0% found this document useful (0 votes)
86 views33 pages

Chapter 5

The document discusses random processes and their properties. It defines a random process as a mapping from a random experiment to an ensemble of sample functions, where each time instance corresponds to a random variable. It describes how to calculate the ensemble average and variance of a random process, and introduces the concepts of stationarity, ergodicity, time averages, autocorrelation, and power spectral density as important properties and analysis tools for random processes.

Uploaded by

Seham Raheel
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
86 views33 pages

Chapter 5

The document discusses random processes and their properties. It defines a random process as a mapping from a random experiment to an ensemble of sample functions, where each time instance corresponds to a random variable. It describes how to calculate the ensemble average and variance of a random process, and introduces the concepts of stationarity, ergodicity, time averages, autocorrelation, and power spectral density as important properties and analysis tools for random processes.

Uploaded by

Seham Raheel
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 33

5

5.1

Random Processes
2 2 4 5 11 11 15 17 28 28 29 30 32

Random Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.1 5.1.2 5.1.3 Denition and Examples . . . . . . . . . . . . . . . . . . . . . Ensemble Averages and Stationarity . . . . . . . . . . . . . . . Time Averages and Ergodicity . . . . . . . . . . . . . . . . . . Autocorrelation and Autocovariance . . . . . . . . . . . . . . . Power Spectral Density . . . . . . . . . . . . . . . . . . . . . Deterministic Power and Energy Signals . . . . . . . . . . . . .

5.2

Autocorrelation and Power Spectra . . . . . . . . . . . . . . . . . . . 5.2.1 5.2.2 5.2.3 5.2.4

Examples of Autocorrelation Functions and Power Spectral Densities 19 Expected Value of the Output Random Process . . . . . . . . .

5.3

Excitation of LTI Systems with Stationary Random Processes . . . . . . . 5.3.1 5.3.2 5.3.3 5.3.4 Autocorrelation Function of the Output Random Process . . . . . Power Spectral Density of the Output Random Process . . . . . . Cross-Correlation between Input and Output Random Process . . .

Dr. Tanja Karp

5.1
5.1.1 Denition and Examples

Random Processes
random variable

random experiment X (t , 1 )

1 1

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

* 1 * i * 3 * 2

* X (t , ) 2

0.5 0 0.5 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

* X (t , 3 )

1 0.5 0 0.5 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

sample space S

X (t , i )

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

t1
Dr. Tanja Karp 5.1 Random Processes

t2
2

Random Process: A random process X (t) describes the mapping of a random experiment with sample space S onto an ensemble of sample functions X (t, i). For each point in time t1, X (t1) describes a random variable. Example: Rolling a Die Random variable: X = i if number i is on top of the die Random process: Y (t) = X cos(0t). Example: Tossing a Coin N Times Random variable: Xn = 0 if the nth result is head, Xn = 1 if the nth result is tail PN Random process: Y (t) = n=1 Xn rect(t n + 0.5).
Y (t , 1 ) t Y (t , 1 ) t Y (t , 1 ) t Y (t , 4 ) t
Dr. Tanja Karp 5.1 Random Processes

Y (t , 2N ) t
3

Example: Filtering a Random Process

Y (t) = h(t) X (t) =

h( )X (t )d

X (t), Y (t): random processes, h(t): lter impulse response


5.1.2 Ensemble Averages and Stationarity

For each time instance of a random process, the average value, variance etc. can be calculated from all sample functions X (t, i). Expected Value E {X (t)}: For a random process X (t) with probability density function fX (t)(x), the expected value E {X (t)} = mX (t) is given by: Z x fX (t)(x)dx = mX (t) E {X (t)} =

Variance X (t):
Z

X (t) = E {|X (t) mX (t)| } =

Dr. Tanja Karp 5.1

|x mX (t)| fX (t)(x)dx

Random Processes

For a stationary random process the probability density function is independent of time t, thus the expected value and the variance are also a constant over time.

fX (t)(x) = fX (t+t0)(x), mX (t) = mX (t + t0) = mX ,


5.1.3 Time Averages and Ergodicity

t, t0

X (t) = X (t + t0) = X

So far, the average value and the variance of a random process X (t) were calculated based on the probability density function fX (t). However, in practical experiments the probability density function of a random process is often unknown. Also, in many cases, there is only one sample function X (t, i) available. Therefore, it is favorable to average over time instead of taking the ensemble average. Average Value mX (i):

mX (i)
2 Variance X ( ) : i 2 X ( ) i

1 = lim T T

T /2

X (t, i)dt
T / 2

1 = lim T T

T /2 T / 2

(X (t, i) mX (i)) dt

Dr. Tanja Karp

5.1

Random Processes

ensemble average

random experiment X (t , 1 )

1 1

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

* 1 * i * 3 * 2

* X (t , ) 2

0.5 0 0.5 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

* X (t , 3 )

1 0.5 0 0.5 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

sample space S

X (t , i )

time average
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

t1

t2

Dr. Tanja Karp

5.1

Random Processes

Ergodicity: A stationary random process X (t) is called ergodic, if the time averages of each sample function X (t, i) converge towards the corresponding ensemble average with probability one. In practical applications ergodicity is often assumed since just one sample function is available and therefore the ensemble averages cannot be taken. Example 1: Random process: X (t) = A cos(0t) A: discrete random variable with P (A = 1) = P (A = 2) = 0.5 0: constant Ensemble average:

mX (t) = E {X (t)} = E {A} cos(0t) = 1.5 cos(0t)


For 0 = 0 the random process is not stationary and we are not allowed to take the time average.

Dr. Tanja Karp

5.1

Random Processes

Example 2: Random process: X (t) = A A: discrete random variable with P (A = 1) = P (A = 2) = 0.5 Ensemble average:

mX (t) = E {X (t)} = E {A} = 1.5 the ensemble average is independent of time.


Time averages:

mX (1)

1 = lim T T 1 = lim T T

1 X (t, 1)dt = lim T T T / 2 1 X (t, 2)dt = lim T T T / 2


T /2

T /2

T /2

1 dt = 1
T / 2

T /2

mX (2)

2 dt = 2
T / 2

time averages taken for dierent sample functions are not identical to the ensemble average, the random process is thus not ergodic.

Dr. Tanja Karp

5.1

Random Processes

Example 3: Tossing a Coin N Times Random variable: Xn = 0 if the nth result is head, Xn = 1 if the nth result is tail. N P Xn rect(t n + 0.5) Random process: Y (t) =
n=1

Y (t , 1 ) t Y (t , 1 ) t Y (t , 1 ) t Y (t , 4 ) t Y (t , 2N ) t

Ensemble average:

mY (t) = E {Y (t)} = 0.5


Time average: Sample function Y (t, i): Coin is tossed N times and we observe n1 times head and n2 times tail. Z N 1 1 n2 m Y ( i ) = Y (t, i)dt = (n1 0 + n2 1) = N 0 N N

for N , n1 and n2 converge towards N/2 and mY (t,i) = mY (t) = mY . The random process is thus ergodic.
Dr. Tanja Karp 5.1 Random Processes 9

random process X (t ) with pdf fX (t ) (x)

stationary random process with pdf f X (x)

ergodic random process: time average=ensemble average

Dr. Tanja Karp

5.1

Random Processes

10

5.2
5.2.1

Autocorrelation and Power Spectra

Autocorrelation and Autocovariance


1

We are interested in how the value of a random process X (t) evaluated at t2 depends on its value at time t1. At t1 and t2 the random process is characterized by random variables X1 and X2, respectively. The relationship between X1 and X2 is given by the joint probability density function

X (t , 1 )

1 1 0.5

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

X (t , 2 )

0 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

0.5 1

X (t , 3 )

0.5 0 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

0.5

fX1 X2 (x1, x2)


Autocorrelation Function:
X (t , i )

RXX (t1, t2) = E {X1 X2} Z = E {X (t1) X (t2)} =

Dr. Tanja Karp

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

t1

t2

x1x2fX1X2 (x1, x2)dx1dx2


11

5.2 Autocorrelation and Power Spectra

Autocovariance Function:

CXX (t1, t2) = E {(X (t1) mX (t1)) (X (t2) mX (t2))} Z Z = (x1 mX (t1))(x2 mX (t2)) fX (t1)X (t2)(x1, x2)dx1dx2

= RXX (t1, t2) mX (t1)mX (t2)


2 CXX (t, t) describes the variance X (t) of a random process.

Autocorrelation and Autocovariance Function of a Stationary Random Process: The joint probability density function of a stationary process does not change if a constant value t is added to both t1 and t2.

fX1 X2 (x1, x2) = fX (t1)X (t2)(x1, x2) = fX (t1+t)X (t2+t)(x1, x2)


The autocorrelation function then only depends on the dierence between t1 and t2 Z Z RXX (t1, t2) = E {X (t1) X (t2)} = x1x2fX (t1)X (t2)(x1, x2)dx1dx2
Z Z

=

Dr. Tanja Karp

x1x2fX (0)X (t2t1)(x1, x2)dx1dx2 = RX,X (0, t2 t1) = RX,X ( )

5.2 Autocorrelation and Power Spectra

12

Since the average value is a constant, the autocovariance function is given by:

CXX (t1, t2) = E {(X (t1) mX ) (X (t2) mX )} Z Z = (x1 mX )(x2 mX ) fX (t1)X (t2)(x1, x2)dx1dx2
Z Z

(x1 mX )(x2 mX ) fX (0)X (t2t1)(x1, x2)dx1dx2

= CXX (0, t2 t1) = CXX ( )

Properties of the Autocorrelation Function of a Stationary Random Process:

Symmetry: RXX ( ) = RXX ( ) Mean Square Average: RXX (0) = E {X (t)2} 0 Maximum: RXX (0) |RXX ( )| Periodicity: if RXX (0) = RXX (t0), then RXX ( ) is periodic with period t0.

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

13

Wide Sense Stationary (WSS) Random Process: A random process X (t) is called WSS if the following three properties are satised:

The average value of the random process is a constant: mX (t) = mX The autocorrelation and autocovariance function only depend on the time dierence = t1 t2: RXX (t1, t2) = RXX (t2 t1) = RXX ( ) CXX (t1, t2) = CXX (t2 t1) = CXX ( )
2 The variance is constant and nite: X = CXX (0) = RXX (0) m2 X <

Autocorrelation and Autocovariance Function of an Ergodic Random Process:

1 RXX ( ) = lim T T 1 T T
T /2 Z

T /2 Z

XT (t, i)XT (t + , i)dt


T / 2

CXX ( ) = lim

(XT (t, i) mX )(XT (t + , i) mX )dt


T / 2

XT (t, i): sample function of random process X (t) windowed to be of length T (starting at T /2 ending at T /2).
Dr. Tanja Karp 5.2 Autocorrelation and Power Spectra 14

5.2.2

Power Spectral Density

Motivation:

Description of random processes in the frequency domain Calculation of the Fourier Transform of a sample function is not useful We assume in the following that the random process considered is at least WSS if not stationary.
The power spectral density (psd) of a WSS random process X (t) is dened as the Fourier Transform of the autocorrelation function RXX ( ): Z j 2f SXX (f ) = F {RXX ( )} = RXX ( ) e d

Inverse transform:
Z

RXX ( ) =

SXX (f ) e

j 2f

df

Properties of the Power Spectral Density:

SXX (f ) = SXX (f ),
Dr. Tanja Karp

SXX (f ) 0,

Im{SXX (f )} = 0
15

5.2 Autocorrelation and Power Spectra

Ergodic Random Process x(t): Autocorrelation Function: T /2 T /2 Z Z 1 1 xT (t, i)xT (t + , i)dt = lim xT (t)xT (t + )dt Rxx( ) = lim T T T T
T / 2 T / 2

Power Spectral Density:

Sxx(f ) = Z

Rxx( ) e

j 2f

d
j 2f

1 = lim T T 1 = lim T T Z

T /2 T / 2

xT (t)xT (t + )dt e Z xT (t)

T /2 T / 2

x T (t + ) e | {z xT (t) e
j 2f t

j 2f

d dt }

1 = XT (f ) lim T T

XT (f )ej 2f t T /2 T / 2

dt

|XT (f )|2 XT (f )XT (f ) = lim = lim T T T T

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

16

5.2.3

Deterministic Power and Energy Signals

Power Signal: The autocorrelation function and the power spectral density can also be calculated for a deterministic power signal x(t). In this case, the signal simply replaces the random process. With xT (t) = x(t) rect(t/T ) we obtain for the autocorrelation function:

1 Rxx( ) = lim T T

T /2 Z

xT (t)xT (t + )dt
T / 2

and for the power spectral density: Z XT (f )XT (f ) |XT (f )|2 Sxx(f ) = Rxx( ) exp(j 2f )d = lim = lim T T T T

Note that we obtain the power of x(t) as

1 P = Rxx(0) = lim T T

T /2 Z

xT (t)dt =
T / 2

Sxx(f )df

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

17

Energy Signals: For an energy signal x(t) an energy autocorrelation function can be dened as Z E Rxx( ) = x(t)f (t + )dt = x( ) x( )

Applying the Fourier Transform to the energy autocorrelation function, we obtain the energy spectral density as: Z E E 2 Sxx(f ) = Rxx( ) exp(j 2f )d = X (f )X (f ) = |X (f )|

We obtain the energy as


Z

E = Rxx(0) =

x(t) dt =

1 2

Sxx(f )df =

|X (f )| df

which restates Parsevals Theorem.

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

18

5.2.4

Examples of Autocorrelation Functions and Power Spectral Densities

Example 1: Sinusoid with Random Phase Angle Random process:

x(t) = A sin(0t + ) A, 0: constant values, : random variable with probability density function f(x): 1/2 for x < f(x) = 0 otherwise
Average value:

mx(t) = E {x(t)} = E {A sin(0t + )} Z =


A sin(0t + x)f(x) dx Z

1 A sin(0t + x) dx = 0 2

The average value is a constant and independent of t. Since mx = 0 the autocorrelation and autocovariance functions are identical.
Dr. Tanja Karp 5.2 Autocorrelation and Power Spectra 19

Autocorrelation function:

Rxx(t1, t2) = E {x(t1)x(t2)} = E {A sin(0t1 + ) A sin(0t2 + )}


Applying:

2 sin(A) sin(B ) = cos(A B ) cos(A + B )


2 2

Rxx(t1, t2) = E {0.5A cos(0(t2 t1)) 0.5A cos(0(t2 + t1) + 2)} = 0.5A E {cos(0(t2 t1))} 0.5A E {cos(0(t2 + t1) + 2)} = 0.5A cos(0(t2 t1)) 0 = Rxx( )
The autocorrelation function only depends on = t2 t1 but not on the absolute values of t1 and t2. Power spectral density:
2 2 2

Sxx(f ) = F{Rxx( )} = F {0.5A cos(0 )} = 0.5A F {cos(0 )} = 0.5A ( (f f0) + (f + f0))


Ergodicity: If the random process is ergodic, we obtain the same results for the average value and the autocorrelation function by taking the time averages over one sample function.
Dr. Tanja Karp 5.2 Autocorrelation and Power Spectra 20

Let us assume the sample function x(t, i) has a phase angle i:

x(t, i) = A sin(0t + i)
Average value (time average):

mx(i)

Z T /2 1 = lim x(t, i)dt T T T / 2 Z T /2 1 = lim A sin(0t + i)dt = 0 = mx T T T / 2

Autocorrelation function (time average): Z T /2 1 Rx(i)x(i)( ) = lim x(t, i)x(t + , i)dt T T T / 2 Z T /2 1 2 A sin(0t + i) sin(0(t + ) + i)dt = lim T T T / 2 Z T /2 A2 1 = lim cos(0 ) + cos(0(2t + ) + 2i)dt 2 T T T /2 A2 cos(0 ) = Rxx( ) = 2 Time averages of one sample function and ensemble averages are identical the random process is ergodic.
Dr. Tanja Karp 5.2 Autocorrelation and Power Spectra 21

A = 1, 0 = 2
Sample Functions 1 x(t,1) 0 1 1 x(t,2) 0 1 1 x(t,3) 0 1 1 x(t,4) 0 1 1 x(t,5) 0 1 0.5 Autocorrelation Function Rxx()

0.5

1.5

2.5

3.5

4.5

5 0

0.5

1.5

2.5

3.5

4.5

5 0.5 3 2 1 0 1 2 3

Power Spectral Density S (f)


xx

0.5

1.5

2.5

3.5

4.5

0.7 0.6 0.5 0.4

0.5

1.5

2.5

3.5

4.5

0.3 0.2 0.1

0.5

1.5

2.5 t in sec

3.5

4.5

0 3

0 f

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

22

Example 2: Binary Data Transmission A binary sequence is transmitted by rectangular pulses of width Tb. The amplitude of the pulse is determined by each bit, i.e. it is one if the bit is one and zero if the bit is zero. We assume that ones and zeros are equally likely and that each bit is statistically independent of all others. Using ergodicity, we obtain the following results from a sample function x(t, 1): Average value (sample function of length N bits with n1 ones): Z T Z N Tb 1 1 x(t, 1)dt = lim x(t, 1)dt mx = E {x(t, 1)} lim N T T N T b 0 0

1 n1 = lim [n1 1 Tb + (N n1) 0 Tb] = lim = 0.5 N N Tb N N


Autocorrelation function:

1 Rxx( ) = E {x(t, 1)x(t + , 1)} = lim T T = 0.25( (t) + ( /Tb))


Power spectral density:

Z
0

x(t, 1)x(t + , 1)dt

Sxx(f ) = F {Rxx( )} = 0.25(1 + sinc (f Tb))

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

23

Sample Function 1.5 1 0.5 0 0.5 0.75 x(t,1)

Autocorrelation Function Rxx()

0.5 0 2 4 6 8 10 t/T 12 14 16 18 20
b

0.25

Shifted Copies of Sample Function 1.5 1 0.5 0 0.5 1.5 1 0.5 0 0.5 1.5 1 0.5 0 0.5 x(t+0.3Tb,1) 0 5 4 3 2 1 0 /Tb 1 2 3 4 5

Power Spectral Density S (f)


xx

10

12

14

16

18

20

0.75

x(t+Tb,1)

0.5 0 2 4 6 8 10 12 14 16 18 20 0.25

x(t+1.5Tb,1)

10 t/Tb

12

14

16

18

20

0 4

0 f Tb

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

24

Example 3: Deterministic Energy Signal

f (t) = rect(t 0.5)


Energy autocorrelation function:
E Rf f ( )

f (t)f (t + )dt = ( )

Energy spectral density:

Sf f (f ) = F{Rf f ( )} = sinc (f )
Autocorrelation Function Power Spectral Density

1 0.8 Rxx() 0.6 0.4 0.2 0 4 S (f) 2 0 2 4


xx

1 0.8 0.6 0.4 0.2 0 4

0 f

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

25

Example 4: White Noise A random process n(t) is called white noise, if it has a constant power spectral density of N0/2 watts per Hz measured over positive frequencies. If in addition the random process has zero mean (mn = 0), the power spectral density is given by:

Snn(f ) = N0/2
Autocorrelation function:

for all f

Rnn( ) = F

{Snn(f )} =

N0 (t) 2

Since only the rst and second moment of the process are known, the probability density function cannot be uniquely determined. In the case of a Gaussian probability density function, the process is called white Gaussian noise. If the white Gaussian noise is added to the signal, we denote it as additive white Gaussian noise (AWGN).

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

26

White Gaussian Noise 3 2

1 n(t) 0

1 2

5 t

10

Rnn () (N0 /2)

Snn ( f ) N0 /2 f

Dr. Tanja Karp

5.2 Autocorrelation and Power Spectra

27

5.3

Excitation of LTI Systems with Stationary Random Processes

Excitation of an LTI system with sample function x(t, i) of a stationary random process x(t).

x(t , i )

LTI system h(t )

y(t , i )

Sample function y (t, i) of the output random process y (t):

y (t, i) = h(t) x(t, i)


5.3.1 Expected Value of the Output Random Process

8 > <Z my = E {y (t)} = E {x(t) h(t)} = E


Z Z

9 > = x(t )h( )d > ;

> :

j 2 0 1

E {x(t )} h( )d =

mx h( ) e | {z } d = mxH (0)

Dr. Tanja Karp

5.3 Excitation of LTI Systems with Stationary Random Processes

28

5.3.2

Autocorrelation Function of the Output Random Process

Ryy ( ) = E {y (t)y (t + )} = E {(x(t) h(t))(x(t + ) h(t + ))} 8 9 > > Z <Z = =E h( )x(t )d h()x(t + )d > > : ;
Z Z

=
Z Z

h( )h()E {x(t )x(t + )} dd

=
Z =

h( )h()Rxx( + )dd
Z

h( )h( + )d Rxx( )d | {z
R E ( ) hh

Rhh()Rxx( )d = Rhh( ) Rxx( )

E Rhh ( ): Energy autocorrelation function of the system impulse response


Dr. Tanja Karp 5.3 Excitation of LTI Systems with Stationary Random Processes 29

5.3.3

Power Spectral Density of the Output Random Process

Syy (f ) = F{Ryy ( )} = F{Rhh( ) Rxx( )} = F{Rhh( )} F{Rxx( )} = F{Rhh( )} Sxx(f )


with
E E E

Rhh( ) = h( ) h( ) h( ) H (f ),
E

h( ) H (f )
2

Rhh( ) H (f )H (f ) = |H (f )|
and thus
2

Syy (f ) = Sxx(f ) |H (f )|

Rxx ()

x(t , i )

LTI system h(t ) RE hh () |H ()|2

y(t , i )

Ryy () = Rxx () RE hh ()

Sxx ( f )

Syy ( f ) = Sxx ( f )|H ( f )|2

Dr. Tanja Karp

5.3 Excitation of LTI Systems with Stationary Random Processes

30

Example: Ideal Lowpass Filtering of White Noise Input random process ni(t):

Snini (f ) = 0.5 0.5N0 ( ) = Rnini ( )


Ideal lowpass lter:

HLP (f ) =

1 for |f | < fc 0 otherwise


2

Output random process no(t):

Snono (f ) = Snini (f )|HLP (f )| =


Power of output random process:
Z fc Z

0.5N0 for |f | < fc 0 otherwise

P no =

Snono (f )df =
fc

0.5N0df = N0fc


N0 /2 fc
Dr. Tanja Karp

HLP ( f )

1 S ni ni ( f )


N0 /2 f fc fc f
31

S no no ( f )

fc

5.3 Excitation of LTI Systems with Stationary Random Processes

5.3.4

Cross-Correlation between Input and Output Random Process

The autocorrelation function describes the statistical properties of two random variables X1 and X2 taken from the same random process at times t1 and t2, respectively, X1 = X (t1) and X2 = X (t2). The cross-correlation function describes the statistical properties of two random variables X1 and Y2 taken from two dierent random processes X (t) and Y (t) (here input and output of an LTI system) at times t1 and t 2, respectively, such that X1 = X (t1) and Y2 = Y (t2). It is dened as:

RXY (t1, t2) = E {X (t1)Y (t2)}


For stationary processes, it simplies to:

RXY ( ) = E {X (t) Y (t + )}
Two random processes X (t) and Y (t) are called uncorrelated if

RXY (t1, t2) = E {X (t1)Y (t2)} = E {X (t1)} E {Y (t2)} = mX (t1) mY (t2)


They are called orthogonal if

RXY (t1, t2) = 0,

for all t1, t2.

Dr. Tanja Karp

5.3 Excitation of LTI Systems with Stationary Random Processes

32

Here:

Rxy ( ) = E {x(t) y (t + )} = E {x(t)(x(t + ) h(t + ))} 8 9 > > Z < = = E x(t) h()x(t + )d > > : ;
Z

=
Z

h()E {x(t)x(t + )} d

h()Rxx( )d = h( ) Rxx( )

Example: System Identication An LTI system with unknown impulse response is excited with a white noise random process ni(t) with power spectral density Snini = N0/2. The output noise process is no(t). The cross-correlation between input and output noise process is given by: N0 N0 ( ) = h( ) Rnino ( ) = h( ) Rni,ni ( ) = h( ) 2 2

Dr. Tanja Karp

5.3 Excitation of LTI Systems with Stationary Random Processes

33

You might also like