Time-Offset Estimation in Multisensor Tracking Systems

Abstract

In this paper, a new algorithm is proposed for time-offset estimation in multisensor target tracking systems. First, the time offset pseudo-measurement equation is derived and calculated in both centralized and distributed scenarios, where measurements and local tracks are available at the fusion center, respectively. Second, the observability of time offset is analyzed theoretically with constant velocity (CV) and constant acceleration (CA) targets, showing that only relative time offsets between sensors are observable. Then, a two-stage relative time-offset estimation method is developed with two different formulations corresponding to different target dynamic models. Finally, simulation results show that the proposed algorithm meets the corresponding posterior Cramér-Rao lower bound (PCRLB), demonstrating the validity of the proposed algorithm.