An Effective Tap-Length NLMS Algorithm for Network Echo Cancellers

Abstract

A novel normalized least mean square (NLMS)-based algorithm is proposed to locate the active taps of the unknown impulse response in the application of network echo cancellation. Starting with the full length of the unknown echo path, the start-tap and end-tap estimates of the active region adapt, respectively, in the form of fractional tap. The effective tap-length estimate is defined by the length between the start-tap and end-tap estimates. The filter tap-weight vector is then updated with this effective tap-length estimate following the NLMS algorithm, while the tap weights located beyond the effective tap-length estimate are directly set to zero. In case that the active region of the unknown echo path shifts in a large scale, the start-tap and end-tap estimates are reset to the full length when the effective tap-length estimate is lower than a predetermined threshold value. Simulation results demonstrate that the proposed fractional tap-based algorithm can efficiently locate and track the effective tap length of the unknown echo path when the active coefficients change abruptly in locations and magnitudes, outperforming the existing stochastic tap-based schemes.