Widrow-Hoff least-mean-squares algorithm

Widrow-Hoff least-mean-squares algorithm

(communications)

An algorithm that is widely used in adaptive signal processing; for time-discrete analysis with a finite-response filter, it is represented by the first-order difference equationW_{k +1}=W_k + 2μ e_k X_k , where k is a time index that takes on integral values;Wis a vector whose components are the coefficients of the filter; μ is the convergence coefficient; e_k is the residual signal or error, equal to y_k -ŷ_k , where y_k and ŷ_k are the outputs of the plant (where the unprocessed signal is generated) and the filter, respectively; andXis a vector whose components are the present value of the input and L- 1 past values of the input, where L is the number of filter coefficients.

McGraw-Hill Dictionary of Scientific & Technical Terms, 6E, Copyright © 2003 by The McGraw-Hill Companies, Inc.