LECTURE NOTES

ON

SIGNLAS AND SYSTEMS

(AECB14-R18)

B. TECH IV Semester ECE

(AUTONOMOUS-R18)

Ms.V.Bindusree
(Assistant Professor)

ELECRTONICS AND COMMUNICATION ENGINEERING

INSTITUTE OF AERONAUTICAL ENGINEERING
(Autonomous)
DUNDIGAL, HYDERABAD - 500043
SYLLABUS:

MODULE -I SIGNAL ANALYSIS Classes: 09

Signal Analysis: Analogy between Vectors and Signals, Orthogonal Signal Space, Signal approximation using
Orthogonal functions, Mean Square Error, Closed or complete set of Orthogonal functions, Orthogonally in
Complex functions, Exponential and Sinusoidal signals, Concepts of Impulse function, Unit Step function,
Signum function.
MODULE - II FOURIER SERIES Classes: 09

Representation of Fourier series, Continuous time periodic signals, Properties of Fourier Series, Dirichlet’s
conditions, Trigonometric Fourier Series and Exponential Fourier Series, Complex Fourier spectrum.
Fourier Transforms:
Deriving Fourier Transform from Fourier series, Fourier Transform of arbitrary signal, Fourier Transform of
standard signals, Fourier Transform of Periodic Signals, Properties of Fourier Transform, Fourier Transform
involving Impulse function and Signum function, Introduction to Hilbert Transforms
MODULE - III SIGNAL TRANSMISSION THROUGH LINEAR Classes: 12
SYSTEMS
Linear System, Impulse response, Response of a Linear System, Linear Time Invariant (LTI) System, Linear
Time Variant (LTV) System, Transfer function of a LTI system, Filter characteristics of Linear Systems,
Distortion less transmission through a system, Signal bandwidth, System bandwidth, Ideal LPF, HPF and BPF
characteristics.
Causality and Paley-Wiener criterion for physical realization, Relationship between Bandwidth and rise time, Convolution
and Correlation of Signals, Concept of convolution in Time domain and Frequency domain, Graphical representation of
Convolution.
MODULE - IV LAPLACE TRANSFORM AND Z-TRANSFORM Classes: 09

Laplace Transforms: Laplace Transforms (L.T), Inverse Laplace Transform, Concept of Region of
Convergence (ROC) for Laplace Transforms, Properties of L.T, Relation between L.T and F.T of a signal,
Laplace Transform of certain signals using waveform synthesis.
Z–Transforms: Concept of Z- Transform of a Discrete Sequence, Distinction between Laplace, Fourier and Z
Transforms, Region of Convergence in Z-Transform, Constraints on ROC for various classes of signals, Inverse
Z-transform, Properties of Z-transforms
MODULE - V SAMPLING THEOREM Classes: 06
Graphical and analytical proof for Band Limited Signals, Impulse Sampling, Natural and Flat top Sampling, Reconstruction
of signal from its samples, Effect of under sampling – Aliasing, Introduction to Band Pass Sampling. Correlation: Cross
Correlation and Auto Correlation of Functions, Properties of Correlation Functions, Energy Density Spectrum, Parseval’s
Theorem, Power Density Spectrum, Relation between Autocorrelation Function and Energy/Power Spectral Density
Function, Relation between Convolution and Correlation, Detection of Periodic Signals in the presence of Noise by
Correlation, Extraction of Signal from Noise by filtering
Text Books:
1. Signals, Systems & Communications, B.P. Lathi, BS Publications, 2009.
2. Signals and Systems, A.V. Oppenheim, A.S. Willsky and S.H. Nawab ,PHI, 2nd Edition 2009.
3. Digital Signal Processing, Principles, Algorithms, and Applications, John G. Proakis, Dimitris G. Manolakis,
Pearson Education / PHI. 2007.
Reference Books:
1. Signals & Systems, Simon Haykin and Van Veen, Wiley, 2nd Edition, 2009.
2. Signals and Signals, Iyer and K. Satya Prasad, Cengage Learning, 2 nd Edition, 2009.
3. Discrete Time Signal Processing, A. V. Oppenheim and R.W. Schaffer, PHI, 2009.
4. Fundamentals of Digital Signal Processing, Loney Ludeman. John Wiley, PHI, 2009.
COURSE OBJECTIVES

CO 1 Summarize the basic signals exponential, sinusoidal, impulse, unit step and signum for
performing mathematical operations on signals.
CO 2 Demonstrate the concepts of vector algebra for approximating a signal with the orthogonal
functions.

MAPPING OF COs WITH POs and PSOs FOR MODULE-I:

Program
Course Program Outcomes Specific
Outcomes Outcomes
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3
CO 1 √ √ - - - - - - - - - - - - -
CO 2 √ - - - - - - - - - - - - - -
 MODULE – I
SIGNAL ANALYSIS
Analogy between Vectors and Signals, Orthogonal Signal Space, Signal approximation using Orthogonal
functions, Mean Square Error, Closed or complete set of Orthogonal functions, Orthogonality in Complex
functions, Classification of Signals and systems, Exponential and Sinusoidal signals, Concepts of Impulse
function, Unit Step function, Signum function.

Analogy Between Vectors and Signals:

There is a perfect analogy between vectors and signals.

Vector
A vector contains magnitude and direction. The name of the vector is denoted by bold face type and their magnitude
is denoted by light face type.

Example: V is a vector with magnitude V. Consider two vectors V1 and V2 as shown in the following diagram. Let
the component of V1 along with V2 is given by C12V2. The component of a vector V1 along with the vector V2 can
obtained by taking a perpendicular from the end of V1 to the vector V2 as shown in diagram:

The vector V1 can be expressed in terms of vector V2

V1= C12V2 + Ve

Where Ve is the error vector.

But this is not the only way of expressing vector V1 in terms of V2. The alternate possibilities are:

V1=C1V2+Ve1
V2=C2V2+Ve2

The error signal is minimum for large component value. If C12=0, then two signals are said to be orthogonal.

Dot Product of Two Vectors V1 . V2 = V1.V2 cosθ

θ = Angle between V1 and V2 V1. V2 =V2.V1
From the diagram, components of V1 a long V2 = C 12 V2

The concept of orthogonality can be applied to signals. Let us consider two signals f1(t) and f2(t).
Similar to vectors, you can approximate f1(t) in terms of f2(t) as f1(t) = C12 f2(t) + fe(t) for (t1 < t < t2)
⇒ fe(t) = f1(t) – C12 f2(t)

One possible way of minimizing the error is integrating over the interval t1 to t2.

However, this step also does not reduce the error to appreciable extent. This can be corrected by taking the square of
error function.

Where ε is the mean square value of error signal. The value of C12 which minimizes the error, you need to calculate
dε/dC12=0

Derivative of the terms which do not have C12 term are zero.

Put C12 = 0 to get condition for orthogonality.

Orthogonal Vector Space

A complete set of orthogonal vectors is referred to as orthogonal vector space. Consider a three dimensional vector
space as shown below:

Consider a vector A at a point (X1, Y1, Z1). Consider three unit vectors (VX, VY, VZ) in the direction of X, Y, Z
axis respectively. Since these unit vectors are mutually orthogonal, it satisfies that
We can write above conditions as

The vector A can be represented in terms of its components and unit vectors as

Any vectors in this three dimensional space can be represented in terms of these three unit vectors only.

If you consider n dimensional space, then any vector A in that space can be represented as

As the magnitude of unit vectors is unity for any vector A The component of A along x axis = A.VX
The component of A along Y axis = A.VY The component of A along Z axis = A.VZ

Similarly, for n dimensional space, the component of A along some G axis

=A.VG (3)
Substitute equation 2 in equation 3.

Orthogonal Signal Space

Let us consider a set of n mutually orthogonal functions x1(t), x2(t)... xn(t) over the interval t1 to t2. As these
functions are orthogonal to each other, any two signals xj(t), xk(t) have to satisfy the orthogonality condition. i.e.
Let a function f(t), it can be approximated with this orthogonal signal space by adding the components along
mutually orthogonal signals i.e.

The component which minimizes the mean square error can be found by

All terms that do not contain Ck is zero. i.e. in summation, r=k term remains and all other terms are zero.
Mean Square Error:

The average of square of error function fe(t) is called as mean square error. It is denoted by ε (epsilon).

The above equation is used to evaluate the mean square error.

Closed and Complete Set of Orthogonal Functions:

Let us consider a set of n mutually orthogonal functions x1(t), x2(t)...xn(t) over the interval t1 to t2. This is called as
closed and complete set when there exist no function f(t) satisfying the condition

If this function is satisfying the equation

For k=1,2,.. then f(t) is said to be orthogonal to each and every function of orthogonal set.
This set is incomplete without f(t). It becomes closed and complete set when f(t) is included.
f(t) can be approximated with this orthogonal set by adding the components along mutually orthogonal signals i.e.

Orthogonality in Complex Functions:

If f1(t) and f2(t) are two complex functions, then f1(t) can be expressed in terms of f2(t) as

f1(t)=C12f2(t).. with negligible error

Where f2*(t) is the complex conjugate of f2(t) If f1(t) and f2(t) are orthogonal then C12 = 0

The above equation represents orthogonality condition in complex functions.

Ramp Signal

Ramp signal is denoted by r(t), and it is defined as r(t) =

Area under unit ramp is unity.

Parabolic Signal

Parabolic signal can be defined as x(t) =

Signum Function

Signum function is denoted as sgn(t). It is defined as sgn(t) =

sgn(t) = 2u(t) – 1

Exponential Signal

Exponential signal is in the form of x(t) = eαt

.The shape of exponential can be defined by α
Case i: if α = 0 → x(t) = e0= 1
Case ii: if α< 0 i.e. -ve then x(t) = e−αt
. The shape is called decaying exponential.

Case iii: if α> 0 i.e. +ve then x(t) = eαt

. The shape is called raising exponential.

Rectangular Signal

Let it be denoted as x(t) and it is defined as

Triangular Signal

Let it be denoted as x(t)

Sinusoidal Signal
Sinusoidal signal is in the form of x(t) = A cos(w0±ϕ) or A sin(w0±ϕ)

Where T0 = 2π/w0

Classification of Signals:

Signals are classified into the following categories:

 Continuous Time and Discrete Time Signals

 Deterministic and Non-deterministic Signals
 Even and Odd Signals
 Periodic and Aperiodic Signals
 Energy and Power Signals
 Real and Imaginary Signals
Continuous Time and Discrete Time Signals

A signal is said to be continuous when it is defined for all instants of time.

A signal is said to be discrete when it is defined at only discrete instants of time/

Deterministic and Non-deterministic Signals

A signal is said to be deterministic if there is no uncertainty with respect to its value at any instant of time. Or,
signals which can be defined exactly by a mathematical formula are known as deterministic signals.

A signal is said to be non-deterministic if there is uncertainty with respect to its value at some instant of time. Non-
deterministic signals are random in nature hence they are called random signals. Random signals cannot be described
by a mathematical equation. They are modelled in probabilistic terms.
Even and Odd Signals

A signal is said to be even when it satisfies the condition x(t) = x(-t)

Example 1: t2, t4… cost etc.

Let x(t) = t2

x(-t) = (-t)2 = t2 = x(t)

∴ t2 is even function
Example 2: As shown in the following diagram, rectangle function x(t) = x(-t) so it is also even function.

A signal is said to be odd when it satisfies the condition x(t) = -x(-t)

Example: t, t3 ... And sin t Let x(t) = sin t

x(-t) = sin(-t) = -sin t = -x(t)

∴ sin t is odd function.

Any function ƒ(t) can be expressed as the sum of its even function ƒe(t) and odd function ƒo(t). ƒ(t ) = ƒe(t ) + ƒ0(t )
where

ƒe(t ) = ½[ƒ(t ) +ƒ(-t )]

Periodic and Aperiodic Signals

A signal is said to be periodic if it satisfies the condition x(t) = x(t + T) or x(n) = x(n + N). Where
T = fundamental time period, 1/T = f = fundamental frequency.

The above signal will repeat for every time interval T0 hence it is periodic with period T0.

Energy and Power Signals

A signal is said to be energy signal when it has finite energy.

A signal is said to be power signal when it has finite power.

NOTE:A signal cannot be both, energy and power simultaneously. Also, a signal may be neither energy nor power
signal.

Power of energy signal = 0 Energy of power signal = ∞

Real and Imaginary Signals

A signal is said to be real when it satisfies the condition x(t) = x*(t) A signal is said to be odd when it satisfies the
condition x(t) = -x*(t) Example:
If x(t)= 3 then x*(t)=3*=3 here x(t) is a real signal.
If x(t)= 3j then x*(t)=3j* = -3j = -x(t) hence x(t) is a odd signal.
Note: For a real signal, imaginary part should be zero. Similarly for an imaginary signal, real part should be zero.
Basic operations on Signals:

There are two variable parameters in general:

1. Amplitude
2. Time

(1) The following operation can be performed with amplitude:

Amplitude Scaling

C x(t) is a amplitude scaled version of x(t) whose amplitude is scaled by a factor C.

Addition

Addition of two signals is nothing but addition of their corresponding amplitudes. This can be best explained by
using the following example:

As seen from the previous diagram,

-10 < t < -3 amplitude of z(t) = x1(t) + x2(t) = 0 + 2 = 2

-3 < t < 3 amplitude of z(t) = x1(t) + x2(t) = 1 + 2 = 3 3 < t < 10 amplitude of z(t) = x1(t) + x2(t) = 0 + 2 = 2

Subtraction

subtraction of two signals is nothing but subtraction of their corresponding amplitudes.

This can be best explained by the following example:
As seen from the diagram above,

-10 < t < -3 amplitude of z (t) = x1(t) - x2(t) = 0 - 2 = -2

-3 < t < 3 amplitude of z (t) = x1(t) - x2(t) = 1 - 2 = -1 3 < t < 10 amplitude of z (t) = x1(t) - x2(t) = 0 - 2 = -2
Multiplication

Multiplication of two signals is nothing but multiplication of their corresponding amplitudes. This can be best
explained by the following example:

As seen from the diagram above,

-10 < t < -3 amplitude of z (t) = x1(t) ×x2(t) = 0 ×2 = 0

-3 < t < 3 amplitude of z (t) = x1(t) - x2(t) = 1 ×2 = 2 3 < t < 10 amplitude of z (t) = x1(t) - x2(t) = 0 × 2 = 0

(2) The following operations can be performed with time:

Time Shifting

x(t ±t0) is time shifted version of the signal x(t). x (t + t0) →negative shift
x (t - t0) →positive shift

Time Scaling

x(At) is time scaled version of the signal x(t). where A is always positive.
|A| > 1 → Compression of the signal
|A| < 1 → Expansion of the signal

Note: u(at) = u(t) time scaling is not applicable for unit step function.

Time Reversal

x(-t) is the time reversal of the signal x(t).

Classification of Systems:

Systems are classified into the following categories:

 Liner and Non-liner Systems

 Time Variant and Time Invariant Systems
 Liner Time variant and Liner Time invariant systems
 Static and Dynamic Systems
 Causal and Non-causal Systems
 Invertible and Non-Invertible Systems
 Stable and Unstable Systems
Linear and Non-linear Systems

A system is said to be linear when it satisfies superposition and homogenate principles. Consider two systems with
inputs as x1(t), x2(t), and outputs as y1(t), y2(t) respectively. Then, according to the superposition and homogenate
principles,

T [a1 x1(t) + a2 x2(t)] = a1 T[x1(t)] + a2 T[x2(t)]

∴ T [a1 x1(t) + a2 x2(t)] = a1 y1(t) + a2 y2(t)

From the above expression, is clear that response of overall system is equal to response of individual system.

Example:

y(t) = x2(t) Solution:

y1 (t) = T[x1(t)] = x12(t)

y2 (t) = T[x2(t)] = x22(t)

T [a1 x1(t) + a2 x2(t)] = [ a1 x1(t) + a2 x2(t)]2

Which is not equal to a1 y1(t) + a2 y2(t). Hence the system is said to be non linear.

Time Variant and Time Invariant Systems

A system is said to be time variant if its input and output characteristics vary with time.
Otherwise, the system is considered as time invariant. The condition for time invariant system is:
y (n , t) = y(n-t)
The condition for time variant system is:

y (n , t) ≠ y(n-t
Where y (n , t) = T[x(n-t)] = input change

y (n-t) = output change

Example:

y(n) = x(-n)

y(n, t) = T[x(n-t)] = x(-n-t)

y(n-t) = x(-(n-t)) = x(-n + t)

∴ y(n, t) ≠ y(n-t). Hence, the system is time variant.

Liner Time variant (LTV) and Liner Time Invariant (LTI) Systems

If a system is both liner and time variant, then it is called liner time variant (LTV) system.

If a system is both liner and time Invariant then that system is called liner time invariant (LTI) system.

Static and Dynamic Systems

Static system is memory-less whereas dynamic system is a memory system.

Example 1: y(t) = 2 x(t)

For present value t=0, the system output is y(0) = 2x(0). Here, the output is only dependent upon present input.
Hence the system is memory less or static.

Example 2: y(t) = 2 x(t) + 3 x(t-3)

For present value t=0, the system output is y(0) = 2x(0) + 3x(-3).

Here x(-3) is past value for the present input for which the system requires memory to get this output. Hence, the
system is a dynamic system.

Causal and Non-Causal Systems

A system is said to be causal if its output depends upon present and past inputs, and does not depend upon future
input.

For non causal system, the output depends upon future inputs also.

Example 1: y(n) = 2 x(t) + 3 x(t-3)

For present value t=1, the system output is y(1) = 2x(1) + 3x(-2).

Here, the system output only depends upon present and past inputs. Hence, the system is causal.
Example 2: y(n) = 2 x(t) + 3 x(t-3) + 6x(t + 3)

For present value t=1, the system output is y(1) = 2x(1) + 3x(-2) + 6x(4) Here, the system output depends upon
future input. Hence the system is non-causal system.

Invertible and Non-Invertible systems

A system is said to invertible if the input of the system appears at the output.

Y(S) = X(S) H1(S) H2(S)

= X(S) H1(S) · 1(H1(S))

Since H2(S) = 1/( H1(S) )

∴ Y(S) = X(S)

→ y(t) = x(t)
Hence, the system is invertible.

If y(t) ≠ x(t), then the system is said to be non-invertible.

Stable and Unstable Systems

The system is said to be stable only when the output is bounded for bounded input. For a bounded input, if the
output is unbounded in the system then it is said to be unstable.

Note: For a bounded signal, amplitude is finite.

Example 1: y (t) = x2(t)

Let the input is u(t) (unit step bounded input) then the output y(t) = u2(t) = u(t) = bounded output.

Hence, the system is stable.

Example 2: y (t) = ∫x(t)dt

Let the input is u (t) (unit step bounded input) then the output y(t) = ∫u(t)dt = ramp signal (unbounded because
amplitude of ramp is not finite it goes to infinite when t → infinite).
Hence, the system is unstable.

1.1 Continuous-time and discrete-time Signals

1.1.1 Examples and Mathematical representation

Signals are represented mathematically as functions of one or more independent variables. Here
we focus attention on signals involving a single independent variable. For convenience, this will
generally refer to the independent variable as time.
There are two types of signals: continuous-time signals and discrete-time signals.

Continuous-time signal: the variable of time is continuous. A speech signal as a function of

time is a continuous-time signal.

Discrete -time signal: the variable of time is discrete. The weekly Dow Jones stock market index
is an example of discrete-time signal.
x(t) x[n]

x[0]

x[-1] x[1]

x[-2] x[2]

-5 -4 -3 3 4 5
n
t -2 -1 0 1 2

Fig. 1.1 Graphical representation of continuous- Fig. 1.2 Graphical representation of discrete-time
time signal. signal.

To distinguish between continuous-time and discrete-time signals we use symbol t to denote the
continuous variable and n to denote the discrete-time variable. And for continuous-time signals
we will enclose the independent variable in parentheses (), for discrete-time signals we will
enclose the independent variable in bracket [].

A discrete-time signal x[n] may represent a phenomenon for which the independent variable is
inherently discrete. A discrete-time signal x[n] may represent successive samples of an
underlying phenomenon for which the independent variable is continuous. For example, the
processing of speech on a digital computer requires the use of a discrete time sequence
representing the values of the continuous-time speech signal at discrete points of time.
1.1.2 Signal Energy and Power

If v(t) and i(t) are respectively the voltage and current across a resistor with resistance R , then
the instantaneous power is

1
p(t)  v(t)i(t)  v 2 (t) . (1.1)
R

The total energy expended over the time interval t1  t  t2 is

1
 p(t)dt  
t2 t2
v 2 (t)dt , (1.2)
t1 t1 R

and the average power over this time interval is

1 1 1
 
t2 t2
p(t)dt  v 2 (t)dt . (1.3)
t2  t1 t1 t2  t1 t1 R

For any continuous-time signal x(t) or any discrete-time signal x[n], the total energy over the
time interval t1  t  t2 in a continuous-time signal x(t) is defined as

t2 2


t1
x(t) dt , (1.4)

where x denotes the magnitude of the (possibly complex) number x . The time-averaged power
1 t2 2
is

t2  t1 t1
x(t) dt . Similarly the total energy in a discrete-time signal x[n] over the time

interval n1  n  n2 is defined as

n2

x[n]
2
(1.5)
n1

n2
1 2

n2  n1  1 
The average power is x[n]
n1

In many systems, we will be interested in examining the power and energy in signals over an
infinite time interval, that is, for    t   or    n   . The total energy in continuous
time is then defined

T
 2
 2
E   lim  x(t) dt  x(t) dt , (1.6)
T  T  
and in discrete time

N  
E x[n]  x[n] .
2
 lim
2
 (1.7)
N 
N 

For some signals, the integral in Eq. (1.6) or sum in Eq. (1.7) might not converge, that is, if x(t)
or x[n] equals a nonzero constant value for all time. Such signals have infinite energy, while
signals with E   have finite energy.

The time-averaged power over an infinite interval

P  lim 1 T x(t) dt

2
(1.8)
T  2T T

 N 2
P  lim 1
N  2N  1
 x[n] (1.9)
N

Three classes of signals: 


 Class 1: signals with finite total energy, E    and zero average power, (Energy Signal)

E
P  lim 0 (1.10)

T  2T

 Class 2: with finite average power P. If P  0 , then E   . An example is the signal
x[n]  4 , it has infinite energy, but has an average power of P =16. (Power Signal)

Class 3: signals for which neither P and E are finite. An example of this signal is x(t)  t .

1.2 Transformations of the independent variable

In many situations, it is important to consider signals related by a modification of the

independent variable. These modifications will usually lead to reflection, scaling, and shift.

1.2.1 Examples of Transformations of the Independent Variable

 x[n] x[n-n 0]

n n
n0

(a) (b)

Fig.1.3 Discrete-time signals related by a time shift.

x(t-t0)
x(t)

t
t0
t

Fig. 1.4 Continuous-time signals related by a time shift.

x[n] x[-n]

n n

(a) (b)

Fig. 1.5 (a) A discrete-time signal x[n]; (b) its reflection, x[n] about n  0 .

x(t) x(-t)

t t
0 0

(a) (b)

Fig. 1.6 (a) A continuous-time signal x(t) ; (b) its reflection, x(t) about t  0 .
 x(t)
x(2t)

t
t 0
(a) (b)
x(t/2)

t
0
(c)

Fig. 1.7 Continuous-time signals related by time scaling.

1.2.2 Periodic Signals

A periodic continuous-time signal x(t) has the property that there is a positive value of T for
which

x(t)  x(t  T ) for all t (1.11)

From Eq. (1.11), we can deduce that if x(t) is periodic with period T, then x(t)  x(t  mT ) for
all t and for all integers m . Thus, x(t) is also periodic with period 2T, 3T, …. The fundamental
period T0 of x(t) is the smallest positive value of T for which Eq. (1.11) holds.
x(t)

...... ......

Fig. 1.8 Continuous-time periodic signal.

A discrete-time signal x[n] is periodic with period N , where N is an integer, if it is unchanged
by a time shift of N,

x[n]  x[n  N] (1.12)

for all values of n. If Eq. (1.12) holds, then x[n] is also periodic with period 2N , 3N , …. The
fundamental period N0 is the smallest positive value of N for which Eq. (1.12) holds.

x[n]

...... ......
n

Fig. 1.9 Discrete-time periodic signal.

1.2.3 Even and Odd Signals

In addition to their use in representing physical phenomena such as the time shift in a radar
signal and the reversal of an audio tape, transformations of the independent variable are
extremely useful in examining some of the important properties that signal may possess.

Signal with these properties can be even or odd signal, periodic signal:

An important fact is that any signal can be decomposed into a sum of two signals, one of which
is even and one of which is odd.
x(t)
x(t)

t
0

t
0

(a) (b)

Fig. 1.10 An even continuous-time signal; (b) an odd continuous-time signal.

 EVx(t)
1
x(t)  x(t) (1.13)
2

which is referred to as the even part of x(t) . Similarly, the odd part of x(t) is given by

ODx(t) 
1
x(t)  x(t) (1.14)
2

Exactly analogous definitions hold in the discrete-time case.

1, n  0 1
x[n] x[n]   x[n] n 0
0, n  0  ,
2

EV x[ n]  1, n 0
1
1 1  , n 0
2
1
2
n n

(a) (b)
 1
x[n]  , n0
 2

ODx[n]  0, n  0
1
 , n  0
2
1
2

n
1

2

(c)

Fig.1.11 The even-odd decomposition of a discrete-time signal.

1.3 Exponential and sinusoidal signals

1.3.1 Continuous-time complex exponential and sinusoidal signals

The continuous-time complex exponential signal

x(t)  Ceat (1. 15)

where C and a are in general complex numbers.

Real exponential signals

x(t) x(t)

C
C
t t

(a) (b)
Fig. 1.12 The continuous-time complex exponential signal x(t)  Ce , (a) a  0 ; (b) a  0 .
at

Periodic complex exponential and sinusoidal signals

If a is purely imaginary, we have

x(t)  e j0 t (1.16)

An important property of this signal is that it is periodic. We know x(t) is periodic with period
T if

e j0t  e j0 ( tT )  e j0 t e j0T (1.17)

For periodicity, we must have

e j0T  1 (1.18)

For  0  0 , the fundamental period T0 is

2 
T0   (1.19)
 0

Thus, the signals e j0 t and e j0 t have the same fundamental period.

A signal closely related to the periodic complex exponential is the sinusoidal signal

x(t)  A cos(0t  ) (1.20)

With seconds as the unit of t, the units of  and 0 are radians and radians per second. It is also
known  0  2f 0 , where f0 has the unit of circles per second or Hz.
The sinusoidal signal is also a periodic signal with a fundamental period of T0 .

x( t)  A cos( 0t   )

2
T 0
0

A cos

Fig. 1.13 Continuous-time sinusoidal signal.

Using Euler’s relation, a complex exponential can be expressed in terms of sinusoidal signals
with the same fundamental period:

e j0 t  cos 0 t  j sin 0 t (1.21)

Similarly, a sinusoidal signal can also be expressed in terms of periodic complex exponentials
with the same fundamental period:

Acos( t  )  Ae j e j0 t  A e j e  j0 t (1.22)

0
2 2

A sinusoid can also be expresses as 



Acos( t0  )  ARe e j (0t  )  (1.23)

and 


Asin( 0t   )  A Im e j(0 t )  (1.24)

Periodic signals, such as the sinusoidal signals provide important examples of signal with infinite
total energy, but finite average power. For example:
T0 T0
e j0 t dt  1dt  T0
Eperiod  
0 0
(1.25)

1 T0 T0
e j0t dt 
Pperiod 
T0 
0  1dt  1
0
(1.26)
Since there are an infinite number of periods as t ranges from   to   , the total energy
integrated over all time is infinite. The average power is finite since
2
1 T
P  lim e j t dt  1 (1.27)

T  2T 
T
0

Harmonically related complex exponentials:

 k(t)  e jk t , k  0,  1,  2, ......

0
(1.28)

0 is the fundamental frequency.

Example:

Signal x(t)  e j 2t  e j 3t can be expressed as x(t)  e j 2.5t (e  j 0.5t  e j0.5t )  2e j 2.5t cos(0.5t) , the
magnitude of x(t) is x(t)  2 cos(0.5t) , which is commonly referred to as a full-wave rectified
sinusoid, shown in Fig. 1.14.

x (t)

t
 4  2 0 2 4

Fig. 1.14 Full-wave rectified sinusoid.

General complex Exponential signals

Consider a complex exponential Ce at , where C  C e j is expressed in polar and a  r  j0 is

expressed in rectangular form. Then

Ce at  C e j e(r j0 )t  C ert e j (0t )  C ert cos( 0t   )  j C e rt sin( 0t   ) . (1.29)

Thus, for r  0 , the real and imaginary parts of a complex exponential are sinusoidal.
For r  0 , sinusoidal signals multiplied by a growing exponential.
For r  0 , sinusoidal signals multiplied by a decaying exponential.

Damped signal – Sinusoidal signals multiplied by decaying exponentials are commonly refereed
to as damped signal.
 x(t)
x(t)

t t

(a) (b)

Fig. 1.15 (a) Growing sinusoidal signal; (b) decaying sinusoidal signal.

1.3.2 Discrete-time complex exponential and sinusoidal signals

A discrete complex exponential or sequence is defined by

x[n]  C n , (1.30)

where C and  are in general complex numbers. This can be alternatively expressed

x[n]  Ce n , (1.31)

where   e  .

Real Exponential Signals

If C and  are real, we have the real exponential signals.

x[n] x[n]

n n

(a) (b)
x[n] x[n]

n n

(c) (d)
 Fig. 1.16 Real Exponential Signal x[n]  C n : (a)  >1; (b) 0< <1; (c) –1< <0; (d)  <-1.

Sinusoidal Signals

x[n]  e j0 n (1.32)

e j0 n  cos n0  j sin n0 (1.33)

Similarly, a sinusoidal signal can also be expresses in terms of periodic complex exponentials
with the same fundamental period:

Acos( n  )  A e j e j0n  A e  j e j0 n (1.34)

0
2 2

A sinusoid can also be expresses as 


Acos( 0n  )  ARe e  j( 0 n)
 (1.35)

and 

Asin( 0n  )  AIm e  j (0 n )
 (1.36)

The above signals are examples of discrete signals with infinite total energy, but finite average
power. For example: every sample of x[n]  e j0 n contributes 1 to the signal’s energy. Thus the
total energy    n   is infinite, while the average power is equal to 1.
 Fig.1.17 Discrete-time sinusoidal signal.

General Complex Exponential Signals

Consider a complex exponential C n , where C  C e j and    e j0 , then

C n  C  cos( 0n   )  j C  sin j( 0n   ) .

n n
(1.37)

Thus, for   1, the real and imaginary parts of a complex exponential are sinusoidal.
For   1, sinusoidal signals multiplied by a decaying exponential.
For   1, sinusoidal signals multiplied by a growing exponential.
 (a) (b)

Fig. 1.18 (a) Growing sinusoidal signal; (b) decaying sinusoidal signal.

1.3.3 Periodicity Properties of Discrete-Time Complex Exponentials

There are a number of important differences between continuous-time and discrete-time

sinusoidal signals. The continuous-time signals e j0 t are distinct for distinct values of 0 . For
discrete-time signals, however, these values are not distinct because the signal with 0 is
identical to the signals with frequencies  0  2 ,  0  4 , and so on,

e j (0 2 ) n  e j(0 4 ) n  e j0 n . (1.38)

In considering discrete-time exponentials, we need only consider a frequency interval of 2 . In

most occasions, we will use the interval 0  0  2 or    0   .

The discrete-time signal x[n]  e j0 n does not have a continuously increasing rate of oscillation
as 0 is increased in magnitude, but as 0 is increased from 0, the signal oscillates more and
more rapidly until 0 reaches  , and when 0 is continuously increased, the rate of oscillation
 decreases until 0 reaches 2 . We conclude that the low-frequency discrete-time exponentials
have values of 0 near 0, 2 , and any other even multiple of  , while the high-frequencies are
located near 0   and other odd multiples of  .

In order for the signal x[n]  e j0 n to be periodic with period N  0 , we must have

e j0 (n  N )  e j0 n , (1.39)

or equivalently

e j0 N  1. (1.40)

For Eq. (1.40) to hold,  0 N must be a multiple of 2 . That is, there must be an integer m such
that

 0 N  2m , (1.41)

or equivalently

0 m
 . (1.42)
2 N

From Eq. (1.40), x[n]  e j0 n is a periodic if  0 / 2 is a rational number and is not periodic
otherwise.

The fundamental frequency of the discrete-time signal x[n]  e j0 n is

2 0
 , (1.43)
N m

and the fundamental period of the signal can be

 2 
N  . (1.44)
m
0 


The comparison of the continuous-time and discrete-time signals are summarized in the table
below:
 Table 1 Comparison of the signals e j0 t and e j0n .

e j 0 t e j 0 n
Distinct signals for distinct values of 0 Identical signals for values of 0 separated
by multiples of 2
Periodic for any choice of 0 Periodic only if  0  2m / N for some
integers N  0 and m .
Fundamental frequency  0 Fundamental frequency  0 / m
Fundamental period Fundamental period
 0  0 : undefined  0  0 : undefined
  0 : 2
0   0 : m 2 
  


0


0
 0 
 




Example : Suppose that we wish to determine the fundamental period of the discrete-time signal

x[n]  e j ( 2 / 3) n  e j( 3 / 4 )n (1.45)

Solution:

The first exponential on the right hand side has a fundamental period of 3. The second
exponential has a fundamental period of 8.

For the entire signal to repeat, each of the terms in Eq. (1.45) must go through an integer number
of its own fundamental period. The smallest increment of n the accomplished this is 24. That is,
over an interval of 24 points, the first term will have gone through 8 of its fundamental periods,
and the second term through three of its fundamental periods, and the overall signal through
exactly one of its fundamental periods.

Harmonically related periodic exponentials

 k[n]  e jk ( 2 / N ) n , k  0,  1, ...... (1.46)

In the continuous-time case, all of the harmonically related complex exponentials e jk ( 2 / N )t ,

k  0,  1,........, are distinct. But this is not the case for discrete-time signals:

 k  N [n]  e j( k N )( 2 / N )n  e j( k 2 / N ) ne j 2n   k[n] (1.47)

There are only N distinct period exponentials in the set given in Eq. (1.46).
1.4 The Unit Impulse and Unit Step Functions

The unit impulse and unit step functions in continuous and discrete time are considerably
important in signal and system analysis.

1.4.1 The discrete-Time Unit Impulse and Unit Step Sequences

Discrete-time unit impulse is defined as

0, n0
 [n]   , (1.48)
1, n0

 [n]

Fig. 1.19 Discrete-time unit impulse.

Discrete-time unit step is defined as

0, n0
u[n]   , (1.49)
1, n0

u [ n]
1

n
0

Fig. 1.20 Discrete-time unit step sequence.

The discrete-time impulse unit is the first difference of the discrete-time step

 [n]  u[n]  u[n  1] , (1.50)

The discrete-time unit step is the running sum of the unit sample:
 

n
u[n]   [m] ,
m
(1.51)


It can be seen that for n  0 , the running sum is zero, and for n  0 , the running sum is 1.


If we change the variable of summation from m to k  n  m we have, u[n]   [n  k] .

k0

The unit impulse sequence can be used to sample the value of a signal at n  0 . Since  [n] is
nonzero only for n  0 , it follows that

x[n] [n]  x[0][n] . (1.52)

More generally, a unit impulse  [n  n0 ] , then

x[n] [n  n0 ]  x[n0 ] [n  n 0 ] (1.53)

This sampling property is very important in signal analysis.

1.4.2 The Continuous-Time Unit Step and Unit Impulse Functions

Continuous-time unit step is defined as

0, t0
u(t)   , (1.54)
1, t0

u (t)
1

t
0

Fig. 1.21 Continuous-time unit step function.

The continuous-time unit step is the running integral of the unit impulse

u(t)    ( )d .

t
(1.55)

The continuous-time unit impulse can also be considered as the first derivative of the continuous-
time unit step,
 du(t)
 (t)  . (1.56)
dt

Since u(t) is discontinuous at t  0 and consequently is formally not differentiable. This can be
interpreted, however, by considering an approximation to the unit step u  (t) , as illustrated in the
figure below, which rises from the value of 0 to the value 1 in a short time interval of length  .

u  (t )
  (t )

1
1


t t
0  0 
(a) (b)

Fig. 1.22 (a) Continuous approximation to the unit step u  (t) ; (b) Derivative of u  (t) .

The derivative is

du (t)
 (t)  , (1.57)

dt

 1
, 0  t  
  (t)     , (1.58)
 0, otherwise

as shown in Fig. 1.22.

Note that   (t) is a short pulse, of duration  and with unit area for any value of  . As 0,
  (t) becomes narrower and higher, maintaining its unit area. At the limit,

 (t)  lim (t) , (1.59)

0

u(t)  limu (t) , (1.60)

0

and
 du(t)
 (t)  . (1.61)
dt

Graphically,  (t) is represented by an arrow pointing to infinity at t  0 , “1” next to the arrow
represents the area of the impulse.

 (t) k  (t )

1 k

t t
0 0

Fig. 1.23 Continuous-time unit impulse.

Sampling property of the continuous-time unit impulse:

x(t) (t)  x(0) (t) , (1.62)

Or more generally,

x(t) (t  t0 )  x(t0 ) (t  t0 ) (1.63)

Example:

Consider the discontinuous signal x(t)

x(t )

2
x (t )

2 t
0
1 -1

t -2
0
-1 -3

Fig. 1.24 The discontinuous signal and its derivative.

 Note that the derivative of a unit step with a discontinuity of size of k gives rise to an impulse of
area k at the point of discontinuity.

1.5 Continuous-Time and Discrete-Time Systems

A system can be viewed as a process in which input signals are transformed by the system or
cause the system to respond in some way, resulting in other signals as outputs.

Examples

+
+
vs (t ) - C v0 (t )
i(t )
-

(a)

f (t )

(a)

Fig. 1. 25 Examples of systems. (a) A system with input voltage vs (t) and output voltage v0 (t) .
(b) A system with input equal to the force f (t) and output equal to the velocity v(t) .

A continuous-time system is a system in which continuous-time input signals are applied and
results in continuous-time output signals.

Continuous-time
x(t ) y (t )
system

A discrete-time system is a system in which discrete-time input signals are applied and results in
discrete-time output signals.

Discrete-time
x[n ] y[n]
system
1.5.2 Simple Examples of Systems

Example 1: Consider the RC circuit in Fig. 25 (a).

The current i(t) is proportional to the voltage drop across the resistor:
v (t)  vC (t)
i(t)  s . (1.64)
R

The current through the capacitor is

dvC (t)
i(t)  C . (1.65)
dt

Equating the right-hand sides of Eqs. 1.64 and 1.65, we obtain a differential equation describing
the relationship between the input and output:

dvC (t) 1 v (t)  1

 C
v (t) ,
s
(1.66)
dt RC RC

Example 2: Consider the system in Fig. 25 (b), where the force f (t) as the input and the velocity
v(t) as the output. If we let m denote the mass of the car and v the resistance due to friction.
Equating the acceleration with the net force divided by mass, we obtain

dv(t)  1  f (t)  v(t) dv(t)  1 f (t) .

   v(t)  (1.67)
dt m dt m m

Eqs.1.66 and 1.77 are two examples of first-order linear differential equations of the form:

dy(t)
 ay(t)  bx(t). (1.66)
dt

Example 3: Consider a simple model for the balance in a bank account from month to month.
Let y[n] denote the balance at the end of nth month, and suppose that y[n] evolves from month
to month according the equation:

y[n]  1.01y[n  1]  x[n], (1.67)

or

y[n]  1.01y[n  1]  x[n] , (1.68)

where x[n] is the net deposit (deposits minus withdraws) during the nth month 1.01y[n  1]
models the fact that we accrue 1% interest each month.
 Example 4: Consider a simple digital simulation of the differential equation in Eq. (1.67), in
which we resolve time into discrete intervals of length  and approximate dv(t) / d (t) at t  n
by the first backward difference, i.e.,

v(n)  v((n  1))




Let v[n]  v(n) and f [n]  f (n) , we obtain the following discrete-time model relating the
sampled signals v[n] and f [n],

m ∆
v[n]  v[n  1]  f [n] . (1.69)
(m  ) (m  )

Comparing Eqs. 1.68 and 1.69, we see that they are two examples of the first-order linear
difference equation, that is,

y[n]  ay[n  1]  bx[n] . (1.70)

Some conclusions:

 Mathematical descriptions of systems have great deal in common;

 A particular class of systems is referred to as linear, time-invariant systems.
 Any model used in describing and analyzing a physical system represents an idealization of
the system.

1.5.3 Interconnects of Systems

Input System1 System1 Output

(a)

System1

Input + Output

System 2

(b)
 System1 System 2

Input + Output

System 3

(c)

Fig. 1.26 Interconnection of systems. (a) A series or cascade interconnection of two systems; (b)
A parallel interconnection of two systems; (c) Combination of both series and parallel systems.

Input + System1 Output

System 2

Fig. 1.27 Feedback interconnection.

Rs
+

Vi A Vo
-
Vs ±

RL
Vf •
R2
R1

(a)

+ vi  vs  v f BASIC
vs v L  A vi
+ AMPLIFIER
A
-

Feedback

Signal FEEDBACK vL
NETWORK
v f  v L
FB

(b)
 Fig. 1.28 A feedback electrical amplifier.

1.6 Basic System Properties

1.6.1 Systems with and without Memory

A system is memoryless if its output for each value of the independent variable as a given time is
dependent only on the input at the same time. For example:

y[n]  (2x[n]  x2 [n])2 , (1.71)

is memoryless.

A resistor is a memoryless system, since the input current and output voltage has the relationship:
i(t )
+
v(t)  Ri(t) , (1.72)
v(t )
where R is the resistance. -

One particularly simple memoryless system is the identity system, whose output is identical to its
input, that is

y(t)  x(t) , or y[n]  x[n]

An example of a discrete-time system with memory is an accumulator or summer.


n n 1
y[n]   x[k ]  x[k ]  x[n]  y[n  1]  x[n] , or (1.73)
k  k 


y[n]  y[n  1]  x[n] . (1.74)

Another example is a delay

y[n]  x[n  1] . (1.75)

A capacitor is an example of a continuous-time system with memory,

i(t )
1 +

t
v(t)  i( )d, (1.76)
C  v(t )
-
where C is the capacitance.

1.6.2 Invertibility and Inverse System

A system is said to be invertible if distinct inputs leads to distinct outputs.

y[n] Inverse
x[n] System w[n]=x[n]
system

y(t)
x(t) y(t)=2x(t) w(t)=0.5y(t) w(t)=x(t)

n
y(t)
x[n] y[n]   x[k ] w[n]  y[n ]  y[ n  1] w[ n]  x[n ]
k  

Fig. 1.29 Concept of an inverse system.

Examples of non-invertible systems:

y[n]  0 ,

the system produces zero output sequence for any input sequence.

y(t)  x 2 (t) ,

in which case, one cannot determine the sign of the input from the knowledge of the output.

Encoder in communication systems is an example of invertible system, that is, the input to the
encoder must be exactly recoverable from the output.

1.6.3 Causality

A system is causal if the output at any time depends only on the values of the input at present
time and in the past. Such a system is often referred to as being nonanticipative, as the system
output does not anticipate future values of the input.

The RC circuit in Fig. 25 (a) is causal, since the capacitor voltage responds only to the present
and past values of the source voltage. The motion of a car is causal, since it does not anticipate
future actions of the driver.
The following expressions describing systems that are not causal:

y[n]  x[n]  x[n  1] , (1.77)

and

y(t)  x(t  1) . (1.78)

All memoryless systems are causal, since the output responds only to the current value of input.

Example : Determine the Causality of the two systems:

(1) y[n]  x[n]

(2) y(t)  x(t) cos(t  1)

Solution: System (1) is not causal, since when n  0 , e.g. n  4 , we see that y[4]  x[4], so
that the output at this time depends on a future value of input.

System (2) is causal. The output at any time equals the input at the same time multiplied by a
number that varies with time.

1.6.4 Stability

A stable system is one in which small inp uts leads to responses that do not diverge. More
formally, if the input to a stable system is bounded, then the output must be also bounded and
therefore cannot diverge.

Examples of stable systems and unstable systems:

+
+
vs (t ) C v0 ( t)
-
i(t) - f (t )

(a) (b)

The above two systems are stable system.



n
The accumulator y[n]   x[k ] is not stable, since the sum grows continuously even if x[n] is
k 
bounded.
Check the stability of the two systems:

 S1; y(t)  tx(t) ;

 S2: y(t)  ex(t )

 S1 is not stable, since a constant input x(t)  1 , yields y(t)  t , which is not bounded – no
matter what finite constant we pick, y(t) will exceed the constant for some t.

 S2 is stable. Assume the input is bounded x(t)  B , or  B  x(t)  B for all t. We then see
that y(t) is bounded e B  y(t)  e B .

1.6.5 Time Invariance

A system is time invariant if a time shift in the input signal results in an identical time shift in
the output signal. Mathematically, if the system output is y(t) when the input is x(t) , a time-
invariant system will have an output of y(t  t0 ) when input is x(t  t0 ) .

Examples:

 The system y(t)  sin[x(t)] is time invariant.

 The system y[n]  nx[n] is not time invariant. This can be demonstrated by using
counterexample. Consider the input signal x1 [n]   [n] , which yields y1[n]  0 . However,
the input x2 [n]   [n 1] yields the output y 2 [n]  n [n 1]   [n  1] . Thus, while x2 [n] is
the shifted version of x1 [n] , y2 [n] is not the shifted version of y1[n] .

 The system y(t)  x(2t ) is not time invariant. To check using counterexample. Consider
x1 (t) shown in Fig. 1.30 (a), the resulting output y1 (t) is depicted in Fig. 1.30 (b). If the
input is shifted by 2, that is, consider x 2 (t)  x1 (t  2) , as shown in Fig. 1.30 (c), we obtain
the resulting output y2 (t)  x2 (2t ) shown in Fig. 1.30 (d). It is clearly seen that
y2 (t)  y1 (t  2) , so the system is not time invariant.
 x1 (t ) y1 (t ) x 2 (t )  x1 (t  2)

1 1 1

-2 2 -1 1 0 4
(a) (b) (c)
y2 (t ) y 2 (t  2 )

1 1

0 2 1 3
(d) (e)

Fig. 1.30 Inputs and outputs of the system y(t)  x(2t) .

1.6.6 Linearity

The system is linear if

 The response to x1 (t)  x2 (t) is y1 (t)  y2 (t) - additivity property

 The response to ax1 (t) is ay1 (t) - scaling or homogeneity property.

The two properties defining a linear system can be combined into a single statement:

 Continuous time: ax1 (t)  bx2 (t)  ay1 (t)  by2 (t) ,
 Discrete time: ax1 [n]  bx2 [n]  ay1 [n]  by2 [n].

Here a and b are any complex constants.

Superposition property: If xk [n], k  1, 2, 3, ... are a set of inputs with corresponding outputs
yk [n], k  1, 2, 3, ... , then the response to a linear combination of these inputs given by

x[n]   ak xk [n]  a1 x1[n]  a2 x2 [n]  a3 x3 [n]  ... , (1.79)

k

is
y[n]   ak yk [n]  a1 y1[n]  a2 y2 [n]  a3 y3 [n]  ... , (1.80)
k

which holds for linear systems in both continuous and discrete time.

For a linear system, zero input leads to zero output.

Examples:

 The system y(t)  tx(t) is a linear system.

 The system y(t)  x 2 (t) is not a liner system.
 The system y[n]  Rex[n], is additive, but does not satisfy the homogeneity, so it is not a
linear system.
 The system y[n]  2x[n]  3 is not linear. y[n]  3 if x[n]  0 , the system violates the “zero-
in/zero-out” property. However, the system can be represented as the sum of the output of a
linear system and another signal equal to the zero-input response of the system. For system
y[n]  2x[n]  3 , the linear system is

x[n]  2x[n] ,

and the zero-input response is

y0 [n]  3

as shown in Fig. 1.31.

y 0 (t )

x(t ) Linear system + y (t)

Fig. 1.31 Structure of an incrementally linear system. y0 (t) is the zero-input response of the
system.

The system represented in Fig. 1.31 is called incrementally linear system. The system responds
linearly to the changes in the input.

The overall system output consists of the superposition of the response of a linear system with a
zero-input respons
After successful completion of the course, students will be able to:
CO No. Course Outcomes Knowledge
Level
(Bloom’s
Taxonomy)
CO 3 Illustrate Fourier series and Fourier transforms for calculating spectral Apply
characteristics of periodic and aperiodic signals.

CO 4 Make use of Fourier transform and its properties for determine the Apply
frequency response of the systems.

MAPPING OF COs WITH POs and PSOs FOR MODULE-II :

Program
Course Program Outcomes Specific
Outcomes Outcomes
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3
CO 3 - √ - - - - - - - - - - - - -
CO 4 - √ √ - - - - - - - - - √ - -
 MODULE – II
FOURIER SERIES
Representation of Fourier series, Continuous time periodic signals, Properties of Fourier Series,
Dirichlet‟s conditions, Trigonometric Fourier Series and Exponential Fourier Series, Complex
Fourier spectrum.
Fourier Transforms: Deriving Fourier Transform from Fourier series, Fourier Transform of
arbitrary signal, Fourier Transform of standard signals, Fourier Transform of Periodic Signals,
Properties of Fourier Transform, Fourier Transforms involving Impulse function and Signum
function, Introduction to Hilbert Transforms.

3.0 Introduction

 Signals can be represented using complex exponentials – continuous-time and discrete-

time Fourier series and transform.
 If the input to an LTI system is expressed as a linear comb ination of periodic complex
exponentials or sinusoids, the output can also be expressed in this form.

3.1 A Historical Perspective

By 1807, Fourier had completed a work that series of harmonically related sinusoids were
useful in representing temperature distribution of a body. He claimed that any periodic signal
could be represented by such series – Fourier Series. He also obtained a representation for
aperidic signals as weighted integrals of sinusoids – Fourier Transform.

Jean Baptiste Joseph Fourier

3.2 The Response of LTI Systems to Complex Exponentials

It is advantageous in the study of LTI systems to represent signals as linear combinations

of basic signals that possess the following two properties:

 The set of basic signals can be used to construct a broad and useful class of signals.
  The response of an LTI system to each signal should be simple enough in structure to provide
us with a convenient representation for the response of the system to any signal constructed
as a linear combination of the basic signal.

Both of these properties are provided by Fourier analysis.

The importance of complex exponentials in the study of LTI system is that the response of an
LTI system to a complex exponential input is the same complex exponential with only a change
in amplitude; that is

Continuous time: est  H (s)e st , (3.1)

Discrete-time: z n  H (z)z n , (3.2)

where the complex amplitude factor H (s) or H (z) will be in general be a function of the
complex variable s or z.

A signal for which the system output is a (possible complex) constant times the input is referred
to as an eigenfunction of the system, and the amplitude factor is referred to as the system’s
eigenvalue. Complex exponentials are eigenfunctions.

For an input x(t) applied to an LTI system with impulse response of h(t) , the output is

d 
 
y(t)   h( )x(t   )d 
s( t )
h()e
 
, (3.3)
 
  h( )e s (t  ) d  e st  h( )e  s d
 

  s
where we assume that the integral
 h()e

d converges and is expressed as

H (s)     )e s d , (3.4)

h(



the response to est is of the form

y(t)  H (s)est , (3.5)

It is shown the complex exponentials are eigenfunctions of LTI systems and H (s) for a
specific value of s is then the eigenvalues associated with the eigenfunctions.

Complex exponential sequences are eigenfunctions of discrete-time LTI systems. That is,
suppose that an LTI system with impulse response h[n] has as its input sequence
 x[n]  z n , (3.6)

where z is a complex number. Then the output of the system can be determined from the
convolution sum as
  

y[n]  h[k ]x[n  k ]  h[k ]z n k
z n
h[k ]z k
. (3.7)
k  k  k 


Assuming that the summation on the right-hand side of Eq. (3.7) converges, the output is the
same complex exponential multiplied by a consta nt that depends on the value of z . That is,

y[n]  H (z)z n , (3.8)



where H (z)  h[k ]z k . (3.9)

k 


It is shown the complex exponentials are eigenfunctions of LTI systems and H (z) for a
specific value of z is then the eigenvalues associated with the eigenfunctions z n .

The example here shows the usefulness of decomposing general signals in terms of
eigenfunctions for LTI system analysis:

Let x(t)  a e s1t  a e s2 t  a e s3 t , (3.10)

1 2 3

from the eigenfunction property, the response to each separately is

a e s1t  a H (s )e s1t
1 11 1

a2e s2t  a2 H 2 (s 2 )e s2t

a3e s3t  a3 H 3 (s 3 )e s3t

and from the superposition property the response to the sum is the sum of the responses,

y(t)  a1H1(s1)es1t  a2 H2 (s2 )es2t  a3H3(s3 )es3t , (3.11)

Generally, if the input is a linear combination of complex exponentials,

x(t)   a k e skt , (3.12)

k

the output will be

 y(t)   a H (s )e skt , (3.13)
k k
k

Similarly for discrete-time LTI systems, if the input is

x[n]   ak zk ,
n
(3.14)
k

the output is

y[n]   ak H (zk ) zk ,
n
(3.15)
k

3.3 Fourier Series representation of Continuous-Time Periodic Signals

3.31 Linear Combinations of harmonically Related Complex Exponentials

A periodic signal with period of T ,

x(t)  x(t  T ) for all t , (3.16)

We introduced two basic periodic signals in Chapter 1, the sinusoidal signal

x(t)  cos 0 t , (3.17)

and the periodic complex exponential

x(t)  e j0t , (3.18)

Both these signals are periodic with fundamental frequency 0 and fundamental period
T  2 / 0 . Associated with the signal in Eq. (3.18) is the set of harmonically related complex
exponentials

 k(t)  e jk t  e jk ( 2 / T )t ,
0
k  0,  1,  2, ...... (3.19)

Each of these signals is periodic with period of T (although for k  2 , the fundamental period of
 k (t) is a fraction of T ). Thus, a linear combination of harmonically related complex
exponentials of the form
  

a a
jk ( 2 /T ) t
x(t)  k e
jk0 t
 k e , (3.20)
k  k 


is also periodic with period of T .

 k  0 , x(t) is a constant.
 k  1 and k  1 , both have fundamental frequency equal to 0 and are collectively
referred to as the fundamental components or the first harmonic components.
 k  2 and k  2 , the components are referred to as the second harmonic components.
 k   N and k   N , the components are referred to as the Nth harmonic components.

Eq. (3.20) can also be expressed as


x(t)  x *(t)  a *
k 
k e  jk0t , (3.21)

where we assume that x(t) is real, that is, x(t)  x *(t) .

Replacing k by  k in the summation, we have


x(t)  a *
k 
k
e jk0t , (3.22)


which , by comparison with Eq. (3.20), requires that ak  a *k , or equivalently

a *k  ak . (3.23)

To derive the alternative forms of the Fourier series, we rewrite the summation in Eq. (2.20) as

0 ae k  ak e . (3.24)

k 1

Substituting a *k for ak , we have

0 ae k  a *k e . (3.25)
k 1

Since the two terms inside the summation are complex conjugate of each other, this can be
expressed as

 

x(t)  a0 
 2 Re ak e jk0t . (3.26)
k 1
If ak is expressed in polar from as

a k A ek jk ,

then Eq. (3.26) becomes

 

x(t)  a0 
 2Re Ak e j( k0t k ) .
k 1

That is



x(t)  a0  2 Ak cos(k0 t   k ) . (3.27)

k1

It is one commonly encountered form for the Fourier series of real periodic signals in continuous
time.

Another form is obtained by writing ak in rectangular form as

ak  Bk  jCk

then Eq. (3.26) becomes


x(t)  a 0  2  Bk cos k 0 t  C k sin k 0 t . (3.28)
k 1

For real periodic functions, the Fourier series in terms of complex exponential has the following
three equivalent forms:

 

x(t)  a ek jk t   a e jk(

0
k
2 / T ) t
k  k 


x(t)  a0  2 Ak cos(k 0t   k )
k 1


x(t)  a 0  2  B k cos k 0 t  C k sin k 0 t 

k 1
 3.3.2 Determination of the Fourier Series Representation of a Continuous-Time Periodic
Signal


Multiply both side of x(t)  a e by e jn0 t , we obtain
jk 0t
k
k 


x(t)e jn0 t  a jk0 t  jn0 t
k e e , (3.29)
k 

Integrating both sides from 0 to T  2 / 0 , we have
T
 jn t   jn t   j ( kn ) t 
 T jk t  T
 x(t)e 0
dt   a k   e 0 e 0
dt    a k   e 0
dt  , (3.30)
0
k 
0  k   0 

Note that

T , kn

T
e j( k n )0 t dt  
0
0, kn

So Eq. (3.30) becomes

1 T
a  x(t)e jn0 t dt , (3.31)
n
T 0

The Fourier series of a periodic continuous-time signal

 
x(t)  a e
k 
k
jk 0t
 a k e jk (2 / T )t
k 
(3.32)

1  jk 0 t 1
ak   dt  
 jk (2 / T )t
x(t)e x(t)e dt (3.33)
T T T T

Eq. (3.32) is referred to as the Synthesis equation, and Eq. (3.33) is referred to as analysis
equation. The set of coefficient ak  are often called the Fourier series coefficients of the
spectral coefficients of x(t) .

The coefficient a0 is the dc or constant component and is given with k  0 , that is

 1
a
0
T  x(t)dt ,
T
(3.34)

Example : consider the signal x(t)  sin 0 t .

1 j0t  1  j t
sin  0 t  2 e  e 0 .
2j

Comparing the right-hand sides of this equation and Eq. (3.32), we have

1 1
a1  , a1  
2j 2j

ak  0 , k  1 or  1

Example : The periodic square wave, sketched in the figure below and define over one period is

1, t  T1
x(t)   , (3.35)
T1  t  T / 2
0,

The signal has a fundamental period T and fundamental frequency  0  2 / T .

x (t )

 2T T  T T T T 2T
 1 1

2 2

To determine the Fourier series coefficients for x(t) , we use Eq. (3.33). Because of the
symmetry of x(t) about t  0 , we choose  T / 2  t  T / 2 as the interval over which the
integration is performed, although any other interval of length T is valid the thus lead to the same
result.

For k  0 ,

1 T1 1 T1 2T1
a x(t)dt  dt 
 
T T
, (3.36)
0
T 1 T 1 T

For k  0 , we obtain
 T
1
a  1
1

 e
T1
 jk0t
T
dt  e  jk0 t
jk 0T
k
T 1
T1

2  e jk0T1 2 ej  jk0T1  


 k T (3.37)
0  

2 sin(k0 T1 ) sin( k 0T1 )
 
k 0 T k


The above figure is a bar graph of the Fourier series coefficients for a fixed T 1 and several
values of T . For this example, the coefficients are real, so they can be depicted with a single
graph. For complex coefficients, two graphs corresponding to the real and imaginary parts or
amplitude and phase of each coefficient, would be required.

3.4 Convergence of the Fourier Series

If a periodic signal x(t) is approximated by a linear combination of finite number of

harmonically related complex exponentials

N
xN (t)  a e
k  N
k
jk0t
. (3.38)
Let eN (t) denote the approximation error,

N
eN (t)  x(t)  xN (t)  x(t)   a k e jk0t . (3.39)
k  N

The criterion used to measure quantitatively the approximation error is the energy in the error
over one period:

2
E N   e N (t) dt . (3.40)
T

It is shown (problem 3.66) that the particular choice for the coefficients that minimize the energy
in the error is

1
a
T 
 jk 0t
k x(t )e
T
dt . (3.41)

It can be seen that Eq. (3.41) is identical to the expression used to determine the Fourier series
coefficients. Thus, if x(t) has a Fourier series representation, the best approximation using only
a finite number of harmonically related complex exponentials is obtained by truncating the
Fourier series to the desired number of terms.

The limit of EN as N   is zero.

One class of periodic signals that are representable through Fourier series is those signals which
have finite energy over a period,

2
x(t ) dt   ,

T
(3.42)

When this condition is satisfied, we can guarantee that the coefficients obtained from Eq. (3.33)
are finite. We define


e(t)  x(t)  a
k
k
e jk0 t , (3.43)

then


2
e(t) dt  0 , (3.44)
T
 The convergence guaranteed when x(t) has finite energy over a period is very useful. In this
case, we may say that x(t) and its Fourier series representation are indistinguishable.

Alternative set of conditions developed by Dirichlet that guarantees the equivalence of the signal
and its Fourier series representation:

Condition 1: Over any period, x(t) must be absolutely integrable, that is

T
x(t) dt   , (3.45)

This guarantees each coefficient ak will be finite, since

1 1
a  x(t)e  jk0 t dt 
k
T 
T
TT
 x(t) dt   . (3.46)

A periodic function that violates the first Dirichlet condition is

1
x(t)  , 0  t  1.
t

Condition 2: In any finite interval of time, x(t) is of bounded variation; that is, there are no
more than a finite number of maxima and minima during a single period of the signal.

An example of a function that meets Condition1 but not Condition 2:

 2 
x(t)  sin , 0  t  1, (3.47)
 
 t 


Condition 3: In any finite interval of time, there are only a finite number of discontinuities.
Furthermore, each of these discontinuities is finite.

An example that violates this condition is a function defined as

x(t)  1 , 0  t  4 , x(t)  1/ 2 , 4  t  6 , x(t)  1/ 4 , 6  t  7 , x(t)  1/ 8 , 7  t  7.5 , etc.

The above three examples are shown in the figure below.

 The above are generally pathological in nature and consequently do not typically arise in practical
contexts.

Summary:

 For a periodic signal that has no discontinuities, the Fourier series representation converges and
equals to the original signal at all the values of t .
 For a periodic signal with a finite number of discontinuities in each period, the Fourier series
representation equals to the original signal at all the values of t except the isolated points of
discontinuity.

Gibbs Phenomenon:

Near a point, where x(t) has a jump discontinuity, the partial sums xN (t) of a Fourier series
exhibit a substantial overshoot near these endpoints, and an increase in N will not diminish the amplitude
of the overshoot, although with increasing N the overshoot occurs over smaller and smaller intervals. This
phenomenon is called Gibbs phenomenon.
A large enough value of N should be chosen so as to guarantee that the total energy in these
ripples is insignificant.

3.5 Properties of the Continuous-Time Fourier Series

Notation: suppose x(t) is a periodic signal with period T and fundamental frequency 0 . Then if
the Fourier series coefficients of x(t) are denoted by ak , we use the notation

x(t) 
FS
 ak ,
 to signify the pairing of a periodic signal with its Fourier series coefficients.

Linearity

Let x(t) and y(t) denote two periodic signals with period T and which have Fourier series
coefficients denoted by ak and bk , that is

x(t) 
FS
 a k and y(t) 
FS
 bk ,

then we have

z(t)  Ax(t)  By(t) 

FS
 c k  Aak  Bbk . (3.48)

3.5.1 Time Shifting

When a time shift to a periodic signal x(t) , the period T of the signal is preserved.

If x(t) 
FS
 a k , then we have

x(t  t 0 ) 
FS
 e  jk 0 t a k . (3.49)

The magnitudes of its Fourier series coefficients remain unchanged.

3.4.3 Time Reversal

If x(t) 
FS
 a k , then

x(t) 
FS
 ak . (3.50)

Time reversal applied to a continuous-time signal results in a time reversal of the corresponding
sequence of Fourier series coefficients.

If x(t) is even, that is x(t)  x(t) , the Fourier series coefficients are also even, ak  ak .
Similarly, if x(t) is odd, that is x(t)  x(t) , the Fourier series coefficients are also odd,
ak  ak .

3.5.4 Time Scaling


If x(t) has the Fourier series representation x(t)  a e k
jk0t
, then the Fourier series


x(t)   a k e jk (0 ) t . (3.51)
k 


The Fourier series coefficients have not changes, the Fourier series representation has changed
because of the change in the fundamental frequency.

3.5.5 Multiplication

Suppose x(t) and y(t) are two periodic signals with period T and that

x(t) 
FS
 ak ,

y(t) 
FS
 bk .

Since the product x(t) y(t) is also periodic with period T, its Fourier series coefficients hk is

x(t) y(t) 
FS
h 
k
a b


l k l
. (3.52)
l 

The sum on the right-hand side of Eq. (3.52) may be interpreted as the discrete-time convolution
of the sequence representing the Fourier coefficients of x(t) and the sequence representing the
Fourier coefficients of y(t) .

3.5.6 Conjugate and Conjugate Symmetry

Taking the complex conjugate of a periodic signal x(t) has the effect of complex conjugation
and time reversal on the corresponding Fourier series coefficients. That is, if

x(t) 
FS
 ak ,

then

x * (t) 
FS
 a * k . (3.53)

If x(t) is real, that is, x(t)  x *(t) , the Fourier series coefficients will be conjugate symmetric,
that is

ak  a *k . (3.54)
From this expression, we may get various symmetry properties for the magnitude, phase, real
parts and imaginary parts of the Fourier series coefficients of real signals. For example:

 From Eq. (3.54), we see that if x(t) is real, a0 is real and ak  ak .
 If x(t) is real and even, we have ak  a k , from Eq. (3.54) ak  a *k , so ak  a *k  the
Fourier series coefficients are real and even.
 If x(t) is real and odd, the Fourier series coefficients are real and odd.

3.5.7 Parseval’s Relation for Continuous-Time periodic Signals

Parseval’s Relation for Continuous-Time periodic Signals is




1 
x(t) dt   ak 2 ,
2

TT
 k 
(3.55)

Since
1 1
 ae jk0t 2 2
dt  a dt  a ,
2

T T
k
T

T
k k

2
so that a k is the average power in the kth harmonic component.

Thus, Parseval’s Relation states that the total average power in a periodic signal equals the sum
of the average powers in all of its harmonic components.
 3.5.8 Summary of Properties of the Continuous-Time Fourier Series

Property Periodic Signal Fourier Series

Coefficients

x(t) Periodic with period T and ak


y(t) fundamenta l frequency   2 /T bk
 0

Linearity Ax(t)  By(t) Aa k  Bbk

Time Shifting x(t  t0 ) e jk0t a k
Frequency shifting e jM0 t x(t) ak M
Conjugation x *(t) a *k
Time Reversal x(t) ak
Time Scaling x(t) ,   0 (Periodic with period T /  ) ak
Periodic Convolution
 x( ) y(t   )d
Tak bk
T

Multiplication x(t) y(t)
 ab l kl
l
Differentiation 2
dx(t) jk a  jk a
0 k
dt T k
t
Integration  1   1 


x(t)dt (finite valued and periodic only 
 jk
 ak  
 jk(2  / T )
ak
if a0  0 ) 0  

Conjugate Symmetry for x(t) real  a k  a *k

Real Signals  Rea   Rea 
 k k

Imak    Im a k 
 a k  ak

 ak  ak
Real and Even Signals x(t) real and even ak real and even
Real and Odd Signals x(t) real and odd ak purely imaginary and
odd
Even-Odd Decomposition
 xe (t)  Evx(t)

x(t) real
of Real Signals
x (t)  Od x(t) x(t) real Rea k 
e
j Ima k 
Parseval’s Relation for Periodic Signals

1 2 2
x(t) dt   ak
T T k 
 Example : Consider the signal g(t) with a fundamental period of 4.

g (t )

1/2

t
2 1 1 2
 1/ 2

The Fourier series representation can be obtained directly using the analysis equation (3.33). We
may also use the relation of g(t) to the symmetric periodic square wave x(t) discussed on page
8. Referring to that example, T  4 and T1  1 ,

g (t)  x(t  1)  1/ 2 . (3.56)

The time-shift property indicates that if the Fourier series coefficients of x(t) are denoted by a k
the Fourier series coefficients of x(t  1) can be expressed as

bk  a ke  jk / 2 . (3.57)

The Fourier coefficients of the dc offset in g(t), that is the term –1/2 on the right-hand side of
Eq. (3.56) are given by

0, for k  0
ck 
 1
  , for k  0 . (3.58)
 2

Applying the linearity property, we conclude that the coefficients for g(t) can be expressed as

a e jk / 2 , for k  0
 k
d k  1 , (3.59)
 , for k  0
a 0
2
sin(k / 2)
replacing a  e jk / 2 , then we have
k
k


sin(k / 2) e jk / 2 , for k  0

d k   k . (3.60)
0, for k  0
 Example : The triangular wave signal x(t) with period T  4 , and fundamental frequency
 0   / 2 is shown in the figure below.

x (t )

t
2 2

The derivative of this function is the signal g(t) in the previous preceding example. Denoting
the Fourier series coefficients of g(t) by dk , and those of x(t) by ek , based on the
differentiation property, we have

d k  jk( / 2)ek . (3.61)

This equation can be expressed in terms of ek except when k  0 . From Eq. (3.60),

2d k 2 sin(k / 2)  jk / 2
ek   e . (3.62)
jk jk 
2

For k  0 , e0 can be simply calculated by calculating the area of the signal under one period and
divide by the length of the period, that is

e0  1 / 2 . (3.63)

Example: The properties of the Fourier series representation of periodic train of impulse,


x(t)   (t  kT ) .
k 
(3.64)


We use Eq. (3.33) and select the integration interval to be  T / 2  t  T / 2 , avoiding the
placement of impulses at the integration limits.
1 T/2 1
a   (t)e jk ( 2 / T )t dt  . (3.65)
k
T T / 2 T
All the Fourier series coefficients of this periodic train of impulse are identical, real and even.
The periodic train of impulse has a straightforward relation to square-wave signals such as g(t)
on page 8. The derivative of g(t) is the signal q(t) shown in the figure below,

x(t )

 2T T T 2T
g (t )

 2T T T  T1 T1 T T 2T

2 2

q(t )

T1
T  T1 T
 2
2

which can also interpreted as the difference of two shifted versions of the impulse train x(t) .
That is,

q(t)  x(t  T1 )  x(t  T1 ) . (3.66)

Based on the time -shifting and linearity properties, we may express the Fourier coefficients bk of
q(t) in terms of the Fourier series coefficient of ak ; that is

b k  e jk 0T1a 
k
1

e jk T a0 1 k e jk T 01

 e jk 0T1 , (3.67)
T

Finally we use the differentiation property to get

bk  jk0 ck , (3.68)

where ck is the Fourier series coefficients of g(t). Thus

 bk
2 j sin( k0T1 ) 2 sin( k0T1 )
ck   , k 0, (3.69)
jk0 jk 0T k0T

c0 can be solve by inspection from the figure:

2T1
c  . (3.70)
0
T

Example: Suppose we are given the following facts about a signal x(t)

1. x(t) is a real signal.

2. x(t) is periodic with period T  4 , and it has Fourier series coefficients ak .
3. a k  0 for k  1 .
4. The signal with Fourier coefficients bk  e  jk / 2 ak is odd.
1 2 1
x(t) dt 
5.
4 4
 2

Show that the information is sufficient to determine the signal x(t) to within a sign factor.

 According to Fact 3, x(t) has at most three nonzero Fourier series coefficients ak : a1 , a0
and a1 . Since the fundamental frequency  0  2 / T  2 / 4   / 2 , it follows that

x(t)  a 0  a1 e jt / 2  a1 e jt / 2 . (3.71)

 Since x(t) is real (Fact 1), based on the symmetry property a0 is real and a1  a *1 .
Consequently,

x(t)  a  a e
0
jt / 2

1
 1

 a e jt / 2 *  a  2 Re a e jt / 2 .
0
 1
 (3.72)

 Based on the Fact 4 and considering the time -reversal property, we note that ak corresponds
to x(t) . Also the multiplication property indicates that multiplication of kth Fourier series
by e jk / 2 corresponds to the signal being shifted by 1 to the right. We conclude that the
coefficients bk correspond to the signal x((t  1))  x(t  1) , which according to Fact 4
must be odd. Since x(t) is real, x(t  1) must also be real. So based the property, the
Fourier series coefficients must be purely imaginary and odd. Thus, b0  0 , b1  b1 .
 Since time reversal and time shift cannot change the average power per period, Fact 5 holds
even if x(t) is replaced by x(t  1) . That is

1 1

2
x(t  1) dt  . (3.73)
4 4 2
 Using Parseval’s relation,

b1  b1  1/ 2 .
2 2
(3.74)

Since b1  b1 , we obtain b1  1/ 2 . Since b1 is known to be purely imaginary, it must be

either b1  j / 2 or b1   j / 2 .

 Finally we translate the conditions on b0 and b1 into the equivalent statement on a0 and
a1 . First, since b0  0 , Fact 4 implies that a0  0 . With k  1 , this condition implies that
a1 e  j / 2 b1   jb1  jb1 . Thus, if we take b1  j / 2 , a1  1/ 2 , from Eq. (3.72),
x(t)  cos(t / 2) . Alternatively, if we take b1   j / 2 , the a1  1 / 2 , and therefore,
x(t)  cos(t / 2) .

3.6 Fourier Series Representation of Discrete-Time Periodic Signals

The Fourier series representation of a discrete-time periodic signal is finite, as opposed to the
infinite series representation required for continuous-time periodic signals

3.6.1 Linear Combination of Harmonically Related Complex Exponentials

A discrete-time signal x[n] is periodic with period N if

x[n]  x[n  N ] . (3.75)

The fundamental period is the smallest positive N for which Eq. (3.75) holds, and the
fundamental frequency is  0  2 / N .

The set of all discrete-time complex exponential signals that are periodic with period N is given
by

 k[n]  e jk n  e jk ( 2 / N ) n ,
0
k  0,  1,  2, ...., (3.76)

All of these signals have fundamental frequencies that are multiples of 2 / N and thus are
harmonically related.

There are only N distinct signals in the set given by Eq. (3.76); this is because the discrete-time
complex exponentials which differ in frequency by a multiple of 2 are identical, that is,

 k [n]  k rN [n] . (3.77)

The representation of periodic sequences in terms of linear combinations of the sequences  k [n]
is

x[n]   a  [n]   a e jk0n  a e jk ( 2 / N ) n . (3.78)

k k k k
k k k

Since the sequences k [n] are distinct over a range of N successive values of k, the summation in
Eq. (3.78) need include terms over this range. We indicate this by expressing the limits of the
summation as k  N . That is,

x[n]   akk [n]   a k e   ak e

jk 0n jk ( 2 / N ) n
. (3.79)
kN k N kN

Eq. (3.79) is referred to as the discrete-time Fourier series and the coefficients ak as the Fourier
series coefficients.

6.2 Determination of the Fourier Series Representation of a Periodic Signal

The discrete-time Fourier series pair:

kN
kk 
k N
k 
k N
k , (3.80)


1  jk (2 / N )n . (3.81)
N n
ak   jk0n 1
x[n]e  x[n]e
N N n N

Eq. (3.80) is called synthesis equation and Eq. (3.81) is called analysis equation.

Example: Consider the signal x[n]  sin 0 n , (3.82)

x[n] is periodic only if 2 / 0 is an integer, or a ratio of integer. For the case the when 2 / 0
is an integer N, that is, when

2
  , (3.83)
0
N

x[n] is periodic with the fundamental period N. Expanding the signal as a sum of two complex
exponentials, we get
 1 j ( 2 / N )n  1  j( 2 / N ) n
x[n]  e  e , (3.84)
2j 2j

From Eq. (3.84), we have

1 1
a1  , a1   , (3.85)
2j 2j

and the remaining coefficients over the interval of summation are zero. As discussed previously,
these coefficients repeat with period N.

The Fourier series coefficients for this example with N  5 are illustrated in the figure below.

When 2 / 0 is a ratio of integer, that is, when

2M
  , (3.86)
0
N
Assuming the M and N do not have any commo n factors, x[n] has a fundamental period of N.
Again expanding x[n] as a sum of two complex exponentials, we have

1 jM ( 2 / N )n  1  jM ( 2 / N ) n
x[n]  e  e , (3.87)
2j 2j

From which we determine by inspection that aM  (1/ 2 j) , aM  (1/ 2 j) , and the remaining
coefficients over one period of length N are zero. The Fourier coefficients for this example with
M  3 and N  5 are depicted in the figure below.
 Example : Consider the signal
 2 
x[n]  1  sin n   2    4 
  3 cos n  cos


n  .


N   N   N 2 


Expanding this signal in terms of complex exponential, we have

3
x[n]  1  (  1 )e j( 2 / N ) n  ( 3  1 )e  j( 2 / N ) n   1 e j / 2  j 2( 2 / N ) n   1 e  j / 2 e  j2 ( 2 / N )n .
 
 


 e  
2 2j 2 2j 2  2 


Thus the Fourier series coefficients for this signal are

a0  1 ,
3 1 3 1
a    j,
1
2 2j 2 2
3 1 3 1
a     j,
1
2 2j 2 2
1
a j,
2
2
1
a2   j .
2

with ak  0 for other values of k in the interval of summation in the synthesis equation. The real
and imaginary parts of these coefficients for N  10 , and the magnitude and phase of the
coefficients are depicted in the figure below.
Example : Consider the square wave shown in the figure below.

Because x[n]  1 for  N1  n  N1 , we choose the length-N interval of summation to include

the range  N1  n  N1 . The coefficients are given

N1  jk ( 2 / N ) n
1
ak 
N e
n N1
, (3.88)

Let m  n  N1 , we observe that Eq. (3.88) becomes

 1 2 N1 jk ( 2 / N )(m N1 )  1 jk ( 2 / N ) N1 2 N1  jk (2 / N ) m
ak  e  e  e , (3.89)
N n0 N n 0

 1  e jk 2 ( 2 N1 1) / N   1 sin2k(N 1  1/ 2) / N  , k  0,  N,  2N, ....

1
a jk ( 2 / N ) N1
(3.90)

 
 1  e  jk ( 2 / N )  N sin(k / N )


k e
N  
and

2N1  1
a , k  0,  N,  2N , .... (3.91)
k
N

The coefficients ak for 2N1  1  5 are sketched for N  10, 20, and 40 in the figure below.

The partial sums for the discrete-time square wave for M  1, 2, 3, and 4 are depicted in the
figure below, where N  9 , 2N1  1  5 .

We see for M  4 , the partial sum exactly equals to x[n]. In contrast to the continuous-time
case, there are no convergence issues and there is no Gibbs phenomenon.
3.7 Properties of Discrete-Time Fourier Series

Property Periodic Signal Fourier Series Coefficients

x[n] ak
 Periodic with period N and Periodic with period N
y[n fundamenta l frequency   2
] 0 bk

Linearity Ax[n]  By[n] Aa k  Bbk

Time Shifting x[n  n0 ] e jk ( 2 / N ) t ak
Frequency shifting e jM ( 2 / N )n x[n] ak M
Conjugation x *[n] a *k
 Time Reversal x[n] ak
Time Scaling x[n/m], if n is a multipleof n 1  viewed as periodic
x(m)[n]   a  
0, if n is a multipleof n m k  with period mN 
(Periodic with period mN )
Periodic Convolution
 x[r]y[n  r]
r [ N ]
Nak bk

Multiplication x[n]y[n]
a b l k l
lN 
Differentiation x[n]  x[n 1] 1  e  jk ( 2 / N )
k a 
Integration n
 1 
 a
 x[k ] (finite valued and periodic
k    1  e jk ( 2 / N ) 
k

only if a0  0 )
Conjugate Symmetry for x[n] real  a k  a *k
Real Signals  Rea   Rea 
 k k

Imak    Im a k 
 a k  ak

a  a
 k k
Real and Even Signals x[n] real and even ak real and even
Real and Odd Signals x[n] real and odd ak purely imaginary and odd
 xe[n]  Evx[n] x[n] real
Even-Odd Decomposition
 Rea k 
x [n]  Od x[n] x[n] real
of Real Signals
e j Ima k 
Parseval’s Relation for Periodic
Signals

 x[n]   a k
1 2 2

T n N  n N 

3.7.1 Multiplication

FS   a l b k l
x[n]y[n]
. (3.92)
l N 

Eq. (3.92) is analogous to the convolution, except that the summation variable is now restricted
to in interval of N consecutive samples. This type of operation is referred to as a Periodic
Convolution between the two periodic sequences of Fourier coefficients.

The usual form of the convolution sum, where the summation variable ranges from   to   ,
is sometimes referred to as Aperiodic Convolution.
3.7.2 First Difference

x[n]  x[n 1]

FS

 1  e  jk ( 2 / N ) a k  . (3.93)

3.7.3 Parseval’s Relation


1 2.
N x[n]  k  N 
T n
2 a k

(3.94)

3.7.4 Examples

Example : Consider the signal shown in the figure below.

x[n]
2

n
-5 0 5

x1 [n]
1

n
-5 0 5

x2 [n]
1

n
-5 0 5
The signal x[n] may be viewed as the sum of the square wave x1 [n] with Fourier series
coefficients bk and x2 [n] with Fourier series coefficients ck .

ak  bk  ck , (3.95)

The Fourier series coefficients for x1 [n] is

1 sin(3k / 5) for k  0,  5,  10, ....

5 ,
bk   sin(  k / 5)
. (3.96)
3 , for k  0,  5,  10, ....
5

The sequence x2 [n] has only a dc value, which is captured by its zeroth Fourier series
coefficient:

14
c0   x [n] 1 ,
5 n0 2
(3.97)

Since the discrete-time Fourier series coefficients are periodic, it follows that ck  1 whenever k
is an integer multiple of 5.

1 sin(3k / 5) for k  0,  5,  10, ....

5 ,
ak   sin(  k / 5)
(3.98)
8 , for k  0,  5,  10, ....
5

Example : Suppose we are given the following facts about a sequence x[n]:

1. x[n] is periodic with period N  6 .


5
2. n0
x[n]  2 .


7
3. n2
(1)n x[n]  1.
4. x[n] has minimum power per period among the set of signals satisfying the preceding three
conditions.

1 5 x[n]  1 .
6 
 From Fact 2, we have a0 
 jn
n 0
 j( 2 / 6)3 n
3 1 1
 Note that (1)  e
n
e , we see from Fact 3 that a 
7
x[n]e  j 3( 2 / N ) n  .
3
6
 2
6
 From Parseval’s relation, the average power in x[n] is
 5

P   ak 2 .
k 0

Since each nonzero coefficient contributes a positive amount to P, and since the values of a0 and
a3 are specified, the value of P is minimized by choosing a1  a2  a4  a5  0 . It follows that
1 1
x[n]  a  a e jn   (1)n ,
0 3
3 6

which is shown in the figure below.

1/2
x[n]

1/6

n
-5 0 5

3.8 Fourier Series and LTI Systems

We have seen that the response of a continuous-time LTI system with impulse response h(t) to a
complex exponential signal est is the same complex exponential multiplied by a complex gain:

y(t)  H (s)est ,
where

s 
H (s) 
 h( )e

d , (3.99)


In particular, for s  j , the output is y(t)  H ( j)e jt . The complex functions H (s) and
H ( j ) a?re called the system function (or transfer function) and the frequency response,
respectively.

By superposition, the output of an LTI system to a periodic signal represented by a Fourier series

  
a a e
jk ( 2 /T ) t
x(t)  k e
jk 0 t
 k is given by
k  k 
  
y(t)  ak H ( jk0 )e jk0t .
k 
(3.99)

That is, the Fourier series coefficients bk of the periodic output y(t) are given by

bk  ak H ( jk 0 ) , (3.100)

Similarly, for discrete-time signals and systems, response h[n] to a complex exponential signal
e jn is the same complex exponential multiplied by a complex gain:

y[n]  H ( jk 0)e jk0 n , (3.101)

where


H (e j )  h[n]e  jn . (3.102)

n 

 3 1
Example : Suppose that the periodic signal x(t)  a k e
jk 2t
with a 0  1 , a1  a1  ,
k 3 4
1
a2  a2  , and a3  a3  1 is the input signal to an LTI system with impulse response
2 3

h(t)  et u(t)

To calculate the Fourier series coefficients of the output y(t) , we first compute the frequency
response:

 
1 1
H ( j )   e  e  j d  e  e  , (3.103)
0 1  j 0
1  j

The output is

3

b e
jk 2t
y(t)  k , (3.104)
k 3

where bk  ak H ( jk 0 )  ak H ( jk 2 ) , so that

1  1  1  1 
b0  0 , b1   , b1   ,
4  1 j2  4 1  j2 
 1  1  1  1 
b2   , b2   ,
4 1  j4  4  1  j4 


1  1  1  1 
b3   , b3   .
4 1  j6  4 1  j6 


Example : Consider an LTI system with impulse response h[n]   nu[n] ,  1    1, and with
the input
 2n 
x[n]  . (3.105)
cos 
 N 


Write the signal x[n] in Fourier series form as

1 j( 2 / N ) n 1  j (2 / N ) n
x[n]  e  e .
2 2

Also the transfer function is


 
e  j  
1
n

  e 
j ) n  jn . (3.106)
H (e  j
n 0 n 0 1e 

The Fourier series for the output

H e j 2 / N e j( 2 / N ) n  H e  j 2 / N e  j( 2 / N )n
1 1
y[n] 
2 2
. (3.107)
 1  1  j( 2 / N )n  1 
e 1   j( 2 / N ) n
e
2 1   e j 2 1   e  j
   


3.9 Filtering

Filtering – to change the relative amplitude of the frequency components in a signal or eliminate
some frequency components entirely.

Filtering can be conveniently accomplished through the use of LTI systems with an appropriately
chosen frequency response.
LTI systems that change the shape of the spectrum of the input signal are referred to as
frequency-shaping filters.

LTI systems that are designed to pass some frequencies essentially undistorted and significantly
attenuate or eliminate others are referred to as frequency-selective filters.

Example : A first-order low-pass filter with impulse response h(t)  et u(t) cuts off the high
frequencies in a periodic input signal, while low frequency harmonics are mostly left intact. The
frequency response of this filter
 1
H ( j )   e  e  j d  . (3.107)
0 1  j


We can see that as the frequency  increase, the magnitude of the frequency response of the
filter H ( j ) decreases. If the periodic input signal is a rectangular wave, then the output signal
will have its Fourier series coefficients bk given by

sin(k0T1 )
bk  ak H ( jk 0 )  , k0 (3.108)
k (1  jk0 )

2T1
b0  a0 H (0)  . (3.109)
T

The reduced power at high frequencies produced an output signal that is smother than the input
signal.

t
T  T1 T1 T
3.10 Examples of continuous-Time Filters Described By Differential
Equations

In many applications, frequency-selective filtering is accomplished through the use of LTI

systems described by linear constant-coefficient differential or difference equations. In fact,
many physical systems that can be interpreted as performing filtering operations are
characterized by differential or difference equation.

3.10.1 A simple RC Lowpass Filter

The first-order RC circuit is one of the electrical circuits used to perform continuous-time
filtering. The circuit can perform either Lowpass or highpass filtering depending on what we
take as the output signal.

v r (t )

+
vs (t ) -
v c (t )

If we take the voltage cross the capacitor as the output, then the output voltage is related to the
input through the linear constant-coefficient differential equation:

dvc (t)
RC  v c (t)  vs (t) . (3.111)
dt

Assuming initial rest, the system described by Eq. (3.111) is LTI. If the input is vs(t)  e jt , we
must have voltage output vc(t)  H ( j )e jt . Substituting these expressions into Eq. (3.111), we
have

RC
d
H ( j )e  H ( j )e
jt jt
 e j t , (3.112)
dt

or

RCjH ( j )e jt  H ( j )e jt  e jt , (3.113)

 1 
Then we have H ( j )  . (3.114)
1  RCj

Te amplitude and frequency response H ( j ) is shown in the figure below.

We can also get the impulse response

1
h(t)  et / RCu(t) , (3.115)
RC

and the step response is

h(t)  (1  e t / RC )u(t) , (3.116)

The fundamental trade-off can be found by

comparing the figures:

 To pass only very low frequencies,

1/ RC should be small, or RC should
be large.

 To have fast step response, we need a

smaller RC .

 The type of trade-off between

behaviors in the frequency domain
and time domain is typical of the
issues arising in the design analysis of
LTI systems.
3.10.2 A Simple RC Highpass Filter

If we choose the output from the resistor, then we get an RC highpass filter.

3.11 Examples of Discrete-Time Filter Described by Difference Equations

A discrete-time LTI system described by the first-order difference equation

y[n]  ay[n  1]  x[n] . (3.116)

Form the eigenfunction property of complex exponential signals, if x[n]  e jn , then
y[n]  H (e j )e jn , where H (e j ) is the frequency response of the system.

j 1 
H (e )  . (3.117)
1  ae  j

The impulse response of the system is

x[n]  a n u[n]. (3.118)

The step response is

1  an1
s[n]  u[n]. (3.119)
1a

From the above plots we can see that for a  0.6 the system acts as a Lowpass filter and
a  0.6 , the system is a highpass filter. In fact, for any positive value of a  1 , the system
approximates a highpass filter, and for any negative value of a  1 , the system approximates a
 highpass filter, where a controls the size of bandpass, with broader pass bands as a in
decreased.

The trade-off between time domain and frequency domain characteristics, as discussed in
continuous time, also exists in the discrete-time systems.

3.11.2.2 Nonrecursive Discrete-Time Filters

The general form of an FIR norecursive difference equation is



M
y[n]  b
k  N
k x[n  k ]. (3.120)

It is a weighted average of the (N  M  1) values of x[n] , with the weights given by the
coefficients bk .

One frequently used example is a moving-average filter, where the output of y[n] is an average
of values of x[n] in the vicinity of n0 - the result corresponding a smooth operation or lowpass
filtering.

1
An example: y[n]  x[n 1]  x[n]  x[n  1]. (3.121)
3

The impulse response is

1
h[n]   [n  1]   [n]   [n  1] , (3.122)
3

and the frequency response

H (e j ) 
1
e j

 1  e  j . (3.123)
3
A generalized moving average filter can be expressed as
M
1
y[n]   b x[n  k] .
N  M  1 k  N k
(3.124)

The frequency response is

j 1 M
1 j  ( N M ) / 2 sin M  N  1/ 2
 j k 
H (e ) e
M  N  1 k  N MN1
e
sin / 2
. (3.125)

The frequency responses with different average window lengths are plotted in the figure below.

FIR norecursive highpass filter

An example of FIR norecursive highpass filter is

 x[n]  x[n  1]
y[n]  . (3.126)
2

The frequency response is

H (e j ) 
1
1  e  je
 j j / 2
sin( / 2) . (3.127)
2
 Continuous-Time Fourier Transform

4.0 Introduction

 A periodic signal can be represented as linear combination of complex exponentials which

are harmonically related.
 An aperiodic signal can be represented as linear combination of complex expone ntials,
which are infinitesimally close in frequency. So the representation take the form of an
integral rather than a sum
 In the Fourier series representation, as the period increases the fundamental frequency
decreases and the harmonically related components become closer in frequency. As the
period becomes infinite, the frequency components form a continuum and the Fourier series
becomes an integral.

4.1 Representation of Aperiodic Signals: The Continuous-Time Fourier

Transform

4.1.1 Development of the Fourier Transform Representation of an Aperiodic Signal

Starting from the Fourier series representation for the continuous-time periodic square wave:

1, t  T1
x(t)   , (4.1)
T1  t  T / 2
0,

x (t )

 2T T T T T T T 2T
 1 1

2 2
The Fourier coefficients ak for this square wave are

2 sin(k0 T1)
ak  . (4.2)
k 0T

or alternatively

74
 2 sin(T1 )
Tak   , (4.3)
  k0

where 2 sin(T1 ) /  represent the envelope of Tak

 When T increases or the fundamental frequency  0  2 / T decreases, the envelope is

sampled with a closer and closer spacing. As T becomes arbitrarily large, the original
periodic square wave approaches a rectangular pulse.

 Tak becomes more and more closely spaced samples of the envelope, as T   , the Fourier
series coefficients approaches the envelope function.

This example illustrates the basic idea behind Fourier’s development of a representation for
aperiodic signals.

Based on this idea, we can derive the Fourier transform for aperiodic signals.

Suppose a signal x(t) with a finite duration, that is, x(t)  0 for t  T1 , as illustrated in the
figure below.

 From this aperiodic signal, we construct a periodic signal ~

x (t) , shown in the figure below.

75
  As T   , ~
x (t)  x(t) , for any infinite value of t .

 The Fourier series representation of ~

x (t) is


x (t)  ak e jk0t ,
~ (4.4)
k 
1 T/2
~
a  x (t)e  jk 0 t dt . (4.5)
k
T T / 2

 Since ~x (t)  x(t) for t  T / 2 , and also, since x(t)  0 outside this interval, so we have
1 T/2 1
a x(t)e  jk0 t dt  x(t)e  jk0t dt .
k
T T / 2 T 



 Define the envelope X ( j ) of Tak as


 j t
X ( j ) 
 x(t)e

dt . (4.6)

we have for the coefficients ak ,

1
a X ( jk )
k 0
T

Then ~
x (t) can be expressed in terms of X ( j ) , that is

~   jk t
x (t)  1 X ( jk0 )e jk0t 1 X ( jk 0 )e 0  .

k  T
 
2 k  0
(4.7)

76
 As T   , ~
x (t)  x(t) and consequently, Eq. (4.7) becomes a representation of x(t) .

 In addition,  0  0 as T   , and the right-hand side of Eq. (4.7) becomes an integral.

We have the following Fourier transform:

1 
2 
x(t)  X ( j  )e j t
d Inverse Fourier Transform (4.8)

and


X ( j  )   x(t)e jt dt Fourier Transform (4.9)

4.1.2 Convergence of Fourier Transform

If the signal x(t) has finite energy, that is, it is square integrable,

 2


x(t) dt   , (4.10)

Then we guaranteed that X ( j ) is finite or Eq. (4.9) converges. If e(t)  ~

x (t)  x(t) , we have

 2

 e(t) dt  0 . (4.11)

An alterative set of conditions that are sufficient to ensure the convergence:

Contition1: Over any period, x(t) must be absolutely integrable, that is



x(t) dt   , (4.12)

Condition 2: In any finite interval of time, x(t) have a finite number of maxima and mi nima.

Condition 3: In any finite interval of time, there are only a finite number of discontinuities.
Furthermore, each of these discontinuities is finite.

77
 4.1.3 Examples of Continuous-Time Fourier Transform

Example : consider signal x(t)  e at u(t) , a  0 .

From Eq. (4.9),


 1  1 a0
X ( j )    j t
e e at
dt   e ( a j )t  , (4.12)
0 a  j 0
a  j


If a is complex rather then real, we get the same result if Rea 0

The Fourier transform can be plotted in terms of the magnitude and phase, as shown in the figure
below.
1  
X ( j)  , X ( j )   tan 1 . (4.13)
 a 

 a 2  2

Example : Let x(t)  ea t , a0

 
 e e  j t dt   e at e  jt dt 
a t 0
X ( j )  e e  jt dt  at 1

1
 2
2a
  0 a  j a  j a   2

The signal and the Fourier transform are sketched in the figure below.

78
 Example : x(t)   (t) . (4.14) x(t)   (t) X ( j)  1

X ( j )   (t)e jt dt  1.


(4.15)



That is, the impulse has a Fourier transform consisting of equal contributions at all frequencies.

Example : Calculate the Fourier transform of the rectangular pulse signal

1, t  T1
x(t)   . (4.16)
0, t  T1
x(t )

 T1 T1
 sin T1
 
T1
X ( j )  x(t)e  j t dt  1e  jt dt  2 .
 T1 
(4.17)

The Inverse Fourier transform is

1  sin T1
x̂(t)   2 e jt d , (4.18)
2  


Since the signal x(t) is square integrable,

 2
e(t)  

x(t)  x̂(t) dt  0 . (4.19)

x̂(t) converges to x(t) everywhere except at the discontinuity, t  T1 , where x̂(t) converges to
½, which is the average value of x(t) on both sides of the discontinuity.

In addition, the convergence of x̂(t) to x(t) also exhibits Gibbs phenomenon. Specifically, the
integral over a finite-length interval of frequencies

1 W sin T1 jt 

2 W
2 e d



79
As W   , this signal converges to x(t) everywhere, except at the discontinuities. More over,
the signal exhibits ripples near the discontinuities. The peak values of these ripples do not
decrease as W increases, although the ripples do become compressed toward the discontinuity,
and the energy in the ripples converges to zero.

Example : Consider the signal whose Fourier transform is

1, W
X ( j)   .
0, W

The Inverse Fourier transform is

x(t)  1 sinWt

W
e jt d  .
2 W t

Comparing the results in the preceding example and this example, we have

FT

Square wave Sinc function


FT 1

This means a square wave in the time domain, its Fourier transform is a sinc function. However,
if the signal in the time domain is a sinc function, then its Fourier transform is a square wave.
This property is referred to as Duality Property.

We also note that when the width of X ( j ) increases, its inverse Fourier transform x(t) will be
compressed. When W   , X ( j ) converges to an impulse. The transform pair with several
different values of W is shown in the figure below.

80
 4.2 The Fourier Transform for Periodic Signals

The Fourier series representation of the signal x(t) is


x(t)  a jk0t
k e . (4.20)
k 


It’s Fourier transform is



X ( j )   2ak  ( k 0 ) .
k 
(4.21)

Example : If the Fourier series coefficients for the square wave below are given

x (t )

 2T T T T T T T 2T
 1 1

2 2

sin k0 T1
a  , (4.22)
k
k

The Fourier transform of this signal is



2 sin k 0T1 
X ( j
)   ( k0 ) . (4.23)
k  k

81
Example : The Fourier transforms for x(t)  sin  0 t and x(t)  cos 0 t are shown in the figure
below.

82
 

Example : Calculate the Fourier transform for signal x(t)   (t  kT ) .

k 


The Fourier series of this signal is

1 T / 2 1
a  (t)e  j0t  .
k
T 
T / 2 T
The Fourier transform is

2 
2k
X ( j)  
T  (  T
k 0
).

The Fourier transform of a periodic impulse train in the time domain with period T is a periodic
impulse train in the frequency domain with period 2 / T , as sketched din the figure below.

4.3 Properties of The Continuous-Time Fourier Transform

4.3.1 Linearity

If x(t) F  X ( j ) and y(t) F Y ( j )

Then

83
ax(t)  by(t) 
F
 aX ( j  )  bY ( j ) . (4. 20)

4.3.2 Time Shifting

If x(t) F  X ( j )

Then

x(t  t0 ) 
F
e  j t X ( j ) . 0
(4. 20)

Or

Fx(t  t0 ) e j t X ( j )  X ( j ) e j X ( j ) t .

0 0
(4. 20)

Thus, the effect of a time shift on a signal is to introduce into its transform a phase shift, namely,
  0t .

Example : To evaluate the Fourier transform of the signal x(t) shown in the figure below.

x(t )

1.5
1
t
1 2 3 4

x2 (t ) x1 (t )

1 1

t t
3 3 1 1
 
2 2 2 2

The signal x(t) can be expressed as the linear combination

1
x(t)  x (t  2.5)  x (t  2.5) . (4. 20)
1 2
2

x1 (t) and x2 (t) are rectangular pulse signals and their Fourier transforms are

84
 2 sin( / 2)  2sin(3 / 2)
X 1( j )  and X 2( j ) 
 

Using the linearity and time -shifting properties of the Fourier transform yields

sin( / 2)  2 sin(3 / 2) 
 X ( j )  e  j5 / 2  
  


4.3.3 Conjugation and Conjugate Symmetry

If x(t) F  X ( j )

Then

x *(t) 
F
 X * ( j ) . (4. 20)

Since X * ( j )   x(t)e dt  
 

 j t
x * (t)e j t dt ,
   


Replacing  by   , we see that

  jt
X *( j ) 
 x * (t)e

dt , (4. 20)


The right-hand side is the Fourier transform of x * (t) .

If x(t) is real, from Eq. (4.20) we can get

X ( j )  X * ( j ) . (4. 20)

We can also prove that if x(t) is both real and even, then X ( j ) will also be real and even.
Similarly, if x(t) is both real and odd, then X ( j ) will also be purely imaginary and odd.

A real function x(t) can be expressed in terms of the sum of an even function
xe (t)  Evx(t)and an odd function xo (t)  Od x(t). That is

x(t)  xe (t)  xo (t)

Form the Linearity property,

85
 F x(t)  F xe (t) F xo (t),

From the preceding discussion, F x e (t) is real function and F x o (t) is purely imaginary. Thus
we conclude with x(t) real,

x(t) F  X ( j )

Evx(t)
F
 ReX ( j  )

Od x(t)F  j ImX ( j )



Example : Using the symmetry properties of the Fourier transform and the result
1
e at u(t) 
F
 to evaluate the Fourier transform of the signal x(t)  ea t , where a  0 .
a  j

Since 
 a t  at  at  
  2Eve u(t) 
 e atu(t)  eat u(t)   at 
x(t) e e u(t) e u( t) 2 

 ,
 2 
 1  2a
So X ( j )  2 Re
 a  j   a   2
2

 


4.3.4 Differentiation and Integration

If x(t) F  X ( j )

Then

dx(t)

F
 jX ( j )
. (4. 20)
dt

1
 x( )d  X ( j )  X (0) ()
t F

 j . (4. 20)

Example : Consider the Fourier transform of the unit step x(t)  u(t) .

It is know that

86
 g(t)   (t) 
F
1

Also note that

t
x(t)   g( )d


The Fourier transform of this function is

1 1
X ( j )   G(0) ( )    ( ) .
j j


where G(0)  1.

Example : Consider the Fourier transform of the function x(t) shown in the figure below.

x(t) 1
1
1 1 1
1 t
t t
1 1
= 1
+ 1 1

dx(t)
g (t) 
dt

From the above figure we can see that g(t) is the sum of a rectangular pulse and two impulses.
 2 sin 
G( j )   e j  e  j 

 
  


Note that G(0)  0 , using the integration property, we obtain

G( j ) 2 sin  2 cos
X ( j)   G(0) ( )   .
j j 2
j


It can be found X ( j ) is purely imaginary and odd, which is consistent with the fact that x(t) is
real and odd.

4.3.5 Time and Frequency Scaling

87
 x(t) F  X ( j ) ,

Then
1 j
x(at) 
F
 X( ). (4. 20)
a a

From the equation we see that the signal is compressed in the time domain, the spectrum will be
extended in the frequency domain. Conversely, if the signal is extended, the corresponding
spectrum will be compressed.

If a  1, we get from the above equation,

x(t) 
F
 X ( j ) . (4. 20)

That is, reversing a signal in time also reverses its Fourier transform.

4.3.6 Duality

The duality of the Fourier transform can be demonstrated using the following example.
t  T1
x (t)  1, F X 2 sinT1
1  ( j) 

1
0, t  T1


x (t) 
sinWT1
F  X 1, W
( j )  
2
t 2
0, W

88
 The symmetry exhibited by these two examples extends to Fourier transform in general. For any
transform pair, there is a dual pair with the time and frequency variables interchanged.

2
Example : Consider using duality and the result e t 
F
 X ( j )  to find the Fourier
1  2
transform G ( j ) of the signal

2
g (t)  .
1t2

2
Since e t 
F
 X ( j )  , that is,
1  2
1   2  jt
e t  2 
 1   2 e d ,
 


Multiplying this equation by 2 and replacing t by  t , we have

2e
t  2   jt 

  
2 
 1   

e d

Interchanging the names of the variables t and  , we find that

 2   jt
e d  F 1 1  t   2e  .
 2
2e

    
 1  2 t
 2 
   


Based on the duality property we can get some other properties of Fourier transform:

dX ( j )
 jtx(t ) 
F

d

e j 0t x(t) 
F
X ( j(  0))

1 
 x(t)  x(0) (t )  x()d
F

jt 

89
4.3.7 Parseval’s Relation

If x(t) F  X ( j ) ,

We have

 1 
   d
2

2
x(t) dt X ( j )
2 

Parseval’s relation states that the total energy may be determined either by computing the energy
2
per unit time x(t) and integrating over all time or by computing the energy per unit frequency
X ( j ) / 2 and integrating over all frequencies. For this reason, X ( j)
2 2
is often referred to
as the energy-density spectrum.

4.4 The convolution properties

y(t)  h(t)  x(t )

F
Y ( j )  H ( j ) X ( j )
The equation shows that the Fourier transform maps the convolution of two signals into product
of their Fourier transforms.

H ( j ) , the transform of the impulse response, is the frequency response of the LTI system,
which also completely characterizes an LTI system.

Example : The frequency response of a differentiator.

dx(t)
y(t)  .
dt

From the differentiation property,

Y ( j )  jX ( j ) ,

The frequency response of the differentiator is

Y( j 
H ( j )  )  j .
X ( j) 

Example : Consider an integrator specified by the equation:

90
 t
y(t)   x( )d .

The impulse response of an integrator is the unit step, and therefore the frequency response of
the system:

1
H ( j )    () .
j

So we have
1
Y ( j )  H ( j ) X ( j )  X ( j )  X (0) () ,
j

which is consistent with the integration property.

Example : Consider the response of an LTI system with impulse response

h(t)  e at u(t) , a0

to the input signal

x(t)  ebt u(t) , b 0

To calculate the Fourier transforms of the two functions:

1 
X ( j)  , and
b  j
1
H ( j )  .
a  j

Therefore,

Y ( j )  1 ,
a  j b  j 

using partial fraction expansion (assuming a  b ), we have

1  1 1
Y ( j )   
 a  j 
b  a  b  j 


The inverse transform for each of the two terms can be written directly. Using the linearity
property, we have

91
 y(t) 
1
b a
 
e at u(t)  e bt u(t) .

We should note that when a  b , the above partial fraction expansion is not valid. However,
with a  b , we have

1
Y ( j )  ,
 a  j 2

Considering a  1j 2  j dd a 1 

  j , and
 

1
e at u(t) 
F
 , and
a  j

d  1 
te at u(t) F  j d  a  j ,
 


so we have

Y (t)  te at u(t) .

4.5 The Multiplication Property

1 
2 
 R( j ) 
r(t)  s(t ) p(t)  S ( j )P( j(   ))d

Multiplication of one signal by another can be thought of as one signal to scale or modulate the
amplitude of the other, and consequently, the multiplication of two signals is often referred to as
amplitude modulation.

Example : Let s(t) be a signal whose spectrum S ( j ) is depicted in the figure below.

92
Also consider the signal

p(t)  cos0 t ,

then

P( j )   (  0 )   (   0 ) .

The spectrum of r(t)  s(t) p(t) is obtained by using the multiplication property,

1 
R( j )   S ( j  )P( j(   ))d
2 
,
1
 S ( j   )  1 S ( j   )
0
2 2
0

which is sketched in the figure below.

From the figure we can see that the signal is preserved although the information has been shifted
to higher frequencies. This forms the basic for sinusoidal amplitude modulation systems for
communications.

Example : If we perform the following multiplication using the signal r(t) obtained in the
preceding example and p(t)  cos0 t , that is,

g (t)  r(t) p(t)

The spectrum of P( j) , R( j) and G ( j ) are plotted in the figure below.

93
If we use a lowpass filter with frequency response H ( j ) that is constant at low frequencies and
zero at high frequencies, then the output will be a scaled replica of S ( j ) . Then the output will
be scaled version of s(t) - the modulated signal is recovered.

94
4.6 Summary of Fourier Transform Properties and Basic Fourier Transform Pairs

95
96
 4.7 System Characterized by Linear Constant-Coefficient Differential
Equations

An LTI system described by the following differential equation:

N
d k y(t) M
d k x(t)
 ak dt k   k 0
bk
dt k
, (4. 67)
k 0

which is commonly referred to as an Nth-order differential equation.

The frequency response of this LTI system

Y ( j )
H ( j )  , (4. 68)
X ( j )

where X ( j ) , Y ( j) and H ( j ) are the Fourier transforms of the input x(t) , output y(t) and
the impulse response h(t) , respectively.

Applying Fourier transform to both sides, we have

N d k y(t)  M d k x(t) 
F  ak dt k   F  bk dt k  , (4. 69)
k 0  k 0 


From the linearity property, the expression can be written as

N
 d k y(t)  M  d k x(t) 
 ak F  dt k    bk F  dt k  . (4. 70)
k 0   k0  


From the differentiation property,

N M
Y ( j ) ( j)k

M
b
a k
( j ) k Y ( j )  b k ( j)k X ( j )  H ( j )   k 0
X ( j)  N ak ( j ) k
k (4. 71)
k 0 k0
k 0

H ( j ) is a rational function, that is, it is a ratio of polynomials in ( j ) .

Example : Consider a stable LTI system characterized by the differential equation

dy(t)
 ay(t)  x(t) , with a  0 .
dt

The frequency response is

97
 1
H ( j )  .
j  a

Te impulse response of this system is then recognized as

h(t)  e at u(t) .

Example : Consider a stable LTI system that is characterized by the differential equation

d 2 y(t) dy(t) dx(t)

2
4  3y(t)   2x(t) .
dt dt dt

The frequency response of this system is

 ( j)  2  j  2
H ( j )   .
( j )  4( j )  3 j  1 j  3
2

Then, using the method of partial-fraction expansion, we find that

1/ 2  1 / 2
H ( j )   .
j  1 j  3

The inverse Fourier transform of each term can be recognized as

1 1
h(t)  et u(t)  e3t u(t) .
2 2

98
After successful completion of the course, students will be able to:
CO No. Course Outcomes Knowledge
Level
(Bloom’s Taxonomy)
CO 5 Identify the linearity and time invariance properties for obtaining the
Apply
behavior of linear time invariant system.
CO 6 Classify the ideal low pass, high pass, band pass and band stop filters
for determining the signal and system bandwidth. Understand

MAPPING OF COs WITH POs and PSOs FOR MODULE-III:

Program
Course Program Outcomes Specific
Outcomes Outcomes
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3
CO 5 √ √ - - - - - - - - - - - - -
CO 6 √ - - - - - - - - - - - - - -
 MODULE – III
SIGNAL TRANSMISSION THROUGH LINEAR SYSTEMS
Linear System, Impulse response, Response of a Linear System, Linear Time Invariant(LTI) System,
Linear Time Variant (LTV) System, Transfer function of a LTI System, Filter characteristic of Linear
System, Distortion less transmission through a system, Signal bandwidth, System Bandwidth, Ideal
LPF, HPF, and BPF characteristics.
Causality and Paley-Wiener criterion for physical realization, Relationship between Bandwidth and
rise time, Convolution and Correlation of Signals, Concept of convolution in Time domain and
Frequency domain, Graphical representation of Convolution.

Linear systems
A system is said to be a linear if it obeys homogeneity and additivity properties. This implies that the
response of a linear system to weighted sum of input signals is equal to the same weighted sum of responses
of the system to each of those signals.
Homogeneity property: This property says if input signal weighted by any arbitrary constant then output
signal also weighted by same arbitrary constant

Additive property: Response of system to sum of two input signals is equal to sum of individual response of
the system.

Combining above two properties

Arbitrary constant
Response

Where is the output of the system in response to

Classification of linear systems
Lumped and Distributed system
Time – Invariant and Time Variant system
Lumped and Distributed system: A Lumped System consists of lumped elements which are interconnected in
particular way. The energy in the system is considered to be stored or dissipated in distinct isolated elements.
The disturbance initiated at any point is propagated instantaneously at every point in the system. The dimension
of elements is very small compared to wave length of the signals to be transmitted. Lumped system obeys Ohms
law and Kirchhoff laws. They can be expressed with ordinary differential equations. Examples are TVS,
motors, computers, any packed sytems
Distributed systems are those in which elements are distributed over a long distances and dimensions of the
circuits are small compared to the wave length of signals to be transmitted. More over such system takes finite
amount of time for disturbance at one point to be propagated to the other point. They can be expressed with
partial differential equations. Example are wave guides, optical fiber, transmission lines, antennas.
Linear Time Invariant (LTI) System: A system said to be LTI if it satisfies linear and invariance properties. Stated in
another way, A LTI system whose parameters do not change with time. LTI system is characterized by linear equations
such as algebraic, differential, or difference equations with constant coefficients.
Example: Circuits using passive elements are LTI systems
For LTI system, if input is delayed by t0 seconds the system satisfies superposition and homogeneity principles. Also, the
output delayed by the same time t0 seconds.

Linear Time Variant (LTV) System: A system said to be LTV if it satisfies the linear property but not the time
invariant. For LTV system, if input delayed by t0 seconds, the system satisfies superposition and homogeneity properties
but output varies with time t0. A LTV system whose parameters change with time. The coefficients in the differential
equations are time variant.

Impulse response and response of LTI system

Let us consider

Signal approximation
Impulse response of LTI system due to an impulse input applied at t=0 is h (t)
Hence
This is known as convolution integral and it gives relationship among input signal, output signal and impulse response of
system.LTI system completely characterized by impulse response

Frequency response of LTI system:

Let us consider LTI system with impulse response h(t) and y(t) is response of input signal x(t) . Input and output
relationship of system given by convolution integral.
.

Fourier transform of input x(t) , output y(t) and impulse response h(t) are X(ω) , Y(ω) and H(ω) respectively.
Magnitude response is symmetric and phase response is anti symmetric.
Response to Eigen functions
If input to the system is an exponential function then output y(t)

Output is a complex exponential of the same frequency as input multiplied by the complex constant . An inputs
signal is called Eigen functions of the system if the corresponding output is a constant multiple of the input signal. Thus
the functions all Eigen functions as we get the same function the output as in input.
Properties of LTI system
Commutative Property

Associate property
This implies that a cascading of two or more LTI system will results to single system with impulse response equal to the
convolution of the impulse response of the cascading systems.

Distributive Property
This property gives that addition of two or more LTI system subjected to same input will results single system with
impulse response equal to the sum of impulse response of two or more individual systems.
Static and dynamic system
A system is static or memory less if its output at any time depends only on the value of its input at that instant of time. For
LTI systems, this property can hold if its impulse response is itself an impulse. But convolution property, we know that
the output depends on the previous samples of the input, therefore an LTI system has memory and hence it is dynamic
system.
Causality
A continuous time LTI system is said to causal if and only if it impulse response is h(t) = 0 for t<0, then integral
becomes

Stability: a continuous time system is bounded input , bounded output stable if and only if the impulse response is
absolutely Integrable.
Consider LTI system with impulse response h(t) . the output y(t) is

If x(t) is bounded and then

For bounded output y(t) < ꝏ , the impulse response should be absolutely integrable. Hence

Above equation gives necessary and sufficient condition for BIBO stability.
Inevitability:
A system T said to be invertible if and only if there exits an inverse system T-1 for such that T T-1 is an identical system .
For an LTI system with impulse response h1(t), this is equivalent to the existence of another system with impulse response
h2(t) such that h1(t)* h2(t) = δ(t).

Transfer Function of LTI System:

Transfer function of LTI system defined as the ratio of Fourier transform of the output signal to Fourier transform
of the input signal .It is expressed as

Inverse Fourier transforms of gives the impulse response of the system. That is h(t) = IFT of

In general Input and output relationship of continuous time causal LTI system described by linear constant coefficient
differential equations with zero initial conditions is given by

Where are constant coefficients the order N refer to the highest derivative of y(t) in above equation.
Apply Fourier Transform on both sides of above equation

Distortion less Transmission System:

Distortion less transmission through the LTI system requires that the response be exact replica of input signal. The replica
may have different magnetude and delayed in time.
Therefore,
Apply the Fourier transform
 Where n is integer number
Therefore, to achieve distortion less transmission through LTI system, magnetude response of system must be
constant over entire frequency range and phase response of the system must be linear with frequency.
Band width of signals and System
Band width of signals: it is the range of significant frequency components present in the signal. A signal may have
frequency components in the entire frequency range from -ꝏ to ꝏ. For any practical signals, the energy content decreases
with frequency, only some of frequency components of signals have significant amplitude within a certain frequency
band; outside this band have negligible amplitude. The amplitude of significant frequency component is within the
times (3dB) of maximum signal amplitude.
System Band width:
The band width of system is defined as the interval of frequencies over which the magnitude spectrum of remains
within times (3dB) its value at the mid band. The band width of system is

times (3dB) its value

at the midband
Times (3dB) of
its value at the midband.
Band width =
For distortion less transmission, a system should have infinite bandwidth. But due to physical limitations it is impossible
to design an ideal filters having infinite bandwidth.
For satisfactory distortion less transmission, therefore, an LTI system should have high bandwidth compared to the signal
bandwidth.

The filter characteristics of linear system:

The system processes the input signal in a way that is characteristics of the system. The system modifies the spectral
density function of input signal according to transfer function. It is observed that the system act as some kind of filter to
various frequency components. Some frequency components are boosted in strength, some are attenuated, and some may
remain unaffected. Similarly, each frequency component suffers a different amount of phase shift in the process of
transmission. LTI system acts as filter depending on the transfer function of system. The transfer function acts as
weighting function to different frequency components of input signal.
LTI system may be classified into five types of filters
Low pass filter
High pass filter
Band pass filter
Band reject filter
All pass filter.
The pass band of a filter the range of frequencies that allowed by the system without distortion. The stop band of filter is
the range of frequencies that attenuated by the system.
Ideal filters:
An Ideal filter passes all frequency components in its pass band without distortion and completely blocks frequency
components outside of pass band. There is discontinuity between pass band and stop band in frequency spectrum. But
practical filters, there is gradual transition gap between pass band and stop band, The range of frequencies over which
there is a gradual attenuation between pass band and stop band is called transition band. Filters with small gap are very
difficult to design.
Ideal Low Pass Filter:
An ideal low pass filter transmits all frequency components below the certain frequency rad/sec called cutoff
frequency, without distortion. The signal above these frequencies is filtered completely.
The transfer function of Ideal Low pass filter given by

Ideal High Pass Filter:

An ideal high pass filter transmits all frequency components above the certain frequency rad/sec called cutoff
frequency, without distortion. The signal below these frequencies is filtered completely.
The transfer function of Ideal high pass filter given by

Ideal Band Pass Filter:

An ideal band pass filter transmits all frequency components within certain frequency band to rad/sec, without
distortion. The signal with frequency outside this band is stopped completely.
The transfer function of Idel band pass filter given by

Ideal Band Reject Filter:

An ideal band reject filter rejects all frequency components within certain frequency band to rad/sec. The signal
outside this band is transmitted without distortion.
The transfer function of Idle band reject filter given by

Causality and Physical Realizability: Paley – Wiener Criterion

For physically realizable systems, that cannot have response before the input signal applied. In time domain approach the
impulse response of physically realizable systems must be causal that is h(t) =0 for t< 0, this is condition known as causal
condition. In frequency domain, this criterion implies that a necessary and sufficient condition for magnetude response
to be physically realizable is

This condition known as the Paley – Wiener criterion. To satisfy this condition the function must be square
integrable that is

All causal systems that satisfy the Paley – Wiener criterion are physically realizable.
Magnetude function may be zero at some discrete frequencies but it cannot be zero over finite band of
frequencies since this will cause the integral to become infinite. Therefore Idle filters are not physically realizable. It can
be concluding that magnetude function cannot fall off to zero faster than exponential order.
is permissible
this Gaussian error curve is not permissible.
But it possible to construct physically realizable filters close to the ideal filter characteristics. Low pass filter having
transfer function

Where an arbitrary small value, produces nearly ideal characteristics shown in fig below
Band Width and Rise Time:
The system band width can be obtained from rise time , which can be derived from output response of the system.
Rise time : the rise time tr of the output response is defined as the time the response takes to reach from 10% to 90% of
the maximum value of the signal or in general it is the time of response to reach from zero to the final value of the signal.

Relationship between Band width and rise time

Consider ideal LPF , its transfer function is given by

Where cut off frequency or 3 dB band width of filter

Apply Inverse Fourier transform
if input is impulse then output is = h(t)

Product of rise time and bandwidth is constant

Rise time inversely proportional to the system band width.
Concept of convolution in time domain:
The process of expressing the output signal in terms of the superposition of weighted and time shifted impulse response is
called convolution. Convolution is a particularly powerful way of characterizing the input – output relationship of LTI
systems. The mathematical tool for evaluating the convolution of continuous time signals is called convolution integral;
for discrete time signals, it is called convolution sum . the convolution integral plays an important role in system analysis
in time and frequency domains. It is important process for signal processing and detection in communication systems.
The convolution integral
Let continuous time signals. Then convolution can be expressed
as

Thus the output of any continuous LTI system is the convolution of the input x(t) with impulse response h(t) of the
system.
Case I : if input signal is causal that is x(t) = 0 for t<0

Case II
System is causal that is h(t) =0 for t<0 then

Case III
Both input signal and system are causal then

Properties of convolution integral

Commutative property
Let continuous time signals
 =
Distributive Property

Associate property

Shift property
If the signal shifted by sec then convolution of
and
If shifted by and respectively

Convolution of function with impulse

=
Convolution of function with unit step
Any arbitrary function x(t) with unit step function u(t)

Proof

Width property
Let us consider finite duration of two signals are T1 and T2 respectively then duration of y(t) =
is equal to the sum of duration of .
T = T1 + T2
Also its area under finite signals are A1 and A2 respectively then the area under y (t) is product of
both areas
A = area under y (t) = area under and area under = A1 A2

Convolution property of Fourier Transform

Fourier transforms pair of two signals given by
 COs COURSE OUTCOMES Blooms level
taxonomy

Convolution in frequency domain:

Fourier Transform of = 2Π Fourier transform of [ x(t) h(t) ]
Fourier transform of [ x(t) h(t) ] =

==

Thus convolution in one domain is transformed a product operation in the other domain

Graphical representation of Convolution

When two signals are provided in graphical form, the convolution can be performed by graphical method. It involves the
following steps.
1. For given signals , draw the signals as function of independent variable.
2. Draw the function of which is time reversal of .then shift function by time t to form .
3. Draw the both signals on the axis with large time shift t along the negative axis.
4. Increase the time t along positive axis . Multiply the signals and integrate over the period
of two signals to obtain convolution at t.
5. Increase the time shift step by step and obtain convolution using step 4.
6. Draw the convolution x (t) with the values obtained in steps 4 and 5 as function of t.
 CO 7 Illustrate the Laplace and Z-transform for analysing the
continuous and discrete time signals and systems. Apply

CO 8 Apply the Region of Convergence Properties of Laplace and z

transform to represent the causal and noncausal Signals. Apply

MAPPING OF COs WITH POs and PSOs FOR MODULE-IV:

Program
Course Program Outcomes Specific
Outcomes Outcomes
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3
CO 7 - √ - - - - - - - - - - - - -
CO 8 - √ √ - - - - - - - - - √ - -
 Lecture Notes Signals & Systems

MODULE – IV
LAPLACE TRANSFORM AND Z TRANSFORM
Laplace Transforms: Laplace Transforms (L.T), Inverse Laplace Transform, Concept of Region of
Convergence (ROC) for Laplace Transforms, Properties of L.T, Relation between L.T and F.T of a signal,
Laplace Transform of certain signals using waveform synthesis.
Z–Transforms: Concept of Z- Transform of a Discrete Sequence, Distinction between Laplace, Fourier and Z
Transforms, Region of Convergence in Z-Transform, Constraints on ROC for various classes of signals, Inverse
Z-transform, Properties of Z-transforms

THE LAPLACE TRANSFORM:

we know that for a continuous-time LTI system with impulse response h(t), the output y(t)of the
system to the complex exponential input of the form est is

A. Definition:

The function H(s) is referred to as the Laplace transform of h(t). For a general continuous-time signal
x(t), the Laplace transform X(s) is defined as

The variable s is generally complex-valued and is expressed as

Relation between Laplace and Fourier transforms:

Laplace transform of x(t)

 Lecture Notes Signals & Systems

Lecture Notes Signals & Systems

Inverse Laplace Transform:

We know that

Conditions for Existence of Laplace Transform:

Dirichlet's conditions are used to define the existence of Laplace transform. i.e.
 The function f has finite number of maxima and minima.
 There must be finite number of discontinuities in the signal f ,in the given interval of time.
 Lecture Notes Signals & Systems
 It must be absolutely integrable in the given interval of time. i.e.

Initial and Final Value Theorems

If the Laplace transform of an unknown function x(t) is known, then it is possible to determine the initial
and the final values of that unknown signal i.e. x(t) at t=0+ and t=∞.

Initial Value Theorem

Statement: If x(t) and its 1st derivative is Laplace transformable, then the initial value of x(t) is given by

Final Value Theorem

Statement: If x(t) and its 1st derivative is Laplace transformable, then the final value of x(t) is given by

Properties of Laplace transform:

The properties of Laplace transform are Linearity Property

 Lecture Notes Signals & Systems

Frequency Shifting Property

Time Reversal Property

Time Scaling Property

Differentiation and Integration Properties

 Lecture Notes Signals & Systems

Multiplication and Convolution Properties

Region of convergence.

The range variation of ζ for which the Laplace transform converges is called region of
convergence.

Properties of ROC of Laplace Transform

 ROC contains strip lines parallel to jω axis in s-plane.
Lecture Notes Signals & Systems

 If x(t) is absolutely integral and it is of finite duration, then ROC is entire s-plane.

 If x(t) is a right sided sequence then ROC : Re{s} > ζo.

 If x(t) is a left sided sequence then ROC : Re{s} < ζo.

 If x(t) is a two sided sequence then ROC is the combination of two regions.

ROC can be explained by making use of examples given below:

Example 1: Find the Laplace transform and ROC of x(t)=e− at u(t) x(t)=e−atu(t)

Example 2: Find the Laplace transform and ROC of x(t)=e at u(−t) x(t)=eatu(−t)
Lecture Notes Signals & Systems

Example 3: Find the Laplace transform and ROC of x(t)=e −at u(t)+e at u(−t)
x(t)=e−atu(t)+eatu(−t)

Referring to the above diagram, combination region lies from –a to a. Hence,

ROC: −a<Res<a
 Lecture Notes Signals & Systems

Causality and Stability

 For a system to be causal, all poles of its transfer function must be right half of s-plane.

 A system is said to be stable when all poles of its transfer function lay on the left half of
s-plane.

 A system is said to be unstable when at least one pole of its transfer function is shifted to
the right half of s-plane.
 Lecture Notes Signals & Systems

 A system is said to be marginally stable when at least one pole of its transfer function
lies on the jω axis of s-plane

LAPLACE TRANSFORMS OF SOME COMMON SIGNALS

A. Unit Impulse Function δ( t ):

B. Unit Step Function u(t ):

CREC Dept. of ECE Page 90

Lecture Notes Signals & Systems

CREC Dept. of ECE Page 91

 Lecture Notes Signals & Systems

Analysis of continuous time LTI systems can be done using z-transforms. It is a powerful
mathematical tool to convert differential equations into algebraic equations.

The bilateral (two sided) z-transform of a discrete time signal x(n) is given as

The unilateral (one sided) z-transform of a discrete time signal x(n) is given as

Z-transform may exist for some signals for which Discrete Time Fourier Transform (DTFT) does
not exist.

Concept of Z-Transform and Inverse Z-Transform

Z-transform of a discrete time signal x(n) can be represented with X(Z), and it is defined as

The above equation represents the relation between Fourier transform and Z-transform

CREC Dept. of ECE Page 92

 Lecture Notes Signals & Systems

Inverse Z-transform:

CREC Dept. of ECE Page 94

 Lecture Notes Signals & Systems

Z-Transform Properties:

Z-Transform has following properties:

Linearity Property:

Time Shifting Property:

Multiplication by Exponential Sequence Property

Time Reversal Property

CREC Dept. of ECE Page 95

 Lecture Notes Signals & Systems

Differentiation in Z-Domain OR Multiplication by n Property

Convolution Property

Correlation Property

Initial Value and Final Value Theorems

Initial value and final value theorems of z-transform are defined for causal signal.

Initial Value Theorem

For a causal signal x(n), the initial value theorem states that

This is used to find the initial value of the signal without taking inverse z-transform

CREC Dept. of ECE Page 96

 Lecture Notes Signals & Systems

Final Value Theorem

For a causal signal x(n), the final value theorem states that

This is used to find the final value of the signal without taking inverse z-transform

Region of Convergence (ROC) of Z-Transform

The range of variation of z for which z-transform converges is called region of convergence of z-
transform.

Properties of ROC of Z-Transforms

 ROC of z-transform is indicated with circle in z-plane.

 ROC does not contain any poles.

 If x(n) is a finite duration causal sequence or right sided sequence, then the ROC is entire z-plane
except at z = 0.

 If x(n) is a finite duration anti-causal sequence or left sided sequence, then the ROC is entire
z-plane except at z = ∞.

 If x(n) is a infinite duration causal sequence, ROC is exterior of the circle with radius a.
i.e. |z| > a.

 If x(n) is a infinite duration anti-causal sequence, ROC is interior of the circle with radius
a. i.e. |z| < a.

 If x(n) is a finite duration two sided sequence, then the ROC is entire z-plane except at z
= 0 & z = ∞.

The concept of ROC can be explained by the following example:

Example 1: Find z-transform and ROC of a n u[n]+a − nu[−n−1] anu[n]+a−nu[−n−1]

The plot of ROC has two conditions as a > 1 and a < 1, as we do not know a.

CREC Dept. of ECE Page 97

 Lecture Notes Signals & Systems

In this case, there is no combination ROC.

Here, the combination of ROC is from a<|z|<1/a

Hence for this problem, z-transform is possible when a < 1.

Causality and Stability

Causality condition for discrete time LTI systems is as follows:

A discrete time LTI system is causal when

 ROC is outside the outermost pole.

 In The transfer function H[Z], the order of numerator cannot be grater than the order of
denominator.

CREC Dept. of ECE Page 98

 Lecture Notes Signals & Systems

Stability Condition for Discrete Time LTI Systems

A discrete time LTI system is stable when

 its system function H[Z] include unit circle |z|=1.

 all poles of the transfer function lay inside the unit circle |z|=1.
Z-Transform of Basic Signals

CREC Dept. of ECE Page 99

 Lecture Notes Signals & Systems

Some Properties of the Z- Transform:

Inverse Z transform:
Three different methods are:
1. Partial fraction method
2. Power series method
3. Long division method

CREC Dept. of ECE Page 100

 Lecture Notes Signals & Systems
Example: A finite sequence x [ n ] is defined as

Find X(z) and its ROC.

Sol: We know that

For z not equal to zero or infinity, each term in X(z) will be finite and consequently X(z) will
converge. Note that X ( z ) includes both positive powers of z and negative powers of z. Thus,
from the result we conclude that the ROC of X ( z ) is 0 < lzl < m.
Example: Consider the sequence

Find X ( z ) and plot the poles and zeros of X(z).

Sol:

From the above equation we see that there is a pole of ( N - 1)th order at z = 0 and a pole at z = a .
Since x[n] is a finite sequence and is zero for n < 0, the ROC is IzI > 0. The N roots of the
numerator polynomial are at

The root at k = 0 cancels the pole at z = a. The remaining zeros of X ( z ) are at

The pole-zero plot is shown in the following figure with N=8

CREC Dept. of ECE Page 101

Lecture Notes Signals & Systems

CREC Dept. of ECE Page 102

Lecture Notes Signals & Systems

CO 9 Identify the similarities between two signals using convolution

and correlation. Apply

CO 10 Make use of cross correlation function for measuring energy

spectral density of a given aperiodic signal. Apply

CO 11 Utilize the power spectral density to measure of power in each

frequency component. Apply

CO 12 Demonstrate the procedure consists of sampling and

reconstruction of bandlimited signals by using various sampling Understand
techniques.

MAPPING OF COs WITH POs and PSOs FOR MODULE-V:

Program
Course Program Outcomes Specific
Outcomes Outcomes
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3
CO 9 √ √ - - - - - - - - - - - - -
CO 10 - √ - - - - - - - - - - √ - -
CO 11 - √ √ - - - - - - - - - - - -
CO 12 √ - - - - - - - - - - - - - -

CREC Dept. of ECE Page 103

 Lecture Notes MODULE – V Signals & Systems
SAMPLING THEOREM
Graphical and analytical proof for Band Limited Signals, Impulse Sampling, Natural and Flat top Sampling,
Reconstruction of signal from its samples, Effect of under sampling – Aliasing, Introduction to Band Pass
Sampling. Correlation: Cross Correlation and Auto Correlation of Functions, Properties of Correlation
Functions, Energy Density Spectrum, Parseval’s Theorem, Power Density Spectrum, Relation between
Autocorrelation Function and Energy/Power Spectral Density Function, Relation between Convolution and
Correlation, Detection of Periodic Signals in the presence of Noise by Correlation, Extraction of Signal from
Noise by filtering.

Graphical and analytical proof for Band Limited Signals:

Sampling theorem: A continuous time signal can be represented in its samples and can be recovered back when
sampling frequency fs is greater than or equal to the twice the highest frequency component of message signal. i. e.
fs≥2fm
Proof: Consider a continuous time signal x(t). The spectrum of x(t) is a band limited to fm Hz i.e. the spectrum of
x(t) is zero for |ω|>ωm.Sampling of input signal x(t) can be obtained by multiplying x(t) with an impulse train δ(t) of
period Ts. The output of multiplier is a discrete signal called sampled signal which is represented with y(t) in the
following diagrams:

Here, you can observe that the sampled signal takes the period of impulse. The process of sampling can be explained
by the following mathematical expression:

CREC Dept. of ECE Page 104

 Lecture Notes Signals & Systems

Take Fourier transform on both sides.

To reconstruct x(t), you must recover input signal spectrum X(ω) from sampled signal spectrum Y(ω), which is
possible when there is no overlapping between the cycles of Y(ω).
There are three types of sampling techniques:
 Impulse sampling.
CREC Dept. of ECE Page 105
  Natural sampling.
Lecture Notes Signals & Systems
 Flat Top sampling.

Impulse Sampling

Impulse sampling can be performed by multiplying input signal x(t) with impulse train of
period 'T'. Here, the amplitude of impulse changes with respect to amplitude of input signal x(t). The output of
sampler is given by

To get the
spectrum of sampled signal, consider Fourier transform of equation 1 on both sides

This is called ideal sampling or impulse sampling. You cannot use this practically because pulse width cannot be
zero and the generation of impulse train is not possible practically.
Natural Sampling
Natural sampling is similar to impulse sampling, except the impulse train is replaced by pulse train of period T. i.e.
you multiply input signal x(t) to pulse train

CREC Dept. of ECE Page 106

 Lecture Notes Signals & Systems

Flat Top Sampling

During transmission, noise is introduced at top of the transmission pulse which can be easily removed if the pulse is
in the form of flat top. Here, the top of the samples are flat i.e. they have constant amplitude. Hence, it is called as
flat top sampling or practical sampling. Flat top sampling makes use of sample and hold circuit.

Theoretically, the sampled signal can be obtained by convolution of rectangular pulse p(t) with ideally sampled
signal say yδ(t) as shown in the diagram:

CREC Dept. of ECE Page 107

 Lecture Notes Signals & Systems

Nyquist Rate
It is the minimum sampling rate at which signal can be converted into samples and can be recovered back without
distortion.
Nyquist rate fN = 2fm hz
Nyquist interval = 1/fN = 1/2fm seconds.
Reconstruction of signal from its samples:
Reconstruction
Assume that the Nyquist requirement ω0 > 2ωm is satisfied. We consider two reconstruction schemes:
• ideal reconstruction (with ideal bandlimited interpolation),
• reconstruction with zero-order hold.
Ideal Reconstruction: Shannon interpolation formula

Our ideal reconstruction filter has the frequency response:

and, consequently, the impulse response

CREC Dept. of ECE Page 108

 Lecture Notes Signals & Systems
Now, the reconstructed signal is

which is the Shannon interpolation (reconstruction) formula. The actual reconstruction system mixes continuous and
discrete time.

The reconstructed signal xr(t) is a train of sinc pulses scaled by the samples x[n]. • This system is difficult to
implement because each sinc pulse extends over a long (theoretically infinite) time interval.

A general reconstruction filter

For the development of the theory, it is handy to consider the impulse-sampled signal xP(t) and its CTFT.

Figure : Reconstruction in the frequency domain is lowpass filtering

CREC Dept. of ECE Page 109

 Lecture Notes Signals & Systems

Effect of under sampling – Aliasing

Possibility of sampled frequency spectrum with different conditions is given by the following diagrams:

Aliasing Effect
The overlapped region in case of under sampling represents aliasing effect, which can be removed by
 considering fs >2fm

 By using anti aliasing filters.

Samplings of Band Pass Signals

In case of band pass signals, the spectrum of band pass signal X[ω] = 0 for the frequencies outside the range f1 ≤ f ≤
f2. The frequency f1 is always greater than zero. Plus, there is no aliasing effect when fs > 2f2. But it has two
disadvantages:
 The sampling rate is large in proportion with f2. This has practical limitations.

 The sampled signal spectrum has spectral gaps.

To overcome this, the band pass theorem states that the input signal x(t) can be converted into its samples and can be
recovered back without distortion when sampling frequency fs < 2f2.
Also,

CREC Dept. of ECE Page 110

Lecture Notes Signals & Systems

CREC Dept. of ECE Page 111

 Lecture Notes Signals & Systems

Correlation
Cross Correlation and Auto Correlation of Functions:
Correlation
Correlation is a measure of similarity between two signals. The general formula for correlation is

There are two types of correlation:

 Auto correlation

CREC Dept. of ECE Page 112

  Cross correlation
Lecture Notes Signals & Systems
Auto Correlation Function
It is defined as correlation of a signal with itself. Auto correlation function is a measure of similarity between a
signal & its time delayed version. It is represented with R(τ).
Consider a signals x(t). The auto correlation function of x(t) with its time delayed version is given by

Where τ = searching or scanning or delay parameter.

If the signal is complex then auto correlation function is given by

Cross Correlation Function

Cross correlation is the measure of similarity between two different signals.
Consider two signals x1(t) and x2(t). The cross correlation of these two signals R12(τ)R12(η) is given by

CREC Dept. of ECE Page 113

 Lecture Notes Signals & Systems

Properties of Correlation Functions:

 Auto correlation exhibits conjugate symmetry i.e. R (τ ) = R*(-τ )

CREC Dept. of ECE Page 114

  Auto correlation function of energy signal at origin i.e. at τ =0 is equal to total energy of that signal, which is
Lecture
given as:Notes Signals & Systems

 Auto correlation function is maximum at τ =0 i.e |R (τ ) | ≤ R (0) ∀ τ

 Auto correlation function and energy spectral densities are Fourier transform pairs. i.e.
F.T[R(τ)]=SXX(ω)
SXX(ω)= ∫R(τ)e−jωτdτ where -∞ < τ<∞

 R(τ)=x(τ)∗x(−τ)

Properties of Cross Correlation Function

 Auto correlation exhibits conjugate symmetry i.e. R12(τ)=R∗21(−τ).

 Cross correlation is not commutative like convolution i.e.

R12(τ)≠R21(−τ)
 If R12(0) = 0 means, if ∫x1(t)x∗2(t)dt=0 over interval(-∞,∞), then the two signals are said to be orthogonal.

 Cross correlation function corresponds to the multiplication of spectrums of one signal to the complex
conjugate of spectrum of another signal. i.e.

R12(τ)←→X1(ω)X∗2(ω)
This also called as correlation theorem.
Energy Density Spectrum:

CREC Dept. of ECE Page 115

Energy spectral density describes how the energy of a signal or a time series is distributed with frequency. Here, the
Lecture
term Notes
energy is used in the generalized sense of signal processing; Signals & Systems
Energy density spectrum can be calculated using the formula:

Parseval’s Theorem:

Power Density Spectrum

The above definition of energy spectral density is suitable for transients (pulse-like signals) whose energy is
concentrated around one time window; then the Fourier transforms of the signals generally exist. For continuous
signals over all time, such as stationary processes, one must rather define the power spectral density (PSD); this
describes how power of a signal or time series is distributed over frequency, as in the simple example given
previously. Here, power can be the actual physical power, or more often, for convenience with abstract signals, is
simply identified with the squared value of the signal.
Power density spectrum can be calculated by using the formula:

 The spectrum of a real valued process (or even a complex process using the above definition) is real and
an even function of frequency:

CREC Dept. of ECE Page 116

  If the process is continuous and purely indeterministic, the autocovariance function can be reconstructed by
Lecture
using theNotes
Inverse Fourier transform Signals & Systems
 The PSD can be used to compute the variance (net power) of a process by integrating over frequency:

CREC Dept. of ECE Page 117

 Lecture Notes Signals & Systems

Relation between Autocorrelation Function and Energy/Power Spectral Density Function:

1. Relation between Autocorrelation Function and Energy Spectral Density Function:

CREC Dept. of ECE Page 118

 Lecture Notes Signals & Systems

2. Relation between Autocorrelation Function and Power Spectral Density Function:

CREC Dept. of ECE Page 119

Lecture Notes Signals & Systems

CREC Dept. of ECE Page 120

 Lecture Notes Signals & Systems
Relation between Convolution and Correlation:

Detection of Periodic Signals in the presence of Noise by Correlation:

Extraction of Signal from Noise by filtering.

Whenever we wish to use correlation for signal detection, we use a two-part system. The first part of
the system performs the correlation and produces the correlation value or correlation signal, depending upon
whether we are doing in-place or running correlation. The second part of the system examines the correlation or
correlation signal and makes a decision or sequence of decisions. See the block diagram given in Figure

CREC Dept. of ECE Page 121

CREC Dept. of ECE Page 94
 Lecture Notes Signals & Systems

Z-Transform Properties:

Z-Transform has following properties:

Linearity Property:

Time Shifting Property:

Multiplication by Exponential Sequence Property

Time Reversal Property

CREC Dept. of ECE Page 95

CREC Dept. of ECE Page 96
 Lecture Notes Signals & Systems

Differentiation in Z-Domain OR Multiplication by n Property

Convolution Property

Correlation Property

Initial Value and Final Value Theorems

Initial value and final value theorems of z-transform are defined for causal signal.

Initial Value Theorem

For a causal signal x(n), the initial value theorem states that

This is used to find the initial value of the signal without taking inverse z-transform

CREC Dept. of ECE Page 96

 Lecture Notes Signals & Systems

Final Value Theorem

For a causal signal x(n), the final value theorem states that

This is used to find the final value of the signal without taking inverse z-transform

Region of Convergence (ROC) of Z-Transform

The range of variation of z for which z-transform converges is called region of convergence of z-
transform.

Properties of ROC of Z-Transforms

 ROC of z-transform is indicated with circle in z-plane.

 ROC does not contain any poles.

 If x(n) is a finite duration causal sequence or right sided sequence, then the ROC is entire
z-plane except at z = 0.

 If x(n) is a finite duration anti-causal sequence or left sided sequence, then the ROC is
entire z-plane except at z = ∞.

 If x(n) is a infinite duration causal sequence, ROC is exterior of the circle with radius a.
i.e. |z| > a.

 If x(n) is a infinite duration anti-causal sequence, ROC is interior of the circle with radius
a. i.e. |z| < a.

 If x(n) is a finite duration two sided sequence, then the ROC is entire z-plane except at z
= 0 & z = ∞.

The concept of ROC can be explained by the following example:

Example 1: Find z-transform and ROC of a n u[n]+a − nu[−n−1] anu[n]+a−nu[−n−1]

The plot of ROC has two conditions as a > 1 and a < 1, as we do not know a.
CREC Dept. of ECE Page 97
Lecture Notes Signals & Systems

CREC Dept. of ECE Page 98

 Lecture Notes Signals & Systems

In this case, there is no combination ROC.

Here, the combination of ROC is from a<|z|<1/a

Hence for this problem, z-transform is possible when a < 1.

Causality and Stability

Causality condition for discrete time LTI systems is as follows:

A discrete time LTI system is causal when

 ROC is outside the outermost pole.

 In The transfer function H[Z], the order of numerator cannot be grater than the order of
denominator.

CREC Dept. of ECE Page 99

 Lecture Notes Signals & Systems

Stability Condition for Discrete Time LTI Systems

A discrete time LTI system is stable when

 its system function H[Z] include unit circle |z|=1.

 all poles of the transfer function lay inside the unit circle |z|=1.
Z-Transform of Basic Signals

CREC Dept. of ECE Page

100
[Type here]

Some Properties of the Z- Transform:

Inverse Z transform:
Three different methods are:
4. Partial fraction method
5. Power series method
6. Long division method

[Type here]
[Type here]

[Type here]
[Type here]

[Type here]
[Type here]

[Type here]
[Type here]

[Type here]
[Type here]

Example: A finite sequence x [ n ] is defined as

Find X(z) and its ROC.

Sol: We know that

[Type here]
[Type here]

For z not equal to zero or infinity, each term in X(z) will be finite and consequently X(z) will
converge. Note that X ( z ) includes both positive powers of z and negative powers of z. Thus, from
the result we conclude that the ROC of X ( z ) is 0 < lzl < m.
Example: Consider the sequence

Find X ( z ) and plot the poles and zeros of X(z).

Sol:

From the above equation we see that there is a pole of ( N - 1)th order at z = 0 and a pole at z = a . Since
x[n] is a finite sequence and is zero for n < 0, the ROC is IzI > 0. The N roots of the numerator
polynomial are at

The root at k = 0 cancels the pole at z = a. The remaining zeros of X ( z ) are at

The pole-zero plot is shown in the following figure with N=8

[Type here]
[Type here]

MODULE – V
SAMPLING THEOREM
Graphical and analytical proof for Band Limited Signals, Impulse Sampling, Natural and Flat top
Sampling, Reconstruction of signal from its samples, Effect of under sampling – Aliasing,
Introduction to Band Pass Sampling. Correlation: Cross Correlation and Auto Correlation of
Functions, Properties of Correlation Functions, Energy Density Spectrum, Parseval’s Theorem,
Power Density Spectrum, Relation between Autocorrelation Function and Energy/Power Spectral
Density Function, Relation between Convolution and Correlation, Detection of Periodic Signals in
the presence of Noise by Correlation, Extraction of Signal from Noise by filtering.
Graphical and analytical proof for Band Limited Signals:
Sampling theorem: A continuous time signal can be represented in its samples and can be recovered back
when sampling frequency fs is greater than or equal to the twice the highest frequency component of message
signal. i. e.
fs≥2fm
Proof: Consider a continuous time signal x(t). The spectrum of x(t) is a band limited to fm Hz i.e. the spectrum
of x(t) is zero for |ω|>ωm.Sampling of input signal x(t) can be obtained by multiplying x(t) with an impulse train
δ(t) of period Ts. The output of multiplier is a discrete signal called sampled signal which is represented with y(t)
in the following diagrams:

Here, you can observe that the sampled signal takes the period of impulse. The process of sampling can be
explained by the following mathematical expression:

[Type here]
[Type here]

Take Fourier transform on both sides.

To reconstruct x(t), you must recover input signal spectrum X(ω) from sampled signal spectrum Y(ω), which is
possible when there is no overlapping between the cycles of Y(ω).
[Type here]
[Type here]

There are three types of sampling techniques:

 Impulse sampling.

 Natural sampling.

 Flat Top sampling.

Impulse Sampling

Impulse sampling can be performed by multiplying input signal x(t) with impulse train of
period 'T'. Here, the amplitude of impulse changes with respect to amplitude of input signal x(t). The output of
sampler is given by

To get the
spectrum of sampled signal, consider Fourier transform of equation 1 on both sides

This is called ideal sampling or impulse sampling. You cannot use this practically because pulse width cannot be
zero and the generation of impulse train is not possible practically.
Natural Sampling
Natural sampling is similar to impulse sampling, except the impulse train is replaced by pulse train of period T.
i.e. you multiply input signal x(t) to pulse train

[Type here]
[Type here]

Flat Top Sampling

During transmission, noise is introduced at top of the transmission pulse which can be easily removed if the
pulse is in the form of flat top. Here, the top of the samples are flat i.e. they have constant amplitude. Hence, it is
called as flat top sampling or practical sampling. Flat top sampling makes use of sample and hold circuit.

Theoretically,
the sampled signal can be obtained by convolution of rectangular pulse p(t) with ideally sampled signal say yδ(t)
as shown in the diagram:

[Type here]
[Type here]

Nyquist Rate
It is the minimum sampling rate at which signal can be converted into samples and can be recovered back
without distortion.
Nyquist rate fN = 2fm hz
Nyquist interval = 1/fN = 1/2fm seconds.
Reconstruction of signal from its samples:
Reconstruction
Assume that the Nyquist requirement ω0 > 2ωm is satisfied. We consider two reconstruction schemes:
• ideal reconstruction (with ideal bandlimited interpolation),
• reconstruction with zero-order hold.
Ideal Reconstruction: Shannon interpolation formula

Our ideal reconstruction filter has the frequency response:

[Type here]
[Type here]

and, consequently, the impulse response

Now, the reconstructed signal is

which is the Shannon interpolation (reconstruction) formula. The actual reconstruction system mixes continuous
and discrete time.

The reconstructed signal xr(t) is a train of sinc pulses scaled by the samples x[n]. • This system is difficult to
implement because each sinc pulse extends over a long (theoretically infinite) time interval.

A general reconstruction filter

For the development of the theory, it is handy to consider the impulse-sampled signal xP(t) and its CTFT.

[Type here]
[Type here]

Figure : Reconstruction in the frequency domain is lowpass filtering

Effect of under sampling – Aliasing

Possibility of sampled frequency spectrum with different conditions is given by the following diagrams:

[Type here]
[Type here]

Aliasing Effect
The overlapped region in case of under sampling represents aliasing effect, which can be removed by
 considering fs >2fm

 By using anti aliasing filters.

Samplings of Band Pass Signals

In case of band pass signals, the spectrum of band pass signal X[ω] = 0 for the frequencies outside the range f1 ≤
f ≤ f2. The frequency f1 is always greater than zero. Plus, there is no aliasing effect when fs > 2f2. But it has two
disadvantages:
 The sampling rate is large in proportion with f2. This has practical limitations.

 The sampled signal spectrum has spectral gaps.

To overcome this, the band pass theorem states that the input signal x(t) can be converted into its samples and
can be recovered back without distortion when sampling frequency fs < 2f2.
Also,

[Type here]
[Type here]

[Type here]
[Type here]

Correlation
Cross Correlation and Auto Correlation of Functions:
Correlation
Correlation is a measure of similarity between two signals. The general formula for correlation is

There are two types of correlation:

[Type here]
[Type here]

 Auto correlation

 Cross correlation

Auto Correlation Function

It is defined as correlation of a signal with itself. Auto correlation function is a measure of similarity between a
signal & its time delayed version. It is represented with R(τ).
Consider a signals x(t). The auto correlation function of x(t) with its time delayed version is given by

Where τ = searching or scanning or delay parameter.

If the signal is complex then auto correlation function is given by

Cross Correlation Function

Cross correlation is the measure of similarity between two different signals.
Consider two signals x1(t) and x2(t). The cross correlation of these two signals R12(τ)R12(η) is given by

[Type here]
[Type here]

Properties of Correlation Functions:

 Auto correlation exhibits conjugate symmetry i.e. R (τ ) = R*(-τ )

[Type here]
[Type here]

 Auto correlation function of energy signal at origin i.e. at τ =0 is equal to total energy of that
signal, which is given as:

 Auto correlation function is maximum at τ =0 i.e |R (τ ) | ≤ R (0) ∀ τ

 Auto correlation function and energy spectral densities are Fourier transform pairs. i.e.
F.T[R(τ)]=SXX(ω)
SXX(ω)= ∫R(τ)e−jωτdτ where -∞ < τ<∞

 R(τ)=x(τ)∗x(−τ)

[Type here]
[Type here]

Properties of Cross Correlation Function

 Auto correlation exhibits conjugate symmetry i.e. R12(τ)=R∗21(−τ).

 Cross correlation is not commutative like convolution i.e.

R12(τ)≠R21(−τ)
 If R12(0) = 0 means, if ∫x1(t)x∗2(t)dt=0 over interval(-∞,∞), then the two signals are said to be orthogonal.

 Cross correlation function corresponds to the multiplication of spectrums of one signal to the complex
conjugate of spectrum of another signal. i.e.

R12(τ)←→X1(ω)X∗2(ω)
This also called as correlation theorem.
Energy Density Spectrum:
Energy spectral density describes how the energy of a signal or a time series is distributed with frequency. Here,
the term energy is used in the generalized sense of signal processing;
Energy density spectrum can be calculated using the formula:

Parseval’s Theorem:

[Type here]
[Type here]

Power Density Spectrum

The above definition of energy spectral density is suitable for transients (pulse-like signals) whose energy is
concentrated around one time window; then the Fourier transforms of the signals generally exist. For continuous
signals over all time, such as stationary processes, one must rather define the power spectral density (PSD); this
describes how power of a signal or time series is distributed over frequency, as in the simple example given
previously. Here, power can be the actual physical power, or more often, for convenience with abstract signals,
is simply identified with the squared value of the signal.
Power density spectrum can be calculated by using the formula:

 The spectrum of a real valued process (or even a complex process using the above definition) is real and
an even function of frequency:

 If the process is continuous and purely indeterministic, the autocovariance function can be reconstructed
by using the Inverse Fourier transform
 The PSD can be used to compute the variance (net power) of a process by integrating over frequency:

[Type here]
[Type here]

[Type here]
[Type here]

Relation between Autocorrelation Function and Energy/Power Spectral Density Function:

1. Relation between Autocorrelation Function and Energy Spectral Density Function:

2. Relation between Autocorrelation Function and Power Spectral Density Function:

[Type here]
[Type here]

Relation between Convolution and Correlation:

[Type here]
[Type here]

Detection of Periodic Signals in the presence of Noise by Correlation:

Extraction of Signal from Noise by filtering.

Whenever we wish to use correlation for signal detection, we use a two-part system. The first part of the
system performs the correlation and produces the correlation value or correlation signal, depending upon whether we
are doing in-place or running correlation. The second part of the system examines the correlation or correlation signal
and makes a decision or sequence of decisions. See the block diagram given in Figure

Correlation
Cross Correlation and Auto Correlation of Functions:
Correlation
Correlation is a measure of similarity between two signals. The general formula for correlation is

There are two types of correlation:

 Auto correlation

 Cross correlation

Auto Correlation Function

It is defined as correlation of a signal with itself. Auto correlation function is a measure of similarity between a
signal & its time delayed version. It is represented with R(τ).
Consider a signals x(t). The auto correlation function of x(t) with its time delayed version is given by
[Type here]
[Type here]

Where τ = searching or scanning or delay parameter.

If the signal is complex then auto correlation function is given by

Cross Correlation Function

Cross correlation is the measure of similarity between two different signals.
Consider two signals x1(t) and x2(t). The cross correlation of these two signals R12(τ)R12(η) is given by

[Type here]
[Type here]

Properties of Correlation Functions:

 Auto correlation exhibits conjugate symmetry i.e. R (τ ) = R*(-τ )

[Type here]
[Type here]

 Auto correlation function of energy signal at origin i.e. at τ =0 is equal to total energy of that
signal, which is given as:

 Auto correlation function is maximum at τ =0 i.e |R (τ ) | ≤ R (0) ∀ τ

 Auto correlation function and energy spectral densities are Fourier transform pairs. i.e.
F.T[R(τ)]=SXX(ω)
SXX(ω)= ∫R(τ)e−jωτdτ where -∞ < τ<∞

 R(τ)=x(τ)∗x(−τ)

[Type here]
[Type here]

Properties of Cross Correlation Function

 Auto correlation exhibits conjugate symmetry i.e. R12(τ)=R∗21(−τ).

 Cross correlation is not commutative like convolution i.e.

R12(τ)≠R21(−τ)
 If R12(0) = 0 means, if ∫x1(t)x∗2(t)dt=0 over interval(-∞,∞), then the two signals are said to be orthogonal.

 Cross correlation function corresponds to the multiplication of spectrums of one signal to the complex
conjugate of spectrum of another signal. i.e.

R12(τ)←→X1(ω)X∗2(ω)
This also called as correlation theorem.
Energy Density Spectrum:
Energy spectral density describes how the energy of a signal or a time series is distributed with frequency. Here,
the term energy is used in the generalized sense of signal processing;
Energy density spectrum can be calculated using the formula:

Parseval’s Theorem:

[Type here]
[Type here]

Power Density Spectrum

The above definition of energy spectral density is suitable for transients (pulse-like signals) whose energy is
concentrated around one time window; then the Fourier transforms of the signals generally exist. For continuous
signals over all time, such as stationary processes, one must rather define the power spectral density (PSD); this
describes how power of a signal or time series is distributed over frequency, as in the simple example given
previously. Here, power can be the actual physical power, or more often, for convenience with abstract signals,
is simply identified with the squared value of the signal.
Power density spectrum can be calculated by using the formula:

 The spectrum of a real valued process (or even a complex process using the above definition) is real and
an even function of frequency:

 If the process is continuous and purely indeterministic, the autocovariance function can be reconstructed
by using the Inverse Fourier transform
 The PSD can be used to compute the variance (net power) of a process by integrating over frequency:

[Type here]