Digital Communication
Dr. Mahadev S. Patil
Professor and Head
Department of ETC engineering
Kasegaon Education Society’s
RAJARAMBAPU INSTITUTE OF TECHNOLOGY,
Islampur, Dist. Sangli, Maharashtra, India - 415 414
Books
Text Books:
• [Link] Shanmugan, Digital & Analog Communication
Systems, Wiley India Edition
• R P Singh, S D Sapre, Communication System-Analog
& Digital, Tata Mc Graw Hill Publication
Reference Books:
• Bernard Sklar, Digital Communication-Fundamentals
and Applications, Pearson Education
• Taub & Schiling, Principles of communication System,
Tata McGraw Hill
Unit- I
Unit –I
Random Signal Theory
Course Outcomes:
1. Explain different concepts of digital communication
systems
Topic Learning Outcomes:
[Link] fundamentals of probability theory
Background- Think, Think, Think…
• What are the examples of random events?
• What are the examples of random signals?
• What is probability?
• Can you illustrate probability with example?
• Usually which examples are used in the study
of probability?
Background- Think, Think, Think…
• The probability of occurrence of 28th Feb in a
year is 1, probability of occurrence of 30th Feb
in a year is 0, probability of occurrence of 29th
Feb in a year is neither 0 not 1. It is 1/4
• Reflected signal in wave propagation are
random in nature, Noise introduced in channel
• Tossing of a coin-has two outcomes
• Throwing a cubic die or dice-has six outcomes
Random signals
• Signals whose behaviour cannot be predicted are
called random signals.
• Because precise value of these signals can not be
predicted in advance before they actually occur.
• Eg. Noise interferences in communication systems
• Noise generated by receiver itself
• Electromagnetic interference is the major source
• Thermal noise caused by random motion of the
electronics
• Reflected signals
Random Signals
• In the analysis of communication systems, we often
encounter random signals-signals whose behavior cannot
be predicted exactly.
• For random signals it is not possible to write an explicit
time function.
• However when examined over a long period, a random
signal may exhibit certain regularities that can be
described in terms of probabilities and statistical
averages.
• Thus although we lack an exact description, we may be
able to model such a signal in terms of average values
and the probability that the random signal will be in a
given range at a specific time.
Random Variables
Random variables can be discrete or continuous
• Discrete random variables have a countable
number of outcomes
– Examples: Dead/alive, treatment/placebo, dice, counts,
etc.
• Continuous random variables have an infinite
continuum of possible values.
– Examples: blood pressure, weight, the speed of a car,
the real numbers from 1 to 6.
Types
Types of RV
Random variables can be discrete or continuous
• Discrete random variables have a countable
number of outcomes
– Examples: Dead/alive, treatment/placebo, dice, counts,
etc.
• Continuous random variables have an infinite
continuum of possible values.
– Examples: blood pressure, weight, the speed of a car,
the real numbers from 1 to 6.
When a coin is tossed
Discrete RV
Illustration
Discrete example: roll of a die
p(x)
1/6
x
1 2 3 4 5 6
P(x) 1
all x
Shooting target by gun
Continuous RV
Two coins are tossed
Probability Mass Function (PMF)
Probability Mass Function (PMF)
– A set of probability value pi assigned to each of the
values taken by the discrete random variable xi
– Let be the probability of xi
Also f(x) such that
1. f(x)≥0
2. ∑ f(x) = 1
Then f(x) is called probability mass function of
discrete random variable
Question
Three fair coins are tossed and random variable ‘x’ is
Number of tails.
(a)Tabulate the pmf
(b)What is the probability of getting two tails.
Homework
Similarly, if three coins are tossed and
random variable x is number of heads then
sample space S and possible random
variable values will be same as shown in
example earlier.
• Tabulate PMF
• What is the probability of getting two
heads?
Shooting target by gun
Continuous RV
Mean, Variance and Standard deviation in RV
Let us see calculations of all these statistical averages
in random variables
Mean:
Variance
SD
Probability Density Function (PDF)
PDF
PDF…..
PDF…..
PDF
PDF…..
PDF…..
Q
Numerical
Random Process
• Let there be a random experiment E having outcome λ
from the sample space S. This means that λ is the subset of
S. Thus every time an experiment is conducted, the
outcome λ will be one of the sample point in sample space.
If this outcome λ is associated with time, then the function
of λ and time t is formed i.e., X(λ, t). Then the function
X(λ, t) is known as Random Process.
• Hence when any random experiment E is given a time
dimension, then each outcome appears at certain time and
the random experiment will be converted to Random
process.
• A random process is the function of two variables λ and t.
• A random variable that is a function of time is called a
random process.
Ensemble
• Here λ belongs to sample space S and -∞<t
< ∞. At some specific time say t=to, X(λ, to)
is a random variable whose value depends
upon λ. Also for a specific outcome λi there
is a single time function X(λi, to). This
function is known as sample function
whereas the collection of all the sample
functions is known as Ensemble.
Illustration
• Let us consider a set of voltage waveforms generated by thermal
electron motion in a large number of identical resistors.
• We can represent the voltage produced at each instant by the resistor as
v. Then the random process X(v, t) denotes the output voltages from
the set of resistors.
• Figure shows the waveforms produced by these resistors randomly and
different parameters of random process.
C
• To determine the statistics of random
process like mean or expected value two
approaches are used.
– Ensemble average
– Time average
Ensemble
Random Process
• Let there be a random experiment E having outcome λ
from the sample space S. This means that λ is the subset of
S. Thus every time an experiment is conducted, the
outcome λ will be one of the sample point in sample space.
If this outcome λ is associated with time, then the function
of λ and time t is formed i.e., X(λ, t). Then the function
X(λ, t) is known as Random Process.
• Hence when any random experiment E is given a time
dimension, then each outcome appears at certain time and
the random experiment will be converted to Random
process.
• A random process is the function of two variables λ and t.
• A random variable that is a function of time is called a
random process.
Ensemble
• Here λ belongs to sample space S and -∞<t
< ∞. At some specific time say t=to, X(λ, to)
is a random variable whose value depends
upon λ.
• Also for a specific outcome λi there is a
single time function X(λi, to).
• This function is known as sample function
whereas the collection of all the sample
functions is known as Ensemble.
Illustration
• Let us consider a set of voltage waveforms generated by thermal
electron motion in a large number of identical resistors.
• We can represent the voltage produced at each instant by the resistor as
v. Then the random process X(v, t) denotes the output voltages from
the set of resistors.
• Figure shows the waveforms produced by these resistors randomly and
different parameters of random process.
C
• To determine the statistics of random
process like mean or expected value two
approaches are used.
– Ensemble average
– Time average
Ensemble Averages
Ensemble
Time averages
Comparison
Stationary Random Process
• A random process X(t) is called stationary if its statistics are not
affected by any shift in the time origin. We may define stationary
process in terms of ensemble averages as:
a. The ensemble mean is independent of time. Mathematically,
mx(t)=mx(t1)=mx(t2)=mx(t3)=……..= constant at all the time instants
b. The autocorrelation function Rτ(t1,t2) depends only upon the time
differences t2-t1.
Rτ(t1,t2)= Rτ(t1 +t ,t2 + t )
• Therefore, the autocorrelation function of any stationary process is a
function of time-difference and is given as
Rτ(t1,t2)= Rτ(t2 + t1 )
• Thus the system of resistors producing thermal noise voltages
represents a stationary process due to the fact that the noise generated
by resistor does not depend on time.
Wide Sense Stationary process
• All the processes in practice, are non-stationary since every
process has some start and end.
• This means that the statistics of such process are dependent
upon time.
• A true stationary process should start at t = - ∞ and should
not stop till t = ∞. Such type of process is not possible
practically.
• The process may appear stationary over a certain period of
time. Then this will be wide sense stationary process.
•Wide sense stationary process is also known as weakly
stationary.
Ergodic Process
• A random process is known as ergodic process if the time-
average are equal to ensemble averages. Hence for a ergodic
process, we have
mx = < mx >
• For any ergodic process, all ensemble averages will be
equal
to corresponding time averages for any particular sample
function. The time averages are not a function of time.
• When time and ensemble averages are the same, it implies
that ensemble averages also are not a function of time.
• Thus an ergodic process is always stationary.
Ergodic Process
Illustration: Ergodic Process
• In an ergodic scenario, the average outcome of the group is the same
as the average outcome of the individual over time.
• An example of an ergodic systems would be the outcomes of a coin
toss (heads/tails). If 100 people flip a coin once or 1 person flips a
coin 100 times, you get the same outcome. (Though the
consequences of those outcomes (e.g. win/lose money) are typically
not ergodic)!
• A way to identify an ergodic situation is to ask do I get the same
result if I:
-look at one individual’s trajectory across time
-look at a bunch of individual’s trajectories at a single point in time
If yes: ergodic.
If not: non-ergodic.
Ergodic Process
Power Spectral Density (PSD)
Take away?
What have we
learnt today??
Thank you