Adaptive filtering using Higher Order Statistics (HOS)

The performed job, in this study, consists in studying adaptive filters and higher order statistics (HOS) to ameliorate their performances, by extension of linear case to non linear filters via Volterra series. This study is, principally, axed on:  Choice of the adaptation step and convergence conditions.  Convergence rate.  Adaptive variation of the convergence factor, according to the input signal. The obtained results, with real signals, have shown computationally efficient and numerically stable algorithms for adaptive nonlinear filtering while keeping relatively simple computational complexity.