02.10.2019 Views

UploadFile_6417

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Round-off Effects in IIR Digital Filters 561<br />

where e(n) isamultiplication quantization error. For analysis purposes<br />

we assume that the conditions on e(n) are similar to those for the A/D<br />

quantization error:<br />

1. The random signal e(n) isuncorrelated with the sequence x(n) for<br />

rounding operation (or two’s-complement truncation operation) in the<br />

quantizer.<br />

2. The signal e(n) isanindependent process (i.e., the samples are independent<br />

of each other).<br />

3. The probability density function (pdf) f E (e) ofe(n) for each n is uniformly<br />

distributed over the interval of width ∆ = 2 −B , which is the<br />

quantizer resolution.<br />

We will emphasize the rounding operation for the rest of this section.<br />

Based on the above model assumptions, the results given in (10.7), (10.9),<br />

and (10.10) are also applicable for the multiplication quantization error<br />

e(n).<br />

We offer the following two MATLAB examples to illustrate this<br />

model. A more thorough investigation of this error can be found in<br />

Rabiner and Tukey [25].<br />

□ EXAMPLE 10.9 Consider the sequence given in Example 10.1, which is repeated here.<br />

x(n) = 1 [sin(n/11) + sin(n/31) + cos(n/67)]<br />

3<br />

This signal is multiplied by c =1/ √ 2, quantized to B bits and the resulting<br />

multiplication is quantized to B bits with rounding. Using the StatModelR<br />

function and 500,000 samples, compute and analyze normalized errors e 1(n)<br />

and e 2(n), defined in (10.2) and (10.3), respectively.<br />

Solution<br />

The following MATLAB script computes error distribution, for B =6bits.<br />

clear; close all;<br />

% Example parameters<br />

B = 6; N = 500000; n = [1:N]; bM = 7;<br />

xn = (1/3)*(sin(n/11)+sin(n/31)+cos(n/67)); clear n;<br />

c = 1/sqrt(2);<br />

% Signal and Coefficient Quantization<br />

xq = (round(xn*(2^B)))/(2^B); c = (round(c*(2^B)))/(2^B);<br />

cxq = c*xq;<br />

% Multiplication of constant and signal<br />

% Quantization error analysis<br />

[H1,H2,Q, estat] = StatModelR(cxq,B,N);<br />

H1max = max(H1); H1min = min(H1);<br />

% Max and Min of H1<br />

H2max = max(H2); H2min = min(H2);<br />

% Max and Min of H2<br />

Copyright 2010 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).<br />

Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!