An optimized adaptive filtering for speech noise cancellation

الملخص EN

The main interest in adaptive filters continues to grow as they begin to find practical applications in areas such as channel equalization, echo cancellation, noise cancellation and many other adaptive signal-processing applications.

The work presented in this paper focuses on optimizing most popular adaptive filtering algorithms namely Least Mean Square (LMS) algorithm, Normalized Least Mean Square (NLMS) and Rec ursive least Squares (RLS) by using genetic optimizer approach.

The tap-length are updated with the three adaptive algorithms according to the value of mean square error based on genetic style.

The simulation results for noise cancellation in speech enhancement demonstrate the good performance of the proposed algorithm in attenuating the noise with less hardware resources complexity.

It is a nice tradeoff between hardware complexity, SNR ratio and the convergence speed.