Hybrid pre-processor based on frequency shifting for stereophonic acoustic echo cancellation

In a multi-channel hands-free communication, reasonable complexity decorrelation algorithms are necessary, along with adaptive filters, to efficiently remove the acoustic echo in real time applications. But, at the same time, the quality of the signals and the spatial position of the sound source must not be perceptually affected by those algorithms. This paper proposes a new hybrid solution that uses frequency shifts to improve the performance of a state-of-art solution based on addition of half-wave rectified signals. The results show that, in a stereophonic echo cancellation environment, the new method achieves a significant improvement in the identification process of the real echo paths as well as in the global perceptual quality of the processed signals.