A Fast Converging and Robust Stability of Acoustic Echo Cancellation using Adaptive Filter

Abstract

In this paper we introduce an Acoustic echo cancellation (AEC) based on an adaptive algorithms to design the adaptive filtering. The main aim in hands-free telephony and in teleconference systems is to provide a good free voice quality when two or more people communicate from different places. The problem that arises during the conversation is often due to acoustic echo. This problem will cause a bad quality of voice signal. Acoustic echo cancellation (AEC) is common in modern telecommunication systems, and it is one of the most popular applications of adaptive filters. This paper focuses on the use of filtering as LMS algorithms to reduce this unwanted echo. The results indicate that the LMS is the simplest to implement and is stable when the step size parameter is selected appropriately. For large step size, a moderate large value of leads to faster convergence but if step size is too high, it leads to instability of LMS and erroneous results. In this paper we propose the robust technique is referred to based diagonal loading (DL). The performance evaluation of the robust adaptive filter using diagonal loading provides faster speed of convergence among other variable step size algorithms while retaining the same small level of misadjustment and the mean square error is almost similar to that of the (fixed step size least mean square) DL-LMS algorithm under similar conditions. more robust variant of the LMS algorithm, exhibits a better balance between simplicity and performance than the pure LMS algorithm.