Filter Design Toolbox

adaptfilt.blmsfft

Construct an FFT-based block LMS FIR adaptive filter

Syntax

• ha = adaptfilt.blmsfft(l,step,leakage,blocklen,coeffs,states)
  

Description

ha = adaptfilt.blmsfft(l,step,leakage,blocklen,coeffs,states) constructs an FIR block LMS adaptive filter object ha where l is the adaptive filter length (the number of coefficients or taps) and must be a positive integer. l defaults to 10. step is the block LMS step size. It must be a nonnegative scalar. The function maxstep may be helpful to determine a reasonable range of step size values for the signals you are processing. step defaults to 0.

leakage is the block LMS leakage factor. It must also be a scalar between 0 and 1. When leakage is less than one, the adaptfilt.blmsfft implements a leaky block LMS algorithm. leakage defaults to 1 (no leakage). blocklen is the block length used. It must be a positive integer such that

•  blocklen + length(coeffs)
  

is a power of two; otherwise, an adaptfilt.blms algorithm is used for adapting. Larger block lengths result in faster execution times, with poor adaptation characteristics as the cost of the speed gained. blocklen defaults to l. Enter your initial filter coefficients in coeffs, a vector of length l. When omitted, coeffs defaults to a length l vector of all zeros. states contains a vector of initial filter states; it must be a length l vector. states defaults to a length l vector of all zeros when you omit the states argument in the calling syntax.

adaptfilt.blmsfft Object Properties

In the syntax for creating the adaptfilt object, the input options are properties of the object you create. This table list all the properties for the block LMS object, their default values, and a brief description of the property.

Property	Default Value	Description
Algorithm	None	Defines the adaptive filter algorithm the object uses during adaptation
FilterLength	Any positive integer	Reports the length of the filter, the number of coefficients or taps
NumSamplesProcessed	Any positive integer	Specifies the number performed during the adaptation process
Coefficients	Vector of elements	Vector containing the initial filter coefficients. It must be a length l vector where l is the number of filter coefficients. coefficients defaults to length l vector of zeros when you do not provide the argument for input.
States	Vector of elementsof length l	Vector of the adaptive filter states. states defaults to a vector of zeros which has length equal to l
Leakage	1	Specifies the leakage parameter. Allows you to implement a leaky algorithm. Including a leakage factor can improve the results of the algorithm by forcing the algorithm to continue to adapt even after it reaches a minimum value. Ranges between 0 and 1.
BlockLength	Vector of length l	Size of the blocks of data processed in each interation
StepSize	0.1	Sets the block LMS algorithm step size used for each iteration of the adapting algorithm. Determines both how quickly and how closely the adaptive filter converges to the filter solution. Use maxstep to determine the maximum usable step size.
ResetBeforeFiltering	off or on	Determine whether the filter states get restored to their starting values for each filtering operation. The starting values are the values in place when you create the filter. ResetBeforeFiltering returns to zero any state that the filter changes during processing. States that the filter does not change are not affected. Defaults to 'on'.

Example

Identify an unknown FIR filter with 32 coefficients using 512 iterations of the adapting algorithm.

• x  = randn(1,512);       % Input to the filter
  b  = fir1(31,0.5);       % FIR system to be identified
  no  = 0.1*randn(1,512);  % Observation noise signal
  d  = filter(b,1,x)+no;   % Desired signal
  mu = 0.008;              % Step size
  n  = 16;                 % Block length
  ha = adaptfilt.blmsfft(32,mu,1,n);
  [y,e] = filter(ha,x,d);
  subplot(2,1,1); plot(1:500,[d(1:500);y(1:500);e(1:500)]);
  title('System Identification of an FIR Filter');
  legend('Desired','Output','Error');
  xlabel('Time Index'); ylabel('Signal Value');
  subplot(2,1,2); stem([b.',ha.coefficients.']);
  legend('actual','estimated'); 
  xlabel('Coefficient #'); ylabel('Coefficient Value'); grid on;
  

As a result of running the adaptation process, filter object ha now matches the unknown system FIR filter b.

See Also

adaptfilt.blms, adaptfilt.fdaf, adaptfilt.lms, filter

Reference

J.J. Shynk, "Frequency-Domain and Multirate Adaptive Filtering," IEEE Signal Processing Magazine, vol. 9, no. 1, pp. 14-37, Jan. 1992.

adaptfilt.blms

adaptfilt.dlms