Enhanced Muting Method in Packet Loss Concealment of ITU-T G.722 Using Sigmoid Function with On-line Optimized Parameters

Abstract

In this paper, we propose an enhanced adaptive muting method using a sigmoid function, which is based on a parameter tracking technique for the packet loss concealment algorithm of ITUT G.722 speech codec. The packet loss concealment algorithm performs an adaptive muting to prevent the generation of unnecessary noises or clicks during packet loss recovery. While a conventional muting method applies the sigmoid function to the muting curve and the principal parameters of the sigmoid function are obtained by using a grid search-based training method, in the proposed muting algorithm, the parameters are substantially obtained from the previous good frames using the steepest descent algorithm, which minimizes the error between the desired signal and the reconstructed signal. From experimental results, the proposed adaptive muting method turns out to improve the performance of the conventional muting method under various experimental conditions.