Font Size: a A A

Integrated Circuit Front-end Design Of Digital Hearing Aid

Posted on:2016-03-04Degree:MasterType:Thesis
Country:ChinaCandidate:Z Q YaoFull Text:PDF
GTID:2284330467472666Subject:Microelectronics and Solid State Electronics
Abstract/Summary:PDF Full Text Request
Digital hearing aid is an important equipment to make up for hearing loss. Its fundamental function is to amplify the sound signals. In order to meet the increasing performance demands of patients, more and more digital speech signal processing algorithms have been implemented to digital hearing aid. The algorithms mainly include loudness compensation, howling suppression, noise control and sound source localization. The loudness compensation and howling suppression algorithms are the most important ones in determining the performance of hearing aid. This research focused on these two key algorithms, and deeply studied their principles. Meanwhile, a comprehensive analysis for existing advantages and disadvantages has been given. And in this paper, an algorithm scheme for hardware implementation has been designed, as well as the verification.First of all, the basic characteristics of the sound signal and the auditory characteristics of patients with hearing impairment has been researched. The key algorithms of digital hearing aid have been analyzed, and the evaluation criteria of hearing aid has been found out.Secondly, an in-depth research on the loudness compensation algorithm and feedback suppression algorithm has been made with analyzing the request and characteristics of hardware design. These algorithms structures for easier hardware implementation have been re-designed. Loudness compensation method is based on the unequal multichannel loudness compensation design for speech signal characteristics, with each band signal in accordance with the patient hearing loudness compensation curve. Howling suppression algorithm used frequency-domain block LMS adaptive filters, which processed the speech signal through a frame by frame mode. This method is more efficient than the individual sampling method.In this paper, two key algorithms were simulated in MATLAB and designed by Verilog HDL. The modules have been optimized with the idea of sacrificing area for speed, and simulated with ModelSim and MATLAB. The two key algorithms have been integrated in the FPGA Verification Platform to build a hearing aid system. The hearing aid system has been tested by special instruments and the results generally meet the expectations of design.
Keywords/Search Tags:digital hearing aid, loudness compensation, howling suppressionmultichannel, frequency domain block
PDF Full Text Request
Related items