Forward/backward spatial reconstruction method for directions of arrival estimation of uncorrelated and coherent signals

An effective method to classify the directions of arrival (DOA) estimation of narrow-band signals in a coherent environment is proposed. The basic idea is to rearrange the correlation matrix. By combining the forward reconstructed correlation matrix with its conjugated backward form together, it is always possible to estimate any ⌊2(M+1)/3⌋ one-dimensional DOAs using an M-element uniform linear array (ULA) and any ⌊2(3⌊2⌊2)MN⌋ two-dimensional DOAs using an M×N-element uniform rectangular array. Other arrays that contain centrosymmetric sub-arrays can also be applied with our method. Simulation results demonstrate that the proposed method could both enlarge the effective aperture and reduce the computational complexity.