DOA Estimation of Coherent Signals Based on Coherent Accumulation Vector

The direction-of-arrival (DOA) method based on multiple Toeplitz matrices reconstruction (MTOEP) method that employs all rows of the array output covariance matrix to reconstruct the Toeplitz matrix, which can effectively resolve coherent signals. However, the existing MTOEP method suffers serious performance degradation when in low signal-to noise (SNR) owing to the square transformation of noise energy. In this paper, a modified MTOEP method based on coherent accumulation processing is proposed. Firstly, because the strong coherence of the narrowband signal and the uncorrelated noise at adjacent snapshots, the accumulation vector is obtained via performing coherent accumulation processing on multiple snapshots. Then, the Toeplitz matrix can be reconstructed by exploiting the coherent accumulation vector and the received data from all sensors, which improves the signal energy, and further results in the improvement of the SNR. Finally, the DOAs can be estimated via the subspace-based methods. Theoretical analysis and simulation results show that compared with the existing MTOEP method, the proposed method provides an improved DOA estimation accuracy and a higher probability of successful resolution at low SNR.