Design of Flexible Signal Processing Algorithm for Reducing Feedback Signal on Digital Hearing Aid System

Abstract

This paper presents adaptive filtering algorithm for processing noise feedback signal in digital hearing aid system. Conventional hearing aid system suffers from noise feedback signal which degrades the original speech signal and cause signal instability. This can be reduced by utilizing adaptive filtering techniques in the filter bank. However this method is very unstable since their weights are modified frequently. Hence a normalized adaptive filtering technique is utilized in the proposed method. Initial study on signal level on hearing output is carried and proposed algorithm is compared with adaptive beam former which works well in noisy environment. Tradeoff between segmented signal to noise ratio and signal quality is considered. The algorithm achieves a better signal gain quality of 16dB over the limit gain.  The algorithm utilizes adaptive feedback noise reduction filter bank having 128 filter coefficients.