An adaptive algorithm using FIR filter for speech enhancement systems

An adaptive algorithm for speech enhancement is studied in this paper. We propose the adaptive Lyapunov speech enhancement algorithm for nonlinear speech enhancement system using FIR filter. By using the Lyapunov stability theory, the filter coefficients can adaptively adjust so that the error converges to zero asymptotically. The adaptive algorithm ensures the speech enhancement system has a better performance than the traditional least mean square algorithm. For the case that the size, weight and power consumption, the design of a microphone array for hearing aids are serious hard to establish, the algorithm still work. At last, an example of speech enhancement system result demonstrated the feasibility and effectiveness of the proposed adaptive algorithm.