Comparative study of track-to-track fusion methods for cooperative tracking with bearings-only measurements

Using a network of spatially distributed sensors to track a moving object can be a challenging task. In applications with limited communication between sensor nodes and packet loss, it may be impossible to process measurements from these distributed sensor nodes in a central unit. Therefore, it is often necessary to use only the locally available measurements at the sensor nodes and afterwards merge all local tracks into one consistent result. In this paper, several different track-to-track fusion algorithms are compared to cooperatively track a moving object using only bearing measurements. It is shown that the Sample-based Fusion that uses a set of deterministic samples to reconstruct the cross-covariances is a suitable fusion algorithm for the considered setup. Furthermore, it provides the means to efficiently keep track of the cross-covariances between sensor nodes and therefore outperforms conservative methods. The proposed approach is also tested in a real-world indoor localization setup using bearings-only acoustic measurements from three microphone arrays.