Evaluation of Sparse Array Geometries to Improve DOA Estimation With Mutual Coupling

Abstract

Translate Article

Direction of arrival (DOA) estimation is a classic problem in array processing. It has wide applications in wireless communications, radar, acoustics, and smart antennas, to name just a few. In this paper, we present a novel comparative study of six array geometries and their impact on mutual coupling and DOA estimation performance. Two scenarios are presented: with and without mutual coupling. A numerical software package was used to simulate the array geometries with dipole antennas. The mutual coupling matrix was calculated from the S‐parameters of the simulated arrays at the operating frequency based on specific sensor locations. The MUSIC algorithm was used to resolve the source location using the difference coarray technique. A well‐known smoothing technique was used to improve the rank of the covariance matrix, allowing MUSIC to compute the DOA using the sparse arrays. Several properties of each array are described, such as the physical sensor location, difference coarray, weight function, and coupling matrix. The sparse array HFCA and UACA achieved an RMSE below 2 degrees (for both 9 and 16 sources and the whole SNR range) and 0.5 degrees (for both 9 and 16 sources and SNR equals to 30 dB), respectively. The performance of the DOA estimation was compared using RMSE as a metric, considering the environment configuration with different number of sources and SNR conditions.