Un Algorithme LMF Efficace à Pas Variable pour la Réduction du Bruit Acoustique

Abstract

As part of this graduation thesis, we are interested in a major subject which is the acoustic noise reduction and its problems, we proposed a new adaptive algorithm LMF (Least mean fourth) and its normalized version NLMF. The advantage of this algorithm is it's efficiency and his features such as its noise cancelation and high convergence speed compared to other adaptive algorithms. In the last part we did a comparison between both of NLMF and NLMS algorithms and the simulation results confirmed the effectiveness of the proposed algorithm over the past ones.