Comparative research of various adaptive algorithms for noise cancellation in speech signals

Abstract

In real life situations, the statistical characteristics of signal and noise are generally unknown & hence a digital filter having ‘constant coefficients’ is hardly of any use. In such situations adaptive filter is desirable. Adaptive filters are capable of adapting their filter coefficients as per the abnormality in characteristics of input signal and noise to achieve a noise free signal. This paper discusses the comparative analysis of various adaptive filter algorithms such as LMS (Least mean square), BLMS (Block LMS), NLMS (Normalized LMS), BNLMS (Block NLMS), VSLMS (Variable step size LMS) and BVSLMS (Block VSLMS) algorithms. As a input we have used Hindi audio speech signal and Babble noise as an interference signal.