Research On Adaptive Acoustic Feedback Cancellation Algorithm For Digital Hearing Aids

Hearing loss is one of the most common chronic health diseases.Due to the increasing serious noise pollution in daily life and the aggravation of the aging,the population of hearing loss is expected to be further increased in the future.Wearing digital hearing aids is an important means to resolve hearing loss and help improve speech intelligibility of the hearing impaired.As hearing aids become smaller and smaller,the phenomenon of acoustic feedback poses major problems for hearing aid users.The phenomenon of acoustic feedback limits the maximum stable gain of the hearing aids,which can cause severe distortion of the useful speech signal or cause howling,affecting the speech intelligibility and comfort of the hearing aid wearer.Based on the adaptive filtering method,in this paper we do some research work on acoustic feedback cancellation in digital hearing aids,the main research results are as follows:1.Firstly,the phenomenon of acoustic feedback in digital hearing aids and its key technologies are studied.The least mean square error algorithm and normalized least mean square error algorithm based on adaptive filtering principle are studied,and the convergence and steady state error performances are analyzed.Secondly,the subband filter bank theory and the performance of the normalized subband adaptive filter algorithm are studied.Compared with the full band input signal,the subband input signal has a lower correlation.Experiment shows that the subband adaptive filtering technology accelerates the convergence rate of the adaptive filter and reduces the steady-state misalignment of the algorithm.2.Based on the subband adaptive filter,an acoustic feedback cancellation model for digital hearing aids is designed,and a new variable step size normalized subband adaptive filtering algorithm is proposed.Based on the principle of cancellation of system noise,the algorithm estimates the power of the noise-free subband priori error by using the shrinkage denoising method and updates the step size.Through mathematical analysis of the proposed algorithm,it can be concluded that thealgorithm is stable and convergent.Simulation experiments show that the proposed algorithm achieves faster convergence rate and obtains lower steady-state misalignment in both the long echo path and the short echo path of hearing aids.3.Aiming at the sparseness of acoustic feedback path,a variable step size proportional normalized subband adaptive filtering algorithm based on adaptive filter state detection is proposed.Firstly,we introduce a coefficient proportional adaptive algorithm when adjusting the coefficient of the adaptive filter.Secondly,according to the variation of the normalized distance between the short-term energy average and the long-term energy average of the adaptive filter coefficients is utilized to classify the filter updating state into convergent state,the transitional state and steady state,use different step size update adaptive filter in different states.Finally,experimental results show that the proposed algorithm has faster convergence rate and lower steady-state misalignment compared with other algorithm,and the speech output signal tracking performance and acoustic feedback suppression are also better.