Scribd Logo
Upload

Welcome to Scribd!

  • 35 views

    EEE 3105: Signals and Linear Systems: 2.7 Correlation

    Uploaded by

    Mozammel Hossain
    The document discusses correlation and autocorrelation of signals. It defines correlation as a measure of similarity between two signals. Autocorrelation is the correlation of a signal with itself. The autocorrelation function provides information about the signal's similarity to time-delayed versions of itself. Key properties are that autocorrelation exhibits conjugate symmetry and is maximum at zero time delay. Autocorrelation and energy spectral density are Fourier transform pairs, with autocorrelation indicating how a signal's power is distributed over time.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    EEE 3105: Signals and Linear Systems: 2.7 Correlation

    Uploaded by

    Mozammel Hossain
    35 views6 pages
    The document discusses correlation and autocorrelation of signals. It defines correlation as a measure of similarity between two signals. Autocorrelation is the correlation of a signal with itself. The autocorrelation function provides information about the signal's similarity to time-delayed versions of itself. Key properties are that autocorrelation exhibits conjugate symmetry and is maximum at zero time delay. Autocorrelation and energy spectral density are Fourier transform pairs, with autocorrelation indicating how a signal's power is distributed over time.

    Original Title

    Signals-and-Linear-Systems-for-Students_31Students

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses correlation and autocorrelation of signals. It defines correlation as a measure of similarity between two signals. Autocorrelation is the correlation of a signal with itself. The autocorrelation function provides information about the signal's similarity to time-delayed versions of itself. Key properties are that autocorrelation exhibits conjugate symmetry and is maximum at zero time delay. Autocorrelation and energy spectral density are Fourier transform pairs, with autocorrelation indicating how a signal's power is distributed over time.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    35 views6 pages

    EEE 3105: Signals and Linear Systems: 2.7 Correlation

    Uploaded by

    Mozammel Hossain
    The document discusses correlation and autocorrelation of signals. It defines correlation as a measure of similarity between two signals. Autocorrelation is the correlation of a signal with itself. The autocorrelation function provides information about the signal's similarity to time-delayed versions of itself. Key properties are that autocorrelation exhibits conjugate symmetry and is maximum at zero time delay. Autocorrelation and energy spectral density are Fourier transform pairs, with autocorrelation indicating how a signal's power is distributed over time.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 6

    11/18/2021

    EEE 3105: Signals and Linear Systems

    2.7 Correlation
    Correlation is a measure of similarity between two signals. The
    general formula for continuous-time and discrete-time
    correlations are

    R( )   x(t ) y (t  )dt (2.7.1)

    R[k ]   x[n]y[n  k ] (2.7.2)
    k  

    There are two types of correlation:


    (a) Auto correlation
    (b) Cross correlation: (i) Linear cross correlation and
    (ii) Circular cross correlation

    1
    11/18/2021

    Difference between Convolution and Correlation Function

    General formulae for Convolution:


     Convolution needs:
    x(t )   x( )h(t  )d
    • Time reverse

    y[n]   x[k ]h[n  k ] • Time shifting
    k  
    • Multiplications

    General formulae for Correlation:


     Correlation needs:
    R( )   x(t ) y (t  )dt

    • Time shifting
    R[k ]   x[n]x[n  k ] • Multiplications
    k  

    Convolution = Folding + Correlation


    3

    2.7.1 Auto Correlation Function


    It is defined as correlation of a signal with itself. Auto correlation function is a
    measure of similarity between a signal and its time delayed version. It is
    represented with R().
    Consider a signal x(t). The auto correlation function of x(t) with its time
    delayed version is given by

    Rxx ( )    x(t ) x(t  )dt [  ve shift] (2.7.3)

       x(t   ) x(t )dt [  ve shift] (2.7.4)
    where,  = searching or scanning or delay parameter.
    If the signal is complex then auto correlation function is given by

    Rxx ( )    x(t ) x  (t  )dt [ ve shift] (2.7.5) where * denotes the
     complex conjugate.
       x(t   ) x  (t ) dt [ ve shift] (2.7.6)
    Similarly the autocorrelation of the discrete time signal x[n] is expressed as

    Rxx [k ]   x[n]x[n  k ] (2.7.7)
    k  
    If the signal is complex then auto correlation function is given by

    Rxx [k ]   x[n]x  [n  k ] (2.7.8)
    k   4

    2
    11/18/2021

    Expressions of Autocorrelation for Energy and Power signals

    Properties of Auto-correlation Function of Energy (/Power) Signal


    • Auto correlation exhibits conjugate symmetry i.e. Rxx(τ) = Rxx*(- τ)

    Proof : Rxx ( )   x(t ) x  (t  )dt

     Rxx * ( )   x  (t ) x(t  )dt

     Rxx * ( )   x  (t ) x(t  )dt
     Rxx * ( )  Rxx ( )
    • Auto correlation function of energy signal at origin i.e. at τ = 0 is the total

    Rxx (0)  E   x(t ) dt
    2
    energy of that signal, which is given as:

    Proof : Rxx ( )   x(t ) x  (t  )dt

     Rxx (0)   x  (t ) x(t )dt [ for τ = 0 ]

     Rxx (0)   x(t ) dt  E x
    2

    1 T
    [ For power  Rxx (0)  lim x(t ) dt  Px ]
    2

    T  2T
    T

    • Auto correlation function  1/τ


    6

    3
    11/18/2021

    • Auto correlation function is maximum at τ = 0 i.e. |Rxx(τ)| ≤ Rxx(0)


    Proof : Consider x(t ) and x(t   ) Integratin g both sides
    [ x(t )  x(t   )]  0
    2   
      x 2 (t )dt   x 2 (t   )dt    2 x(t ) x(t   )dt
     x 2 (t )  x 2 (t   )  2 x(t ) x(t   )  0   x 2 (t )dt   x 2 (t )dt  2 R ( ) [  0 in left - hand side]
      xx
     x 2 (t )  x 2 (t   )  2 x(t ) x(t   )  
      x(t ) dt   x(t ) dt   2 Rxx ( )
    2 2

     E  E  2 Rxx ( )
     2 E  2 Rxx ( )
     E  Rxx ( )
     Rxx (0)  Rxx ( )

    • Auto correlation function and energy spectral densities are Fourier transform
    pairs. i.e. F.T[R( )]  X ( )
    X ( )   R( )e j d

    • R(τ ) = x(τ ) ∗ x(−τ )


    7

    Example: Autocorrelation of the signal x[n] = {-1, 2, 1} can be computed as


    shown in Figure below:

    Rxx = {-1, 0, 6, 0, -1} 8

    4
    11/18/2021

    Here, the first set of samples (those in the first row of every table)
    refers to the given signal. The second set (in the second row of
    every table) refers to the samples of its time-shifted version. Next,
    the samples shown in red color in the third row are obtained by
    multiplying the corresponding samples of the first two rows.
    Finally, we add the samples in the last row of the sample
    (contained within the curly brackets) so as to obtain the samples of
    the auto-correlated signal.
    Thus, here we find that the samples of the autocorrelated
    signal Rxx are {-1, 0, 6, 0, -1}, where 6 is the zeroth sample.
    The example presented shows that the sample of the
    autocorrelated signal will be at its maximum value when the
    overlapping signal best matches the given signal. In this case, it
    happens when time-shift is zero. 9

    Tabular Method of Auto Correlation: Auto-correlate the signal


    x[n] = {-1, 2, 1} using Tabular method.
    The tabular convolution method involve with time reversal, whereas no time reversal is
    needed in correlation. But if we time reverse the 2nd sequence, then it form the method
    of convolution.
    x[n] = { -1, 2, 1} and x[-n] = {1, 2, -1}

    • Total number of terms in cross correlation are {(3) + (3) -1} = 5.


    • Range = -2 to 2
    Interchanging Column and Row:

    Rxx =x[n]*x[-n]= {-1, 0, 6, 0, -1} Rxx =x[-n]*x[n]= {-1, 0, 6, 0, -1}


    10

    5
    11/18/2021

    Tabular Method of Auto Correlation: Auto-correlate the signal


    x[n] = {-1, 2, 1} using Tabular method.
    The tabular convolution method involve with time reversal, whereas no time reversal is needed
    in correlation. But if we time reverse the 2nd sequence, then it form the method of convolution.

    Location 0 1 2 • Total number of terms in cross correlation


    Value -1 2 1 are {(3) + (3) -1} = 5.
    • Sum of indexes at zero marked = 1 + 1 = 2
    Location 2 1 0
    Value
    • Range = -2 to 2
    -1 2 1

    Here, x[-n] = {1, 2, -1}


    Interchanging Column and Row:

    Rxx =x[n]*x[-n]= {-1, 0, 6, 0, -1} Rxx =x[-n]*x[n]= {-1, 0, 6, 0, -1}


    11

    Example: Find the autocorrelation function and power of the sinusoidal


    signal x(t) = A sin(2fot).

    12

    You might also like