PERFORMANCE OF REAL TIME VOICE IMPROVEMENT USING LMS, NLMS AND UNANR FILTERS

Abstract

Speech intelligibility can be enhanced using acoustic properties of “clear speech”, the speech produced by a speaker with an intention to improve intelligibility in a difficult communication environment. The research objective is to devise a signal processing technique based on the properties of clear speech for improving perception of stop consonants for use in speech communication devices and hearing aids. This study was made to estimate the frequency of measuring voltage or current signal in presence of random noise and distortion. Here we are first using linear techniques such as least mean square (LMS), algorithm for measuring the frequency from the distorted voltage signal. Then comparing these results with nonlinear techniques such as nonlinear least mean square (NLMS), and UNANR algorithms with different modulation techniques was Amplitude Modulation and Frequency Modulation and communication channels i.e. AWGN Channel. Signal performance parameter PSNR measured and compared with respect to Signal to Noise Ratio. The performances of these algorithms are studied through MATLAB R2012a simulation.