ON THE PRACTICAL IMPLEMENTATION OF MULTICHANNEL ADAPTIVE FILTERS BASED ON LMS, RLS, FTF AND FAP ALGORITHMS FOR ACTIVE CONTROL

摘要

In this work a comparative study of the adaptive algorithms: LMS (Least Mean Squares), RLS (Recursive Least Squares), FTF (Fast Transversal Filters) and FAP (Finite Affine Projection), for multichannel active control is presented. This study is based on simulations using real data and is focused on: their computational aspects (computational cost and memory load), their convergence properties, their stability and their ability to create quiet zones around listener ears. Simulations show that algorithms based on Fast Transversal Filters exhibit a good trade-off between computational cost and convergence speed. On the other hand, those based on Recursive Least Squares are slightly faster but their present higher computational load and stability problems in their practical implementation. Simulations using different initialization parameters for the RLS showed its stability problems. Finally, algorithms based on low order Finite Affine Projections present less computational cost than the FTF based ones but slower convergence speed. By increasing the FAP order, their computational cost increases as well as their convergence speed. Therefore these algorithms show a desirable versatility for practical applications.