DOA estimation and beamforming using spatially under-sampled AVS arrays

Abstract

In this paper, we show the advantages of spatially under-sampled acoustic vector sensor (AVS) arrays over conventional acoustic pressure sensor (APS) arrays for performing direction-of-arrival (DOA) estimation and interference cancellation. We provide insights into the theoretical performance of an under-sampled AVS array with respect to its DOA estimation performance using the Cramér-Rao lower bound (CRLB). We also show that the minimum variance distortionless response (MVDR) beamformer suppresses the grating lobes considerably as compared to the classical (or Bartlett) beamformer leading to unambiguous DOA estimates. Finally, through zero-forcing (ZF) and minimization of maximum side lobe beamformers, the advantages of an under-sampled AVS array for interference cancellation are presented.