A ML-Based Direct Localization Method for Multiple Sources with Moving Arrays

Abstract

The problem of localizing multiple sources with moving arrays is considered. The conventional two-step methods first extract measurement parameters from the received signals, and then calculate the position via these parameters. Without estimating intermediate parameters, direct position determination (DPD) methods can locate emitter from the signals directly. Due to the great computation load of the straight implement of maximum likelihood (ML) estimators, existing DPD algorithms for multiple sources based on ML criterion remain few. Thus we apply a decoupled iterative idea to the ML-based DPD method. The proposed method decouple one emitter from the other emitters in one step instead of traversal search for all sources, reducing computation complexity substantially. Simulation results demonstrate that our algorithm performs superior and more robust to low signal-to-noise ratios (SNRs) than other location methods.