Direction of arrival estimation for spherical microphone arrays by combination of independent component analysis and sparse recovery

Abstract

Spherical microphone arrays provide a powerful tool for examining source localization and direction of arrival (DOA) estimation in the spherical harmonic domain. In previous work, we have investigated applying instantaneous independent component analysis (ICA) or sparse recovery separately in the spherical harmonic domain for DOA estimation. These algorithms work reasonably well, but rely on different signal characteristics: namely statistical independence or the spatial distribution of sources. In this paper, we describe methods to combine the ICA and sparse recovery algorithms to improve DOA estimation. The simulation results indicate that combining ICA and sparse recovery leads to more robust DOA estimation.