Trajectory estimation algorithm for mobile nodes using encounter information and geographical information

In this paper, we propose a centralized algorithm to estimate trajectories of mobile nodes using a reasonable number of landmarks, ad hoc wireless communication facility and geographical information. The method identifies such trajectories where nodes move between two landmarks with the shortest time, which may be the shortest paths with high possibility, and uses their intermediate positions as “quasi-landmarks” to localize the others. The method can deal with a realistic case where mobile nodes move along pathways among obstacles. From simulation results, we show that the average estimation error is less than 40% of the wireless range under various scenarios.