Adaptive Beamforming for Audio Signal Acquisition

Abstract

This chapter provides an overview of adaptive beamforming techniques for speech and audio signal acquisition. We review basic concepts of optimum adap-tive antenna arrays and show how these methods may be applied to meet the re-quirements of audio signal processing. In particular, we derive optimum beamform-ers using time-domain least-squares instead of frequency-domain minimum mean-squares criteria, and, thereby, are not constrained by the commonly used narrow-band and stationarity assumption. We thus obtain a more general representation of various beamforming aspects relevant to our application. From this, a robust generalized sidelobe canceller (GSC) [1] results as an attractive solution for prac-tical audio acquisition systems. Moreover, the general theoretical framework leads to new insights for the GSC behavior in complex practical situations.