Please use this identifier to cite or link to this item: http://nopr.niscair.res.in/handle/123456789/55120
Title: Multiple Bandwidth FIR Filter Design with Adaptive Algorithms for Hearing Aid Systems
Authors: Mariammal, K
SherinMary, R
Brittopari, J
Vaithiyanathan, D
Keywords: FIR filter;Bandwidth;Windowing;Hearing losses modelling;LMS;NLMS;RLS
Issue Date: Aug-2020
Publisher: NISCAIR-CSIR, India
Abstract: Digital Filter design plays a vital role in signal processing and communication applications. This paper proposes a hearing loss system model with Variable Bandwidth FIR Filter (VBF) and adaptive algorithms for the application to listening. The tunable band filter is designed to provide an appropriate sound level. This filter has several sub-filters each of which is designed with set of selected bandwidths. The sub-bands obtained are adjusted with proper magnitude by trial and error method. Algorithms such as Least Mean Square (LMS), Normalized Least Mean Square (NLMS) and Recursive Least Squares (RLS) are incorporated to improve the quality of the signal. The filter thus designed is examined by taking a number of audio signals. The tests on various hearing loss cases with different type of input signal suggest that this method is capable of reproducing a signal which sounds exactly the same as the original signal. The multiple bandwidth filters is analyzed with mild, moderate, profound and severe hearing loss patterns and the results are reported. The matching error is calculated between ideal response and actual response. The result show that the designed filter provides acceptable minimum matching error and it lies in the range 0 to 2.5dB.This filter design is implemented in TMS320C6711 processor and is tested for sinusoidal input signal.
Page(s): 605-623
URI: http://nopr.niscair.res.in/handle/123456789/55120
ISSN: 0975-0959 (Online); 0301-1208 (Print)
Appears in Collections:IJPAP Vol.58(08) [August 2020]

Files in This Item:
File Description SizeFormat 
IJPAP 58(8) 605-623.pdf1.82 MBAdobe PDFView/Open


Items in NOPR are protected by copyright, with all rights reserved, unless otherwise indicated.