Reducing inter-channel coherence in stereophonic acoustic echo cancellation using partial update adaptive filters

The use of partial update adaptive filters for stereophonic acoustic echo cancellation is investigated. It is proposed to employ subsampling of the tap-input vector, that is intrinsic to partial update schemes, to decorrelate the two tap-input vectors of the stereo adaptive structure thereby improving convergence performance. The problem has been structured as a joint optimization between inter-channel coherence and the L₁ norm of the inputs corresponding to the selected filter coefficients. We investigate the trade-off between improvement in convergence due to decorrelation of the two tap-input vectors and any degradation in convergence due to subsampling in the MMax-NLMS partial update scheme. The exclusive MMax approach is proposed which efficiently approximates the joint optimization of these factors. Simulation results show an improvement in the rate of convergence over standard NLMS approaches with a performance close to that of fast-RLS in combination with a non-linear preprocessor.