Comparative Study of Weiner-Hopf and LMS algorithm for Adaptive Noise Cancellation of Speech

Download

PDF Restricted to Registered users only Download (751Kb)

Description/Abstract

The performance of Wiener-Hopf and Least Mean Square (Online and Batch) methods are considered for recovering the desired signal buried under noise using adaptive noise cancellation techniques. A reference signal and a desired signal are available. To combat the noisy environment an adaptive noise cancellation filter is developed. This filter tries to converge in least mean square sense to the optimal Wiener-Hopf solution to filter the noise from the signal. In the analysis part the weight and learning tracks are shown. Effects of leakage and effects of interchanging the primary and reference inputs to the adaptive filter are also discussed. In the end a comparison is drawn on intelligibility of the recovered signal from both the methods.