Stereo acoustic echo cancellation using independent component analysis

This paper proposes a new adaptive algorithm for stereo acoustic echo cancellation. The proposed method are based on the independent component analysis (ICA) technique. By minimizing the mutual information between the estimated near- end signal and the transmitted signals, we separate the near- end speech from the acoustic echoes at the output side of the echo canceller. Simulation results show that after a same preprocessing stage the proposed method leads to a faster decrease in misalignment than the normalized least-mean-square (NLMS) algorithm while computational complexities are almost the same. It is also shown that the proposed algorithm is robust to the double-talk situation.