3D Sparse Array and Its DOA Estimation Based on Coupled Tensor Decomposition

To improve the performance of two-dimensional direction-of-arrival (DOA) estimation of signals, a three-dimensional sparse array structure and its coupled tensor decomposition-based DOA estimation method was proposed. By utilizing the second-order statistics of the signals from these subarrays, a virtual three-dimensional cross array was constructed, where the element spacing in the z-axis direction was sparse. Analysis showed that when N32+5N22+3 (where N was an odd number greater than 2) physical elements were used in the array, the corresponding virtual array possessed an aperture of (N3+2N-1)d in the x-axis and y-axis directions and (N3+3N)d in the z-axis direction. To fully exploit the large array aperture of the virtual array for enhancing DOA estimation performance and eliminating phase ambiguity caused by the element spacing greater than d in the z-axis direction, a method using coupled tensor decomposition for resolving phase ambiguity and achieving DOA estimation was developed. Theoretical analysis and simulation results demonstrate that, when the number of physical elements used in arrays is identical, the proposed method can yield a better estimation performance than existing three-dimensional arrays, because it has a larger array aperture.