Tracking of Multiple Targets Using Optimal Number of UAVs

Abstract

In this paper, we study the problem of mobile target tracking using the fewest number of mobile trackers for two different types of targets. Given the target trajectories and the period of observation, we propose techniques to compute the minimum number of trackers and their trajectories required to track all mobile targets. Two classes of mobile targets are considered in this paper: 1) targets that need tracking for the entire duration of observation and 2) targets that need tracking at least once during the period of observation. We show that even when target trajectories are known in advance, the problem is computationally hard, i.e., NP-complete. We formulate the problem as a network flow problem and propose algorithms for its solution. We evaluate the performance of our algorithms through simulation and study the impact of parameters such as the speed and sensing range of the trackers.