Acoustic Signal Enhancement in Hearing Aids and Mobile Audio Devices

Abstract

In the past few decades, adaptive filtering techniques have been effectively employed in various fields such as linear and nonlinear acoustic echo cancellation and acoustic feedback cancellation, communication channel equalisation, adaptive beamforming, and control applications. Both the adaptive algorithms and the overall filter design play an essential role in balancing the performance of the adaptive digital filter. The design of the adaptive algorithms is based on the rate of convergence, steady-state, robustness, computational complexity, etc. On the other hand, the structural design complexity in adaptive filters largely depends on the total number of coefficients, i.e., tap-length. An improper selection of these critical parameters may result in undermodelling, slow convergence, adaptation noise, and increased computational complexity. Fast converging adaptive filtering algorithms with optimum tap-length counterbalance overall complexity and steady-state performance. The adaptive filters have grabbed attention due to their applicability in various fields, such as acoustic feedback cancellation in digital hearing aids and acoustic echo cancellation in hands-free telephone and audio conferencing applications.