Utilization of a correlation RLS algorithm for nonlinear echo cancellation

Abstract

In hands-free systems, teleconference systems, and hybrid telephone networks, an acoustic feedback signal, existing in the error signal under a double-talk situation, may cause some fluctuation in the parameter estimates of an echo path model. Also, digital to analog converter (DAC) and analog to digital converter (ADC) may yield nonlinear distortion in the hybrid system. In this paper, utilization of adaptive Volterra filtering with a correlation RLS (CRLS) algorithm is proposed to compensate for nonlinear distortion in the echo path and to solve the double-talk problem. Simulation results, showing fast convergence and robustness, are provided to illustrate the validity and performance of the proposed nonlinear echo cancellation approach.