Fast DOA estimation algorithm for improving performance of SPT-LCMV beamformer

To eliminate coherent interferences, a conventional split polarity transformation-linearly constrained minimum variance (SPT-LCMV) beamformer uses the SPT matrix to remove any correlation between the desired signal and coherent interferences. After removing the correlation, an LCMV beamforming is applied to obtain the desired signal. However, it needs to estimate direction of arrivals (DOAs) for all coherent interferences. Conventional algorithms to estimate the DOAs of coherent interferences are computationally complexity. Moreover, the performance of the SPT-LCMV beamformer degraded severely as the estimated DOA errors increased. Accordingly, high-order constraints are required to compensate the DOA estimation errors. As a result, the SPT-LCMV beamformer needs many array elements in proportion to the number of the constraints applied to generate the SPT matrix. This paper proposes a DOA estimation algorithm to overcome the disadvantages of the SPT-LCMV beamformer. The proposed algorithm has a significantly reduced computational complexity and can also reduce the number of the high-order constraints compared to the conventional DOA estimation algorithms.