Direction of arrival estimation based on spherical harmonics decomposition

Abstract

Direction of arrival (DOA) estimation is of great importance to signal processing, and, as a consequence, various arrays have recently been studied to get good DOA estimation performance. Particularly, spherical arrays have the advantage of three-dimensional orientation which benefits three-dimensional processing and analysis. Also, spherical symmetric structure of spherical arrays could promote the computation of decomposing the wave field into spherical harmonics. Spherical harmonics are mainly studied in this paper as they form a set of complete orthonormal basis to represent signal functions on the 2-sphere space. In previous papers, the theoretical formulation of the spherical harmonics decomposition and spherical array beamforming has been presented. On the basis of existed theories, a spatial sampling scheme where hydrophones are uniformly distributed on the surface of an open sphere is proposed in this paper. The hydrophone array which employs 36 array elements around the sphere is used to compute spherical harmonics up to the 4th order. Beamforming is used to estimate DOA after spherical harmonics decomposition. From this dissertation we analyze the DOA estimation in underwater waveguide environment using simulation for different cases of single source and double sources. Simulation results show that the direct sound and several reflections can be identified.