Method for adapting the coefficients of a digital filter according to the Equation-Error principle

摘要

The method involves forming a scaled error and minimizing the sum of the squares of such scaled errors over a quoted range by optimal selection of the elements of a vector. The solution of the Wiener-Hopf equation is determined recursively starting from a known optimal set of coefficients at a certain time so that only one 2 by 2 matrix needs to be inverted during the recursive computation.