, does have a
finite power.
23
Contd.
• When x(t) is periodic, 𝑥 𝑡 2 is also periodic. Hence, the power of x(t) can be computed from by
averaging 𝑥 𝑡 2 over one period.
• A periodic signal can be a power signal or neither energy nor power signal (NENP).
• But a periodic signal cannot be an energy signal because integration of a periodic signal from
− ∞ 𝑡𝑜 + ∞, we will get infinite area.
24
Singularity Functions
• These singularity functions serve as the basic building blocks for the construction of more complex signals.
• These signals may be used to model a large number of physical signals which occur in nature.
25
Unit Step Function
• The function only exits for positive side and zero for negative side.
𝑢 𝑡 = 0 𝑡<0
1 t ≥0
• Sometimes, if we want a signal to start at t = 0 so that it may have a value of 0 for t < 0, then we will
multiply the given signal with unit-step function.
• The function is also called as Heaviside Step function, named after Oliver Heaviside. 26
Shifted Unit Step Function
• In many circuits, waveforms are applied at specified intervals other than t =0. Such a function may be
described using the shifted/delayed unit step function.
• A function which has value 0 up to the time t = a and thereafter has value 1, is written:
• For example, f(t) = u(t - 3) i.e. the equation means f(t) has value
of 0 when t < 3 and 1 when t > 3.
27
Unit Impulse Function
• An unit-impulse function is denoted as 𝛿(𝑡) and may be expressed as 𝛿 𝑡 = 0 when 𝑡 ≠ 0.
∞
• It can also be expressed as:
𝛿 𝑡 𝑑𝑡 = 1
−∞
• The unit-impulse function is also called as Dirac-Delta function.
• In Fig., the width of the pulse is zero, which means pulse exists
only at t = 0.
• The height of the pulse goes to infinity, also the area under the pulse curve is always unity.
• The total area under the pulse is called as the strength of the impulse. 28
Shifted Unit Impulse Function
• An unit-impulse function centered at 𝑡 = 𝑡0 . It is denoted by 𝛿 𝑡 − 𝑡0 .
29
Unit Ramp Function
• A ramp function or ramp signal is a type of standard signal which starts at 𝑡 = 0 and increases linearly
with time.
• The unit ramp function has unit slope and represented by r(t).
30
Classification of Signals
Continuous Time Discrete Time Non-Periodic
Periodic Signal
Signal Signal Signal
Deterministic
Random Signal Energy Signal Power Signal
Signal
• If t is a discrete variable, then x(t) is a discrete time signal, which is generally denoted by x[n] with n being
an integer.
32
Deterministic and Non-Deterministic Signal
• The signal which can be specified in time, or can be characterized mathematically, or the pattern of the
signal is regular, or the nature or amplitude of the signal can be predicted at any time is called as
Deterministic signal e.g. x(t) = bt
• A non-deterministic or random signal is the one whose occurrence is always random in nature. The pattern
of such signal is irregular e.g. thermal noise generated in electrical circuit.
33
Periodic and Aperiodic Signal
• A periodic signal has a definite pattern and repeats over and over with a repetition period of T.
• If a signal does not repeat then it is called as Aperiodic signal. Sometimes, aperiodic signal have a period =
∞
34
Even and Odd Signal
• The signal which is identical about the origin is called a even signal. This signal exhibits symmetry in the
time domain e.g. 𝑥 𝑡 = 𝑥 −𝑡 .
• The signal which does not exhibit symmetry or is anti-symmetry is called an odd signal. These signals are
not identical about its origin.
35
Energy and Power Signal
• The signal which has finite energy and zero average power is called an Energy Signal. Hence, x(t) is an
energy signal if 0 < 𝐸 < ∞ and P = 0.
• The signal which has infinite energy and finite average power is called a Power Signal. Hence, x(t) is a
power signal if 0 < 𝑃 < ∞ and E = ∞.
• However, if the signal does not satisfy any of the above two conditions, then it is neither an energy signal
nor a power signal.
36
Fourier Series (FS)
• Sine and cosine waves are the basic building functions for any periodic signal, which means that any
periodic signal basically consists of sine waves having different amplitudes of different frequencies and
having different relative phase shifts.
• Fourier series represents a periodic waveform in the form of sum of infinite number of sine and cosine
terms.
Trigonometric or
Polar FS Exponential FS
Quadrature FS
37
Trigonometric Fourier Series
• A periodic function x(t) may be expressed in the form of trigonometric Fourier Series comprising the
following sine and cosine terms:
𝑥 𝑡
= 𝑎0 + 𝑎1 𝑐𝑜𝑠𝑤𝑜 𝑡 + 𝑎2 𝑐𝑜𝑠 2𝑤0 𝑡
+ ⋯ … … … . + 𝑎𝑛 cos 𝑛𝑤0 𝑡 + ⋯ + 𝑏1 sin 𝑤0 𝑡 + 𝑏2 sin 2𝑤0 𝑡 + ⋯ + 𝑏𝑛 sin 𝑛𝑤0 𝑡
2𝜋
• Here, 𝑇0 = ; 𝑎0 , 𝑎𝑛 and 𝑏𝑛 are known as Fourier coefficients.
𝑤0
• 𝑤0 is called the fundamental frequency and 2𝑤0 , 3𝑤0 ,… are called the harmonics of 𝑤0 .
𝑡+𝑇0
2
𝑎𝑛 = 𝑥 𝑡 . cos(𝑛𝑤0 𝑡)𝑑𝑡
𝑇0
𝑡
𝑡+𝑇0
2
𝑏𝑛 = 𝑥 𝑡 . sin(𝑛𝑤0 𝑡)𝑑𝑡
𝑇0
𝑡
39
Polar Fourier Series
• The polar FS is derived from trigonometric FS by combining the sine and cosine terms of same
frequency. It is represented as follows:
𝑥 𝑡 = 𝐶0 + 𝐶𝑛 cos(𝑛𝑤0 𝑡 + ϕ𝑛 )
𝑛=1
𝑏
Where Cn = [𝑎𝑛2 + 𝑏𝑛2 ]1/2 and ϕ𝑛 = − tan−1 [ 𝑛 ]
𝑎𝑛
40
Exponential Fourier Series
• The sine and cosine terms can also be expressed in exponential terms:
𝑒 𝑗𝜃 − 𝑒 −𝑗𝜃 𝑒 𝑗𝜃 + 𝑒 −𝑗𝜃
sin 𝜃 = cos 𝜃 =
2 2
∞
𝑥 𝑡 = 𝐶𝑛 𝑒 𝑗2𝜋𝑛𝑡/𝑇0
𝑛=−∞
1 𝑡+𝑇0
Where Cn = 𝑥 𝑡 . 𝑒 −𝑗2𝜋𝑛𝑡/𝑇0 𝑑𝑡
𝑇0 𝑡
• Here, n is extending from -∞ to ∞ instead of 0 to ∞. Due to this, frequencies in frequency spectrum will
extend from -∞ to ∞.
• If we express a signal using exponential Fourier Series, we obtain a double sided frequency spectrum.
• However, negative freq. signals does not exist physically. They are just used as a mathematical concept
41
and for convenience.
Dirichlet’s conditions in Fourier Series
• FS is used for analysis of Periodic signal while Fourier Transform is used for aperiodic signal.
• The continuous-time Fourier series exists only when the function x(t) satisfy the following three
conditions called as Dirichlet’s conditions.
a) 𝑥(𝑡) must possess only a finite number of discontinuities in the period T
b) 𝑥(𝑡) must have a finite number of positive and negative maxima in the period T.
c) 𝑥(𝑡) should be absolutely integrable over the range of time period.
42
Symmetry Conditions
• If the function x(t) is even, then x(-t) = x(t), e.g. cos t, 𝑡 2
• The graph of even function is symmetrical w.r.t. y axis and in FS representation of even function, only
cosine terms are present (dc term is optional).
• Mathematically, 𝑎 𝑎
𝑥 𝑡 𝑑𝑡 = 2 𝑥 𝑡 𝑑𝑡
−𝑎 0
43
Contd.
• If the function x(t) is odd, then x(-t) = -x(t), e.g. sin t, 𝑡 3
• Mathematically, 𝑎
𝑥 𝑡 𝑑𝑡 = 0
−𝑎
• The sum or product of two or more even functions is an
even function.
44
Properties of Continuous Time FS
Time Conjuga
tion Linearity
Different
iation
Time-
Multipl shifting
ication
Frequency
Time shifting
Scaling
Symmetry
45
Linearity
• Let there are two signals 𝑥1 (𝑡) and 𝑥2 (𝑡). Then, the complex exponentials of these signals are 𝐶1𝑛 and 𝐶2𝑛 .
• Both the signals are having fundamental time period 𝑇0 . Since performing scaling function and adding them
results in
𝛼. 𝑥1 𝑡 + 𝛽. 𝑥2 𝑡 = 𝑥(𝑡).
• Therefore, complex exponential of 𝑥(𝑡) ↔ 𝐶𝑛
1 𝑇0 1 𝑇0
• Since we know, 𝐶𝑛 = 𝑥 𝑡 . 𝑒 −𝑗𝑛𝜔0𝑡 𝑑𝑡 = [𝛼. 𝑥1 𝑡 + 𝛽. 𝑥2 𝑡 ]. 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡
𝑇0 0 𝑇0 0
1 𝑇0 1 𝑇0
Cn = [𝛼 ∗ 𝑥1 𝑡 . 𝑒 −𝑗𝑛𝜔0𝑡 𝑑𝑡]+[𝛽 ∗ 𝑥2 𝑡 . 𝑒 −𝑗𝑛𝜔0𝑡 𝑑𝑡]
𝑇0 0 𝑇0 0
𝐶1𝑛 𝐶2𝑛
• Therefore, 𝑪𝒏 = 𝜶𝐂𝟏𝐧 + 𝜷𝑪𝟐𝒏
• Since this is not a formula of complex exponential Fourier coefficient. Therefore, replacing n by –n.
𝑇0
∗
1
𝐶−𝑛 = 𝑥 ∗ 𝑡 . 𝑒 −𝑗𝑛𝜔0𝑡 𝑑𝑡
𝑇0
0
∗ ∗
𝑥 𝑡 ↔ 𝐶−𝑛
• Therefore, if we take complex conjugate of signal x 𝑡 , then its complex exponential Fourier expansion will
∗
become 𝐶−𝑛
47
Time Reversal
• Let there is a signal with fundamental time period 𝑇0 , 𝑥(𝑡) ↔ 𝐶𝑛
1 𝑇0
• Since we know, 𝐶𝑛 = 𝑥 𝑡 . 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡
𝑇0 0
• We will perform the reversal operation by subjecting 𝑥(𝑡) to a system performing reversal and output will be
𝑥(−𝑡).
• The operation time reversal is only for 𝑥(𝑡). The t in exponential part will not change (because time reversal
operation will not expand or compress the signal waveform, it will just flip the waveform w.r.t. y axis.
Therefore, whenever 𝒙(𝒕) will go time reversal 𝒙 −𝒕 , then its complex exponential Fourier coefficient will also
48
undergo time reversal operation i.e. 𝐂−𝐧
Time Scaling
• Let there is a signal with fundamental time period 𝑇0 , 𝑥(𝑡) ↔ 𝐶𝑛
1 𝑇0
• Since we know, 𝐶𝑛 = 𝑥 𝑡 . 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡
𝑇0 0
• We will perform the time scaling operation by subjecting 𝑥(𝑡) to a system performing scaling and output will
be 𝑥 𝛼𝑡 ; Al𝑠𝑜 α > 1 which means signal is compressing.
• Hence, time period will reduce and freq. will increase. Therefore new time period is 𝑇0′ = 𝑇0 /𝛼
2𝜋 2𝜋𝛼
• Angular Freq. 𝜔′0 = = = 𝜔0 𝛼.
𝑇0′ 𝑇0
𝑇0′ Let α𝑡 = 𝜏, then α 𝑑𝑡 = 𝑑𝜏 𝑜𝑟 𝑑𝑡 = 𝑑𝜏/α
1 If 𝑡 = 0, then 𝜏 = 0
𝐶𝑛′ = 𝑥 𝛼𝑡 . 𝑒 −𝑗𝑛𝜔′0𝑡 𝑑𝑡
𝑇′0 If 𝑡 = 𝑇0′ = 𝑇0/α , then 𝜏 = 𝑇0
0
𝑇0 𝑇0
α 1
𝐶𝑛′ = 𝑥 𝜏 . 𝑒 −𝑗𝑛𝜔0α𝜏/α 𝑑𝜏/α 𝐶𝑛′ = 𝑥 𝜏 . 𝑒 −𝑗𝑛𝜔0𝜏 𝑑𝜏 𝐶𝑛
𝑇0 𝑇0
0 0
Therefore, whenever 𝒙(𝒕) will go time scaling 𝒙 𝜶𝒕 , then its complex exponential Fourier coefficient will not
49
undergo any change i.e. 𝐂𝐧 but 𝑻𝟎 will change.
Time Shifting
• Let there is a signal with fundamental time period 𝑇0 , 𝑥(𝑡) ↔ 𝐶𝑛
1 𝑇0
• Since we know, 𝐶𝑛 = 𝑥 𝑡 . 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡
𝑇0 0
• We will perform the time shifting operation by subjecting 𝑥(𝑡) to a system performing time shifting and
output will be 𝑥 𝑡 − 𝑡0 which means right shifting is performed. Also, time period and angular freq. will not
change.
𝑇0 Let 𝑡 − 𝑡0 = 𝜏, then 𝑑𝑡 = 𝑑𝜏
1
𝐶𝑛′ = 𝑥 𝑡 − 𝑡0 . 𝑒 −𝑗𝑛𝜔0𝑡 𝑑𝑡 If 𝑡 = 0, then 𝜏 = −𝑡0
𝑇0 If 𝑡 = 𝑇0 , then 𝜏 = 𝑇0 − 𝑡0
0 𝑇0 −𝑡0
𝑒 −𝑗𝑛𝜔0 𝑡0
𝐶𝑛′ = 𝑒 −𝑗𝑛𝜔0 𝑡0 . 𝐶𝑛
• We will perform the frequency shifting by an amount m i.e. 𝐶𝑛−𝑚 and let new time domain signal be 𝑥′(𝑡).
𝑇0 𝑇0
1 1
𝐶𝑛−𝑚 = 𝑥 𝑡 . 𝑒 −𝑗(𝑛−𝑚)𝜔0 𝑡 𝑑𝑡 𝐶𝑛−𝑚 = 𝑥 𝑡 . 𝑒 +𝑗𝑚𝜔0𝑡 . 𝑒 −𝑗𝑛𝜔0𝑡 𝑑𝑡
𝑇0 𝑇0
0 0
𝒙′(𝒕)
51
Question 1
• Obtain the Trigonometric Fourier series for the rectangular pulse train shown here.
𝜏 𝜏
• We will consider the time period from − to
2 2
𝐴 𝜏 𝜏 𝑨𝝉
𝑎0 = [ − (− )] 𝒂𝟎 = 52
𝑇0 2 2 𝑻𝟎
Contd.
𝑇0/2 𝜏/2
2 2𝜋𝑛𝑡 2 2𝜋𝑛𝑡
• Now estimate 𝑎𝑛 . 𝑎𝑛 = 𝑥 𝑡 cos[ ]𝑑𝑡 𝑎𝑛 = 𝑥 𝑡 cos[ ]𝑑𝑡
𝑇0 𝑇0 𝑇0 𝑇0
−𝑇0 /2 −𝜏/2
𝜏 𝜏/2
2𝐴 1 2𝜋𝑛𝑡 2 2𝐴 2𝜋𝑛𝑡
𝑎𝑛 = ∗ sin 𝑎𝑛 = cos[ ]𝑑𝑡
𝑇0 2𝜋𝑛 𝑇0 −𝜏/2 𝑇0 𝑇0
𝑇0 −𝜏/2
𝐴 𝜋𝑛𝜏 𝜋𝑛𝜏
𝑎𝑛 = sin + sin 𝟐𝑨 𝝅𝒏𝝉
𝑛𝜋 𝑇0 𝑇0 𝒂𝒏 = 𝐬𝐢𝐧
𝒏𝝅 𝑻𝟎
𝑇0/2 𝜏/2
2 2𝜋𝑛𝑡 2𝐴 2𝜋𝑛𝑡
• Now estimate 𝑏𝑛 . 𝑏𝑛 = 𝑥 𝑡 sin[ ]𝑑𝑡 𝑏𝑛 = sin[ ]𝑑𝑡
𝑇0 𝑇0 𝑇0 𝑇0
−𝑇0 /2 −𝜏/2
𝜏
−2𝐴 1 2𝜋𝑛𝑡 2
−𝐴 𝜋𝑛𝜏 𝜋𝑛𝜏 𝑏𝑛 = ∗ cos
𝒃𝒏 = 𝟎 𝑏𝑛 = cos − 𝑐𝑜s 𝑇0 2𝜋𝑛 𝑇0
𝑛𝜋 𝑇0 𝑇0 −𝜏/2
𝑇0
53
Question 2
• Obtain the Exponential Fourier series for the rectangular pulse train shown here.
𝜏 𝜏
• We will consider the time period from − to
2 2
• Estimate Cn .
𝑡+𝑇0 𝜏/2
𝑗2𝜋𝑛𝑡 1 𝑗2𝜋𝑛𝑡
1 − 𝑇 − 𝑇
Cn = 𝑥 𝑡 .𝑒 0 𝑑𝑡 Cn = 𝐴 ∗𝑒 0 𝑑𝑡
𝑇0 𝑇0
𝑡 −𝜏/2
54
Contd.
𝜏
𝜏/2 𝜏/2 𝑗2𝜋𝑛𝑡 2
1 −
𝑗2𝜋𝑛𝑡 𝐴 −
𝑗2𝜋𝑛𝑡 −𝐴 1 −
Cn = 𝐴∗𝑒 𝑇0 𝑑𝑡 Cn = 𝑒 𝑇0 𝑑𝑡 Cn = ∗ 𝑒 𝑇0 −𝜏
𝑇0 𝑇0 𝑇0 2𝑗𝜋𝑛 2
−𝜏/2 −𝜏/2 𝑇0
𝑗2𝜋𝑛𝜏
−
𝑗2𝜋𝑛𝜏 −𝐴 −𝑗2𝜋𝑛𝜏 𝑗2𝜋𝑛𝜏
𝑒 𝑗𝜃 − 𝑒 −𝑗𝜃
𝑆𝑖𝑛𝑐𝑒, 𝑠𝑖𝑛𝜃 =
2𝑗 𝝉 𝝅𝒏𝝉
𝜏
𝐴∗ 𝜋𝑛𝜏 𝑪𝐧 = 𝑨 ∗ ∗ 𝒔𝒊𝒏𝒄( )
Now, multiply and divide RHS by 𝜏/𝑇0
𝑇0 𝑻𝟎 𝑻𝟎
𝐶n = 𝜏 sin( 𝑇 )
𝜋𝑛 ∗ 0
𝑇0
55
• The function (sin 𝑥)/𝑥 is the sine over argument function and is denoted by
sin 𝑐(𝑥) .
• sin 𝑐 0 = 1 at t=0.
56
Question 3
• Obtain the Trigonometric Fourier series for the pulse train shown here.
𝑦2 − 𝑦1
𝑦= ∗𝑥
𝑥2 − 𝑥1
1−0 1
𝑦= ∗𝑥 = 𝑥
2𝜋 − 0 2𝜋
𝒕
• Now, Put y = x(t) and x = t in above equation. 𝒙(𝒕) =
𝟐𝝅
2𝜋 2𝜋 2π
• Now, solve for 𝑎0 1 1 𝑡 1 t2
𝑎0 = 𝑥 𝑡 𝑑𝑡 𝑎0 = 𝑑𝑡 𝑎0 = 2
2𝜋 2𝜋 2𝜋 4𝜋 2 0
0 0
2π
1 4π2
𝟏 𝑎0 = 2 −0
𝒂𝟎 = 4𝜋 2 0
𝟐
58
𝑇0/2 2𝜋
• Now, solve for 𝑎𝑛 2 2 𝑡
𝑎𝑛 = 𝑥 𝑡 cos 𝑛𝜔0 𝑡𝑑𝑡 𝑎𝑛 = cos 𝑛𝜔0 𝑡𝑑𝑡
𝑇0 𝑇0 2𝜋
−𝑇0 /2 0
• First estimate 𝜔0 2𝜋 2𝜋
𝜔0 = = =1
𝑇0 2𝜋
t is algebraic function;
cosnt is trigonometric function
2𝜋 2𝜋
• Now replace this value 2 𝑡 1
𝑎𝑛 = cos 𝑛𝑡𝑑𝑡 𝑎𝑛 = 𝑡cos 𝑛𝑡𝑑𝑡
2𝜋 2𝜋 2𝜋 2
0 0
• If we have two functions, 𝑢𝑣 𝑑𝑥; here u is first function and v is second function
𝑑𝑢
𝑢𝑣 𝑑𝑥 = 𝑢 𝑣𝑑𝑥 − [ 𝑣𝑑𝑥] 𝑑𝑥
𝑑𝑥
• It is important to choose first function carefully, so that integration will be easy. Follow ILATE
59
2𝜋
1 𝑑𝑢
𝑎𝑛 = 2 𝑡cos 𝑛𝑡𝑑𝑡 𝑢𝑣 𝑑𝑥 = 𝑢 𝑣𝑑𝑥 − [ 𝑣𝑑𝑥] 𝑑𝑥
2𝜋 𝑑𝑥
0
𝑑𝑡 𝑡𝑠𝑖𝑛𝑛𝑡 𝑠𝑖𝑛𝑛𝑡
𝑡𝑐𝑜𝑠𝑛𝑡𝑑𝑡 = 𝑡 𝑐𝑜𝑠𝑛𝑡𝑑𝑡 − [ 𝑐𝑜𝑠𝑛𝑡𝑑𝑡] 𝑑𝑡 = − 1∗ 𝑑𝑡
𝑑𝑡 𝑛 𝑛
𝟏
𝒃𝒏 = −
• Here, first term = 0; second term = 0; third term = 1; fourth term = 1 𝒏𝝅
• Non-periodic signals are conveniently represented by Fourier Transform in the frequency domain.
∞
𝑋 𝜔 =𝐹 𝑥 𝑡 = 𝑥 𝑡 𝑒 −𝑗𝜔𝑡 𝑑𝑡
−∞
or
∞
𝑋 𝑓 = 𝑥 𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
−∞
61
Contd.
• 𝑋 𝑓 or 𝑋 𝜔 are called the Fourier transform of x 𝑡 . These are also called as the frequency domain
representation of time domain function x 𝑡 .
• Therefore, we are converting a time domain signal into its frequency domain representation with the
help of FT.
∞
Here, 𝒙 𝒕 and 𝐗 𝝎 are FT pair 𝑥 𝑡 = 𝑋(𝑓)𝑒 𝑗2𝜋𝑓𝑡 𝑑𝑓
−∞ 62
Contd.
FT
• Symbolically, this may be expressed as 𝑥(𝑡) ↔ 𝑋(𝜔)
• Where |𝑋(𝜔)| is called the amplitude spectrum and < 𝑋(𝜔) is called the phase spectrum.
63
Properties of Fourier Transform
• Some of the important
properties of FT are
listed as follows:
64
Parseval's Power Theorem
• The Parseval's power theorem states that the total average power of a periodic signal 𝑥(𝑡) is equal to the
sum of the average powers of the individual Fourier coefficients i.e. 𝐶𝑛
or ∞
2
𝑃= 𝐶𝑛
𝑛=−∞
65
Spectral Density Functions
• The spectral density of a signal defines the distribution of energy and power per unit bandwidth as a
function of frequency.
• The spectral density of energy signals is called as Energy spectral density (ESD) while that of the power
signal is called as Power spectral density (PSD).
66
Power Spectral Density (PSD)
• Let 𝑥 𝑡 be a real valued power signal. Then, the average normalized power of this signal may be
described as under:
1 𝑇/2 2
𝑃 = lim 𝑥 𝑡 𝑑𝑡
𝑇→∞ 𝑇 −𝑇/2
• If 𝑥 𝑡 is periodic with a period T0 , Then according to Parseval’s power theorem, the average signal
power is given by
∞
𝑃= 𝐶𝑛 2
𝑛=−∞
67
Contd.
• Substituting the value of 𝐶𝑛 in power equation
∞
1 2
𝑃= 2 𝑋 𝑛𝑓0
𝑇0
𝑛=−∞
• Therefore,
∞
1 2
𝑃= 2 𝑋 𝑛𝑓0
𝑇0
𝑛=−∞
• But the FT of a periodic signal is not continuous, it is discrete and present only at frequencies
+ 𝑓0 , +2𝑓0 …
∞
1
𝑆(𝑓) = 2 𝑋 𝑛𝑓0 2 𝛿(𝑓 − 𝑛𝑓0 )
𝑇0
𝑛=−∞
• This relation shows that PSD is a discrete function of frequency indicated by 𝛿 𝑓 − 𝑛𝑓0 .
• The average power of a periodic signal is given by the area under the power spectral density function.
68
Frequency domain
1
• Mathematically, 𝑓𝑟𝑒𝑞 =
𝑇0
• Or we can also say that, Frequency is how fast the signal is changing.
69
Freq. distribution of slow song Freq. distribution of fast song
70
Relation between Time & Frequency domain
Case 1 Suppose if I say, “Hi, this is Communication subject”
Case 2 If I will increase the speed 2 times for speaking same line
73
Freq. domain representation of the signal
Signal
Filter
0 0
∞
The FT is represented as:
𝐹[𝑥 𝑡 ] = 𝑒 −𝛼𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
−∞
∞ ∞
𝐹[𝑥 𝑡 ] = 𝑒 −𝛼𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡 𝐹[𝑥 𝑡 ] = 𝑒 −(𝛼+𝑗2𝜋𝑓)𝑡 𝑑𝑡 1 𝛼+𝑗2𝜋𝑓 𝑡 ∞
0 0 =− 𝑒−
𝛼 + 𝑗2𝜋𝑓 0
𝟏 1 1
𝑭[𝒙(𝒕)] = = − 𝛼+𝑗2𝜋𝑓 [0 − 1] = − 𝛼+𝑗2𝜋𝑓 [𝑒 −∞ −𝑒 0 ]
𝜶 + 𝒋𝟐𝝅𝒇 75
17/3/2022
Tutorial
2. Find the Fourier Transform of the decaying exponential pulse shown here.
𝑥 𝑡 = 𝑒 𝛼𝑡 for t < 0
=0 for t > 0
∞
The FT is represented as:
𝐹[𝑥 𝑡 ] = 𝑒 𝛼𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
−∞
0 0
𝐹[𝑥 𝑡 ] = 𝑒 𝛼𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡 𝐹[𝑥 𝑡 ] = 𝑒 (𝛼−𝑗2𝜋𝑓)𝑡 𝑑𝑡 1 𝛼−𝑗2𝜋𝑓 𝑡 0
−∞
= 𝑒
−∞ 𝛼 − 𝑗2𝜋𝑓 −∞
𝟏 1 1
𝑭[𝒙(𝒕)] = = 𝛼−𝑗2𝜋𝑓 [1 −0] = 𝛼−𝑗2𝜋𝑓 [𝑒 0 −𝑒 −∞]
𝜶 − 𝒋𝟐𝝅𝒇 76
17/3/2022
Tutorial
2. Find the Fourier Transform of the decaying exponential pulse shown here.
= −𝑒 𝛼𝑡 for t < 0
𝑥 𝑡 = |1| for t = 0
= 𝑒 −𝛼𝑡 for t > 0
The FT is represented as:
0 0 ∞
𝐹[𝑥 𝑡 ] = −𝑒 𝛼𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡 + 1 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡 + 𝑒 −𝛼𝑡 𝑒 −𝑗2𝜋𝑓𝑡 𝑑𝑡
−∞ 0 0
1 𝛼−𝑗2𝜋𝑓 𝑡 0
1 𝛼+𝑗2𝜋𝑓 𝑡 ∞
=− 𝑒 + 𝑒−
0 ∞ 𝛼 − 𝑗2𝜋𝑓 ∞ −(𝛼 + 𝑗2𝜋𝑓) 0
𝐹[𝑥 𝑡 ] = −𝑒 (𝛼−𝑗2𝜋𝑓)𝑡 𝑑𝑡 + 𝑒 −(𝛼+𝑗2𝜋𝑓)𝑡 𝑑𝑡
−∞ 0
−𝒋𝟒𝝅𝒇 1 1 1 1
𝑿(𝒇) = − 𝟐 =− + =− +
𝜶 + 𝟒𝝅𝟐 𝒇𝟐 𝛼 − 𝑗2𝜋𝑓 𝛼 + 𝑗2𝜋𝑓 𝛼 − 𝑗2𝜋𝑓 𝛼 + 𝑗2𝜋𝑓
3. Evaluate the energy of a signal shown here.
∞ 1 −2𝛼𝑡
−2𝛼𝑡 ∞
𝐸= 𝑒 𝑑𝑡 𝐸=− 𝑒 0
0 2𝛼
𝟏
𝑬=
𝟐𝜶
78
3. Evaluate the energy of a signal shown here.
0 1 2𝛼𝑡
2𝛼𝑡 𝐸= 𝑒 0
𝐸= 𝑒 𝑑𝑡 ∞
−∞
2𝛼
𝟏
𝑬=
𝟐𝜶
79
3. Evaluate the energy of a signal shown here.
1
We know, the energy of signal for t > 0 is .
2𝛼
1
We also know that the energy of a signal for t < 0 is .
2𝛼
𝐸 = 𝐸𝑙 + 𝐸𝑟
1 1
𝐸= +
2𝛼 2𝛼
𝟏
𝑬=
𝜶
80
Filtering
• A filter is a device that suppresses unwanted components or features from a signal.
• This usually means removing some frequencies to suppress interfering signals and to reduce
background noise.
• The most commonly used filters are low-pass, high-pass, band-pass and band-stop.
• Characteristics that describe filter are its type, cutoff frequency, etc.
• Cutoff frequency - the frequency beyond which the filter will not pass a signal.
81
Types of Filters
1
1
0 0
-5 5 -10 10
1 1
-10 -5 5 10 -10 -5 5 10
• System in which the output depends • System in which the output of the
only on the present values of input. system depends upon the past or future
values of input at any instant of time.
• For example, 𝑦 𝑡 = 𝑥 𝑡 or 𝑦 𝑡 =
2𝑥 𝑡 • For example, 𝑦 𝑡 = 𝑥(𝑡 + 2)
• 𝑦 𝑡 = 𝑥 𝑡 + 𝑥 𝑡 − 1 + 𝑥(𝑡 + 1)
84
Causal System Non-Causal System
• System in which the output of the
• System in which the output is system depends upon the future values
independent of future values of input. of input at any instant of time.
• The system may depend upon present • It may depend upon past and present
and past input. value but it should also depend upon the
future values.
• All real life systems or practical
systems are causal systems. • For example, 𝑦 𝑡 = 𝑥(𝑡 + 2)
= 𝑦(𝑡 − 𝑡0 )
𝑥 𝑡 − 𝑡0 𝑦′ 𝑡
Delay by 𝑡0 System ≠ 𝑦(𝑡 − 𝑡0 )
• A Time-Invariant system is a system in which any delay provided in the input must be reflected in
the output.
• For example, 𝑦 𝑡 = 𝑥 2𝑡 ; which means 𝑥 𝑡 is the input and output is 𝑥 2𝑡 which is equal to 𝑦 𝑡 .
We need to find whether the system is time-invariant or time-variant.
• Step 1: provide delay to output i.e. 𝑦 𝑡 − 𝑡0 = 𝑥 2 𝑡 − 𝑡0 = 𝑥(2𝑡 − 2𝑡0 )
• Step 2: provide delay to input i.e. 𝑥 𝑡 − 𝑡0 and provide this to system which perform time scaling i.e.
𝑥 2𝑡 − 𝑡0 system performs time scaling only on t and 𝑡0 is constant.
• Compare output from step 1 and 2. Both are different and hence system is time variant system. 86
Linear and Non-Linear system
• The system which follows the principle of superposition is known as linear system.
• The system which do not follow the principle of superposition is known as non-linear system.
• Law of Superposition includes two laws i.e. Law of Additivity and law of Homogeneity.
Law of Additivity 𝑥1 𝑡 𝑦1 𝑡
System
𝑦1 𝑡 + 𝑦2 (𝑡)
+
𝑥2 𝑡 𝑦2 𝑡
System
𝑦′ 𝑡 = 𝑦1 𝑡 + 𝑦2 (𝑡)
𝑥1 𝑡
+ System ≠ 𝑦1 𝑡 + 𝑦2 (𝑡)
87
Law of Homogeneity
𝑥 𝑡 𝑦 𝑡 𝑘𝑦 𝑡
System k
= 𝑘𝑦 𝑡
𝑘𝑥 𝑡 𝑦′ 𝑡
k System ≠ 𝑘𝑦 𝑡
88
Invertible and Non-Invertible System
• Before studying invertible and non-invertible systems, we will see one to one mapping and many to one
mapping.
1 a 1
2 b 2 a
3 b 3 b
• For a non-invertible system, there should be many to one mapping between input and output.
• For a stable system, o/p should be bounded for bounded i/p at each and every instant of time.
• For a unstable system, the o/p of the system is unbounded when you provide a bounded i/p.
• Example of Stable system, a dc value; sin t or cos t (amplitude ranges from -1 to +1); unit step function.
• Example of Unstable system, 𝑦 𝑡 = 𝑡𝑥(𝑡); Let the i/p be u(t) then 𝑦 𝑡 = 𝑡𝑢 𝑡 = 𝑟(𝑡) ramp function,
which is unbounded o/p when i/p is bounded.
90
END OF UNIT 1…THANK YOU!!
91