TDoA and Monostatic Radar Data Fusion for Single Object Localization and Tracking

In this paper, we present the fusion of passive Time-Difference-of-Arrival (TDoA) and active monostatic radar measurements in the context of single object localization and tracking. We assume a target equipped with a wireless transmitter so that the passive-TDoA and active-radar sensors can simultaneously detect it. For the localization problem, we base our approach on the Constrained-Weighted-Least-Squares (CWLS) method, which has better accuracy than the Least-Squares (LS) and Weighted-LS approaches. Nevertheless, CWLS suffers from estimation bias when the measurement noise is large. We show that it is possible to reduce such a bias by using both TDoA and monostatic radar measurements. For the tracking problem, we base our approach on Bayesian filtering. Therefore, we detail the TDoA, radar and combined measurement models required for the filtering process regardless of the Bayesian Filter implementation. We show that the combined measurements result in enhanced tracking capabilities, e.g., improved reliability and larger tracking area compared to the radar-only case. In addition, we present other possible combinations of radar and TDoA measurements that one could encounter in practical scenarios. Finally, we present experimental results to validate and support our work.