Multi-sensor Bearings-only Target Tracking Using Two-stage Multiple Hypothesis Tracking

Abstract

The problem of multi-sensor bearings-only target tracking is addressed in this work. A practical challenge in this problem stems from the data association in dense clutter. In order to avoid the increase of computational complexity with the increase of the number of sensors and clutter density, a two-stage multiple hypothesis tracking (MHT) algorithm is proposed. First, the first stage MHT is performed at each local sensor, and the measurements with effective data association are sent to the fusion center. Second, the measurements from different sensors are combined and augmented to form new measurement vectors, which are then converted to Cartesian coordinates by iterated least squares (ILS) estimator. Therefore, in the second stage MHT, the Kalman filter can be used to update the target track. The performance of the two-stage MHT algorithm is illustrated with an example.