Performance Analysis of Adaptive Filters for Noise Cancellation in Audio Signal for Hearing Aid Application

G K Girisha, S L Pinjare

Abstract: This paper represents the experimental development of software and hardware configuration to implement adaptive algorithms for noise cancellation in audio signal processing for hearing aid applications. Adaptive FIR filters play an essential role in many DSP systems including noise cancellers, signal prediction, adaptive feedback cancellation and echo cancellation. In this paper, we describe adaptive filter which is used for the cancellation of the noise component which is overlapped with the desired signal in the same frequency range. The study includes comparison between three different types of adaptive filters, the LMS algorithm, NLMS algorithm and Recursive Least Squares Filter (RLS) algorithm Mean Squared Error (MSE), Peak Signal to Noise Ratio (PSNR) and Mean Absolute Error (MAE) are calculated for all the three adaptive algorithms. The efficient Adaptive Noise Canceller is realized using MATLAB / Simulink.

Keywords: Adaptive filter algorithms, MSE, PSNR, MAE, microphone arrays