Acoustic Feedback Reduction Based on LMS and Normalized LMS Algorithms in WOLA Filters Bank Based Digital Hearing Aids
Acoustic feedback phenomenon can disturb a digital hearing aid performance at high gains, causing instability in the haring aid and degradation in the speech. In order to restore a stable situation, it is needed an acoustic feedback reduction (AFR) subsystem using adaptive algorithms such as the least-mean square (LMS) algorithm. This algorithm has a reduced computational cost, but it is very unstable. In order to avoid this situation, it is used another feedback reduction system based on a modified version of the LMS algorithm. Such algorithm is: the Normalized LMS (NLMS). These two algorithms are tested in two digital hearing aid categories: the In-The-Ear and the In-The-Canal. These categories are selected because they have great feedback effects, so robust AFR subsystems are needed. The added stable gain (ASG) over the limit gain when AFR subsystem is working in the digital hearing aid is obtained for each category. The ASG is determined as a trade-off between two measurements: the Segmented Signal-to-Noise Ratio (objective measurement) and the speech quality (subjective measurement). The results show how the digital hearing aids working with a feedback reduction adaptive filter adapted with the NLMS algorithm is able to achieve up to 18 dB of increase over the limit gain.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is temporarily free for AES members.