부산대학교 도서관

This article analyzes the utilization of multipath measurements in multitarget tracking to achieve better track formation in the presence of false alarms, missed detections, and a known reflection surface for automotive applications. In complex urban environments, the transmitted signal of the automotive radar gets scattered or reflected and creates multipath returns, which are undesirable in tracking as they generate false tracks and degrade the tracking performance. In this article, a technique is proposed not only to mitigate but also to leverage multipath returns to improve the tracking performance. Target-to-measurement and measurement-to-propagation association uncertainties need to be resolved in order to incorporate the multipath returns. Although the standard data association could partially handle these uncertainties for already initialized tracks, a joint approach that considers track initialization with data association is required to eliminate the redundant track formation with multipath. Therefore, a multiframe assignment technique that jointly considers both track initialization and data association is proposed in this work. The performance of the proposed algorithm is compared with two other approaches. The simulation results show that the proposed algorithm can handle the multipath efficiently.