Scribd
Upload
8 views

Random Process

Uploaded by

jaiswal.mohit27
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
8 views

Random Process

Uploaded by

jaiswal.mohit27
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

Random Process

• A random process is a collection of random


variables Xt indexed by time. Each realization of the process is
a function of t. For every fixed time t, Xt is a random variable.
• Random processes are classified as continuous
time or discrete-time, depending on whether time is continuous
or discrete. We typically notate continuous-time random
processes as {X(t)} and discrete-time processes as {X[n]}.
• Example: A sine wave having amplitude as R.V.
Basic Definitions
1.

2.

3.
Hermitian Symmetry of correlation and Covariance
functions

Notes:
1. Each correlation or covariance matrix of a random vector must be positive
definite.
2. The two-dimensional function g(t,s) is positive semidefinite if for all N>0 and
all t1<t2…<tN.

3. This is a necessary and sufficient condition.


4. Diagonal dominance
Power Spectral Density

• Time domain description of the second-order statistics of a


stochastic process: Auto correlation function.
• Frequency domain counter part of this statistical parameter is
the power spectral density. Also known as ‘Power spectrum’
or ‘Spectrum’.
Consider again a wide-sense stationary discrete-time stochastic
process : (a) mean=0 (b) auto-correlation function is denoted by
r(l) for lag l=…-2,-1,0,1,2,…

107
Power Spectral Density
• Let the infinitely long time series u(n) denote
a single realization of the process. Windowed
portion of this time series is given by writing
ìu(n), n = 0, ±1, ±2,... ± N
u N (n) = í
î0, n > N
• The discrete-time fourier transform of the
windowed time series u N (n) is given by
N
U N (w ) = åu N (n) e - jw n

n=- N
Angular frequency lies between –Pi to Pi. 108
Power Spectral Density
Complex conjugate of u N (n) can be given by
N
U N *(w ) = åu N *(k) e jw k

k=- N
From previous 2 equations,
N N

å åu
2 jw (n-k )
U N (w ) = N (n)u N *(k) e
n=- N k=- N

Each realization U N (n)produces such results.


Expected result is obtained by taking expectations
On both sides: 109
Power
N
Spectral
N
Density
E U N (w ) = å å E éëu N (n)u N (k) ùûe
é ù 2 * - jw (n-k )
ë û n=- N k=- N

ìï E éë u N (n)u N * (k) ùû = r(n - k), - N £ (n, k) £ N


rN (n - k) = í
ïî0 Otherwise
N N
E U N (w ) = å å r(n - k)e
é ù 2 - jw (n-k )
ë û n=- N k=- N
Let l=n-k and re-arranging above equation
110
Power Spectral Density
1 é N
æ l ö
E U N (w ) = å ç 1- ÷ r(l)e
ù
2 - jw l

N ë û l=- N è Nø
Above fourier transform is function of two time
functions: the autocorrelation functionrN (l) for lag
l, and a triangular window known as the Bartlett
Window, which is defined by
ìæ l ö
ïç 1- ÷ ; l £ N
w B (l) = íè N ø
ï0 l ³ N;
î
111
Power Spectral Density
If N is very high
¥
1 é
lim E U N (w ) = å r(l)e
ù
2 - jw l
N®¥ N ë û l=-¥
Based on above expression, we can define a
quantity 1
E é U N (w ) ù
2
S(w ) = lim
N®¥ N ë û

Which is spectral density of expected power!!


“Also known as power spectral density”
112
Poisson Process

• The Poisson process is one of the most widely used counting processes.
• It is usually used in scenarios where we are counting the occurrences of certain
events that appear to happen at a specific rate but entirely at random.
• For example, suppose that from a dataset, we know that a bus arrives at a
particular station at a rate of 3 per hour.
• The timings of the bus are completely random. Thus, we conclude that the Poisson
process might be a good model for counting the arrival of the bus. In practice, the
Poisson process or its extensions have been used to model following cases:

 The number of car accidents at a site or in an area;


 The location of users in a wireless network;
 The requests for individual documents on a web server;
 The outbreak of wars/disease;
 Photons landing on a photodiode.
Sample function of Poisson Process [0,∞) is given
below:

The probability Mass function of the Poisson counting process is given as:

We can note that it is the PMF of a Poisson random variable with mean λt, which is
also called intensity is given as:
Examples

1. The number of customers arriving at a grocery store can be modeled by a Poisson


process with intensity λ=10 customers per hour.
a. Find the probability that there are 2 customers between 10:00 and 10:20.
b. Find the probability of 3 customers between 10:00 and 10:20 and 7 customers
between 10:20 and 11.00 am.

2. In a earthquake prone area, the intensity of earth quake happening is 6


earthquakes in a month.
What is probability of having maximum 2 earth quakes in 3 months.
What is probability of having more than 2 earthquake in 3 months.
What is probability of having no earthquake in 3 months.

3. Let X be a random variable with zero mean and variance 1. Y is another R.V.
which is independent of X with mean 3 and variance 4. If Z=Y-X
(a) Calculate the correlation between Y and Z.
(b) Calculate the covariance between Y and Z.
4. Check whether the following function is a valid auto-correlation function
(a) 5 sin(nτ)
(b) 1/(1+9τ2)

5. Find the mean and variance of a stationary random process, whose auto correlation
function is given as

You might also like