Tracking/fusion and deghosting with Doppler frequency from two passive acoustic sensors

It is known that using two passive sensors that provide bearings-only information cannot eliminate false intersections in a multi-target environment, and results in ghost tracks. In this paper, we propose a tracking/fusion and deghosting method using two stationary acoustic sensors. The existing approaches for bearings-only information require the number of sensors to be greater than or equal to three. The method proposed in this paper makes target tracking possible in a two-sensor system through making use of frequency information. This method generates tentative tracks from all possible bearing track combinations, and distinguishes target tracks from ghost tracks using a 2-D assignment algorithm. Tentative tracks are initiated using an iterated Least Squares (LS) algorithm, and updated by an Unscented Kalman Filter (UKF) based on Doppler-Bearing Tracking (DBT) approach. The likelihood ratios of tentative tracks, which are computed based on track innovation probability density functions, are assigned as costs in the 2-D assignment algorithm. Simulation tests have been conducted to validate the effectiveness of the proposed method.