Sparse DOA Estimation Based on Multi-Level Prime Array With Compression

Abstract

A signal emitter can be located using diverse types of direction finding (DF) techniques. One of the most widely used techniques is the direction of arrival (DOA) estimation using antenna arrays. An array configuration that can increase the degrees of freedom (DOF) or the number of estimated sources is desired. Multi-level prime array (MLPA) uses multiple uniform linear subarrays where the number of elements in the subarrays is a pairwise coprime integer. Compared with nested and coprime arrays, the MLPA requires a smaller aperture size which is important in mobile applications. The different MLPA configurations can be constructed for a given number of antennas and the one that maximizes the DOF is exploited. These configurations have a difference coarray with a large number of consecutive lags and few holes. The number of consecutive lags can be increased by properly compressing the inter-element spacing of one subarray under a fixed number of antennas and without changing the aperture size. This paper proposes a new compressed MLPA configuration and demonstrates its performance in sparse DOA estimation. The resultant array, MLPA with compressed subarray (MLPAC), can have a hole-free difference coarray as in nested array case. The MLPAC can estimate a larger number of sources using both MUSIC and sparse reconstruction algorithms. Mutual coupling between sensors has also been evaluated. The simulation results confirm the achievable DOF and the advantage of the proposed configuration in the DOA estimation.