Underdetermined 2D DOA Estimation Based on Bi-directional Parallel Sparse Linear Array with Reduced Mutual Coupling

Compared with uniform arrays, sparse arrays can achieve same estimation accuracy and degrees of freedom (DOFs) with less sensors and reduced mutual coupling, so a novel parallel sparse linear array to achieve undetermined two-dimensional (2-D) DOA estimation is proposed. To improve the estimation accuracy of sparsity-based DOA estimator, this paper introduce Low-rank matrix model for precise covariance matrix reconstruction and propose covariance vector sparse representation(CVSR) for 1-D DOA estimation. Moreover, TLS-CVSR based on proposed 1-D DOA estimator CVSR for 2-D DOA estimation of parallel sparse linear array is proposed. By implementing simulation experiments, it is demonstrated that TLS-CVSR have superior performance in terms of underdetermined 2-D DOA estimation and high estimation accuracy.