Filter Design Toolbox

nlm

Use the noise loading method to estimate the frequency response of a quantized filter

Syntax

• [h,w,pnn,nf] = nlm(hq,n,l)
  [h,w,pnn,nf] = nlm(hq,n,l,'whole')
  [h,f,...] = nlm(hq,n,l,fs)
  [h,f,...] = nlm(hq,n,l,'whole',fs)
  nlm(hq,...)
  

Description

[h,w,pnn,nf] = nlm(hq,n,l) uses the noise loading method to estimate the complex frequency response of quantized filter hq. Using nlm returns the complex frequency response h, frequency vector w, in radians/sample, power spectral density pnn, and noise figure nf, for the quantized filter hq, at n equally-spaced points around the upper half of the unit circle. Noise figure nf and power spectral density pnn are given in dB. nlm averages over l Monte Carlo trials. The Monte Carlo trials result in a noise-like signal that contains complete frequency content across the spectrum. When you omit n or l from the command, or leave them empty, n defaults to 512 and l defaults to 10.

[h,w,pnn,nf] = nlm(hq,n,l,'whole') uses n points around the entire unit circle, rather than the upper half.

[h,f,...] = nlm(hq,n,l,fs) and [h,f,...] = nlm(hq,n,l,'whole',fs) returns frequency vector f, in Hz, where fs is the sampling frequency in Hz.

nlm(hq,...) without output arguments plots the magnitude and unwrapped phase of hq, comparing the estimated response to the theoretical frequency response calculated by [h,w] = freqz(hq,n) in the current figure window.

Examples

Use the noise loading method to determine the frequency response of a quantized IIR filter Hq.

• [b,a] = butter(6, 0.5);
  hq = qfilt('df2t',{b,a});
  nlm(hq,1024,20)
  

For comparison, the plot shows the theoretical response and the response estimated by nlm. Additionally, you see the estimated phase response for comparison.

See Also
freqz, qfilt

References

McClellan, et al., Computer-Based Exercises for Signal Processing Using MATLAB 5, Prentice-Hall, 1998, 243.

msesim

noperations