P. 1
Unit 4 Random Signal Processing

Unit 4 Random Signal Processing

|Views: 32|Likes:
Published by manojniranj

More info:

Published by: manojniranj on Oct 25, 2010
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PPT, PDF, TXT or read online from Scribd
See more
See less

12/14/2012

pdf

text

original

Unit 4 Random Signal Processing

MULTIRATE SIGNAL PROCESSING-LAST PART OF UNIT 4

BASICS OF RANDOM SIGNAL PROCESSING ‡ Introduction to probability function, joint probability, conditional probability ± estimation parameters ± joint distribution function, probability density function, ensemble average ± mean squared value, variance, standard deviation, moments, correlation, covariance, orthogonality, auto-covariance, auto-correlation, crosscovariance and cross-correlation ± stationarity ± ergodic ± white noise ± energy density spectrum ± power density spectrum estimation ± periodogram ± direct method, indirect method, Bartlett method ± Welch method. Decimator (down sampling) ± frequencydomain analysis of decimator ± interpolation (up sampling) ± frequency-domain analysis of interpolator

‡ HOD WILL HANDLE: ‡ power density spectrum estimation ± periodogram ± direct method, indirect method, Bartlett method ± Welch method ‡ ProF.J.Valarmathi:VIT ‡ Introduction to probability function, joint probability, conditional probability ± estimation parameters ± joint distribution function, probability density function, ensemble average ± mean squared value, variance, standard deviation, moments, correlation, covariance, orthogonality, auto-covariance, auto-correlation, crosscovariance and cross-correlation ± stationarity ± ergodic ± white noise ± energy density spectrum

‡ When the sampling rate of the signals is unequal at various parts of the systemThose discrete time systems that process data at more than one sampling rate are known as multirate systems.Definition of Multirate Signal Processing ‡ The systems that uses single sampling rate from A/D converter to D/A converter are known as single rate systems. .

For example a CD is sampled at 44.Narrow band filtering for fetal ECG and EEG .Where Multirate signal processing is used: ‡ 1. ‡ 2.In audio signal processing. ‡ 3.In high quality data acquisition and storage systems. Conversion between DAT and CD use multirate signal processing technique.1KHz but DAT is sampled at 48 KHz.

‡ 4. . ‡ 6.In transmultiplexers. Therefore to watch an American program in Europe one need a sampling rate converter.In speech processing to reduce the storage space or the transmitting rate of the speech data.In video PAL and NTSC run at different sampling rates. ‡ 5.

Basic operations: Interpolation and decimation.Used to decrease the Downsampling rate by an integer factor.Used to increase the Upsampling rate by an integer factor ‡ Down-sampler . . Basic Sampling Rate Alteration Devices ‡ Up-sampler .

otherwise .. s L.UpUp-Sampler ‡ Up-sampling operation is implemented by inserting L  1 equidistant zero-valued samples between two consecutive samples of x[n] ‡ Input-output relation x ®[n / L]. xu [n] ! ¯ ° 0. n ! 0. s 2 L.

5 Amplitude 0.5 -1 0 -1 0 10 20 30 Time index n 40 50 .UpUp-Sampler Input Sequence 1 Output sequence up-sampled by 3 1 0.5 Amplitude 10 20 30 Time index n 40 50 0 0 -0.5 -0.

the zero-valued samples inserted by the up-sampler are replaced with appropriate nonzero values using some type of filtering process ‡ Process is called interpolation and will be discussed later .UpUp-Sampler ‡ In practice.

where M is a positive integer.DownDown-Sampler TimeTime-Domain Characterization ‡ An down-sampler with a down-sampling downfactor M. develops an output sequence y[n] with a sampling rate that is (1/M)-th of that of the input sequence x[n] ‡ Block-diagram representation x[n] M y[n] .

DownDown-Sampler ‡ Down-sampling operation is implemented by keeping every M-th sample of x[n] and removing M  1 in-between samples to generate y[n] ‡ Input-output relation y[n] = x[nM] .

5 0.5 -0.5 -1 -1 0 10 20 30 Time index n 40 50 .5 Amplitude 0 10 20 30 Time index n 40 50 Amplitude 0 0 -0.DownDown-Sampler Input Sequence 1 Output sequence do n-sampled by 3 1 0.

Basic Sampling Rate Alteration Devices ‡ Sampling periods have not been explicitly shown in the block-diagram representations of the up-sampler and the down-sampler ‡ This is for simplicity and the fact that the mathematical theory of multirate systems can be understood without bringing the sampling period T or the sampling frequency FT into the picture .

DownDown-Sampler ‡ Figure below shows explicitly the timedimensions for the down-sampler x[ n ] ! xa ( nT ) Input sampling frequency M y[ n ] ! xa ( nMT ) Output sampling frequency ' ! FT ! 1 FT M T' 1 FT ! T .

n !0. L .UpUp-Sampler ‡ Figure below shows explicitly the timedimensions for the up-sampler x[ n ] ! xa ( nT ) L y[n] x ( nT / L ).!¯ a other ise ° 0 Input sampling frequency Output sampling frequency ' ! LF ! 1 FT T T' 1 FT ! T . 2 L .

Basic Sampling Rate Alteration Devices ‡ The up-sampler and the down-sampler updownare linear but time-varying discrete-time timediscretesystems ‡ We illustrate the time-varying property of a down-sampler ‡ The time-varying property of an upsampler can be proved in a similar manner .

Basic Sampling Rate Alteration Devices ‡ Consider a factor-of-M down-sampler defined by y[n] = x[nM] ‡ Its output y1[ n] for an input x1[ n] ! x[n  n0 ] is then given by y1[ n] ! x1[ Mn] ! x[ Mn  n0 ] ‡ From the input-output relation of the down-sampler we obtain y[n  n0 ] ! x[ M (n  n0 )] ! x[ Mn  Mn0 ] { y1[ n] .

DownDown-Sampler FrequencyFrequency-Domain Characterization ‡ Applying the z-transform to the input-output relation of a factor-of-M down-sampler we get y[n] ! x[ Mn] Y ( z) ! § x[Mn] z n ! g g n ‡ The expression on the right-hand side cannot be directly expressed in terms of X(z) .

xint [n] ! ¯ otherwise ° 0. ‡ Then Y ( z) ! ! n ! g g § x[Mn] z ! § xint [Mn] z n ! g g n g n k ! g xint [k ] z k / M ! X int ( z1 / M ) § . n ! 0. define a new sequence xint [n] : x ®[n].DownDown-Sampler ‡ To get around this problem. s M . s 2 M .

n ! 0. ‡ A convenient representation of c[n] is given by 1 M 1 kn c[n] ! § WM M k !0  j 2T / M where WM ! e where . M . xint [n] can be formally related to x[n] through xint [n] ! c[n] ™ x[n] 1.DownDown-Sampler ‡ Now.c[n] ! ¯ 0 other ise °. 2 M .

DownDown-Sampler ‡ Taking the z-transform of xint [n] ! c[n] ™ x[n] and making use of 1 c[n] ! M M 1 § k !0 kn WM ¨ M 1 kn ¸ n © § WM ¹ x[n] z  n § © int ( z ) ! § c[n]x[n] z ¹ º n ! g n ! g ª k !0 ¸ 1 M 1 1 M 1¨ g  kn © § x[n]WM z  n ¹ ! ! X z WMk §© ¹ M § M k !0 ª n ! g º k !0 1 ! M we arrive at g g .

.

DownDown-Sampler ‡ Consider a factor-of-2 down-sampler with an input x[n] whose spectrum is as shown below ‡ The DTFTs of the output and the input sequences of this down-sampler are then related as j[ ) ! 1 { X (e j[ / 2 )  X ( e j[ / 2 )} Y (e 2 .

DownDown-Sampler ‡ Now ( e j[ / 2 ) ! (e j ([ 2 T) / 2 ) implying that the second term (e j[ / 2 ) in the previous equation is simply obtained by shifting the first term (e j[ / 2 ) to the right by an amount 2T as shown below .

the original ³shape´ of (e j[ ) is lost when x[n] is downsampled as indicated below .DownDown-Sampler ‡ The plots of the two terms have an overlap. in general. and hence.

e.DownDown-Sampler ‡ This overlap causes the aliasing that takes place due to under-sampling ‡ There is no overlap. i. only if ( e j[ ) ! 0 for [ u T / 2 Y (e j[ ) is indeed periodic with a ‡ Note: period 2T. no aliasing. even though the stretched version of (e j[ ) is periodic with a period 4T ..

the relation between the DTFTs of the output and the input of a factor-of-M down-sampler is given by 1 M 1 Y (e j[ ) ! X (e j ([2 Tk ) / M ) § M k !0 Y (e j[ ) is a sum of M uniformly ‡ shifted and stretched versions of (e j[ ) and scaled by a factor of 1/M .DownDown-Sampler ‡ For the general case.

DownDown-Sampler ‡ Aliasing is absent if and only if (e j[ ) ! 0 for [ u T / M as shown below for M = 2 (e j[ ) ! 0 for [ u T / 2 .

prior to downsampling. avoid aliasing caused by downy[n ] sampling x[n ] H ( ) M ‡ The above system is called a decimator . the signal v[n] should be bandlimited to [ T / M by means of a lowpass filter. called the decimation filter as indicated below to filter.Filter Specifications ‡ On the other hand.

s 4. . n ! 0.x u [n ] ! ¯ otherwise ° 0. s 2.UpUp-Sampler FrequencyFrequency-Domain Characterization ‡ Consider first a factor-of-2 up-sampler whose input-output relation in the timedomain is given by x ®[n / 2].

the inputoutput relation is then given by X u ( z) ! n ! g g § xu [n] z ! § x[n / 2] z n ! g n even g n g n ! § x[m] m !g z 2 m ! X (z ) 2 .UpUp-Sampler ‡ In terms of the z-transform.

we can show that for a factor-of-L up-sampler factor-of.upL X u ( z) ! X ( z ) ‡ On the unit circle. the inputoutput relation is given by X u (e j[ ) ! X (e j [L ) . for z ! e j[ .UpUp-Sampler ‡ In a similar manner.

UpUp-Sampler ‡ Figure below shows the relation between j[ j[ X (e ) and u (e ) for L = 2 in the case of a typical sequence x[n] .

UpUp-Sampler ‡ As can be seen. a factor-of-2 sampling rate expansion leads to a compression of X (e j[ ) by a factor of 2 and a 2-fold repetition in the baseband [0. 2T] ‡ This process is called imaging as we get an additional ³image´ of the input spectrum .

UpUp-Sampler ‡ Similarly in the case of a factor-of-L sampling rate expansion.L  1 valued samples in xu [n] with interpolated sample values . there will be L  1 additional images of the input spectrum in the baseband ‡ Low pass filtering of xu [n] removes the images and in effect ³fills in´ the zero.

the unwanted images in the spectra of the up-sampled signal xu [n] must be removed by using a lowpass filter H(z). indicated below x[n ] L xu [n] y[n ] ‡ The above system is called an interpolator .Filter Specifications ‡ Since up-sampling causes periodic repetition of the basic spectrum. called the interpolation filter as filter.

Cascade Equivalences ‡ A complex multirate system is formed by an interconnection of the up-sampler. and the components of an LTI digital filter ‡ In many applications these devices appear in a cascade form ‡ An interchange of the positions of the branches in a cascade often can lead to a computationally efficient realization . the down-sampler.

if we pass x[n] through a factor-of-L up-sampler xu [n] generating xu [n] .Interpolation Filter Specifications ‡ On the other hand. the output of the filter will be precisely y[n] . the relation between the Fourier transforms of x[n] and are given by X u (e j[ ) ! X (e j[ L) ‡ It therefore follows that if xu [n] is passed through an ideal lowpass filter H(z) with a cutoff at T/L and a gain of L.

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->