The Research On Feedback Cancellation Of Digital Hearing Aids

Digital hearing aids can improve the aural comprehension of the patients and play an important role in their daily life. This paper carries on a deep research on basic principles of the adaptive feedback cancellation algorithm. The main work includes:Firstly, the basic knowledge of voice, incentive model, channel model and radiation model are respectively discussed. And then, this paper studies the mechanisms of hearing impairment, investigates the clinical hearing test method, and analyses the key algorithms of digital hearing aids including feedback cancellation, acoustic source localization, speech enhancement and loudness compensation. This paper places emphasis on the adaptive feedback cancellation algorithm.Secondly, this paper proposes a new variable step-size affine projection algorithm to settle the problem that the performance of the conventional algorithms will get worse when the input is speech signal. The new algorithm establishes a new nonlinear function between step-size and estimation error. The step-size of the new algorithm is adjusted automatically according to the change of the estimation error, which leads to low misalignment and fast convergence speed. The simulation results show that the proposed algorithm achieves lower mi salignment and faster convergence speed compared to fixed step-size affine projection algorithm and conventional adaptive algorithms.Finally, this paper puts forward a new variable regularization affine projection algorithm to further optimize the performance of digital hearing aids which can also simplify calculations. Mathematical analysis of the new algorithm provides theoretical basis for its capacity in converging to the objective system. The simulation results verified that the new algorithm can achi eve a better performance. And its low computational complexity satisfies the requirement of real-time system.