Abstract:

A gridless direction-of-arrival (DOA) estimation method based on covariance matrix reconstruction in monostatic multiple-input multiple-output (MIMO) arrays is proposed. In this method, the MIMO arrays are equivalent to uniform linear arrays with enhanced signal-to-noise ratio (SNR) by the dimension reduction, and the problem of the target azimuth estimation is converted to sparse signal reconstruction based on mixed norm minimization (MixNM). Furthermore, a grid-based convex optimization problem equivalent is presented to the sparse reconstruction problem, which is modeled as a semi-definite programming problem. In order to solve the problem of the performance degradation caused by the grid size, the Toeplitz structure of the equivalent uniform linear array is used to model it into a semi-definite programming problem to reconstruct the noise-free covariance matrix. Finally, the target azimuth is estimated by Vandermonde decomposition. Compared with the traditional method based on the MixNM, the proposed method reduces the number of optimization variables. In addition, compared with other off-grid sparsity representation methods, the accuracy of the proposed method is not affected by the grid size, and it can estimate the coherent source targets. The simulation results show the effectiveness of the proposed method.

Key words: multiple-input multiple-output (MIMO), direction-of-arrival (DOA) estimation, joint sparsity signal recovery, gridless sparse representation, noise-free covariance matrix reconstruction