Maximum Likelihood DOA Estimator Based on Particle Filtering

Maximum Likelihood Estimator (MLE) has been shown to be the best performance in Direction-Of-Arrival (DOA) estimation. However, the computational burden of a multidimensional grid search for MLE is very large. In order to resolve the question of computation burden, particle filtering methods are combined with maximum likelihood DOA estimator. A novel Maximum Likelihood DOA Estimator based on Particle Filtering (MLE-PF) is proposed. Numerical simulations illustrate the fact that MLE-PF keeps the perfect performance of MLE and reduces the computational complexity of MLE from O(L^K) to O(K times N_s). Also MLE-PF performs better than MUSIC and MiniNorm, especially in low SNRs.