Weighted Maximum-likelihood source localization for wideband signals in the near-field

In this paper, the source localization of wideband signals in near-field is investigated. Based on the conventional Maximum-likelihood source localization, we propose a more efficient algorithm called Weighted Maximum-likelihood source localization (WML). In this proposed algorithm, the ML estimate is obtained by using the spectrum on each frequency bin, and then all estimates are weighted according to the corresponding SNRs, where the fact that the SNRs are different at each frequency bin mostly for wideband signals is utilized. Simulation results demonstrate that the proposed algorithm outperforms the conventional Maximum-likelihood solution.