Аннотация
The presence of acoustic feedback in digital hearing aids is almost an unavoidable phenomena, as the input microphone and loudspeaker are located very close to each other. The presence of strong acoustic feedback may result in howling effect, which degrades the performance of hearing aid. This paper presents an efficient solution for mitigating acoustic feedback in modern hearing aids equipped with a very short processing delays. The developed method employs two adaptive filters (AFs) and a delay-based strategy to monitor their convergence. The hearing aid received (amplified) signal acts as the input to the main adaptive filter (AF) which is adapted using lattice filtering-based algorithm with improved decorrelation properties. An internally generated random probe signal is input to the auxiliary AF which is adapted using a hybrid adaptive algorithm combining feature of normalized least mean square (NLMS) and maximum Versoria-criterion (MVC) cost functions. The computer simulation demonstrate that the proposed is quite efficient in mitigating the effect of the acoustic feedback in comparison with the other counter parts.