2D DOA estimation by a large-space T-shaped array

Abstract

In this paper, a two-dimensional (2D) direction of arrival (DOA) estimation algorithm by a large-space T-shaped array is proposed. The used T-shaped array consists of a generalized unfolded co-prime array and a symmetrically unfolded co-prime array. The minimum element space (MES) of the proposed T-shaped array can exceed half-wavelength of received signals. The symmetry of the T-shaped array is used to construct an extended cross-covariance matrix, singular value decomposition (SVD) of which can obtain multiple signal subspaces. By combining the co-prime characteristic of element spacing with the rotational invariance of uniform linear array, three extended signal subspaces can be obtained without using virtual elements. Using the three extended signal subspaces, high-precision 2D DOA estimation can be got. The proposed T-shaped array has strong robustness to mutual coupling. The proposed method can eliminate angle ambiguity caused by large-space and has higher estimation precision than many similar 2D DOA estimation algorithms. Many simulation results can prove the performance of the proposed scheme.