Title:Multi-mode Tracking of a Group of Mobile Agents

Abstract:We consider the problem of tracking a group of mobile nodes with limited available computational and energy resources given noisy RSSI measurements and position estimates from group members. The multilateration solutions are known for energy efficiency. However, these solutions are not directly applicable to dynamic grouping scenarios where neighbourhoods and resource availability may frequently change. Existing algorithms such as cluster-based GPS duty-cycling, individual-based tracking, and multilateration-based tracking can only partially deal with the challenges of dynamic grouping scenarios. To cope with these challenges in an effective manner, we propose a new group-based multi-mode tracking algorithm. The proposed algorithm takes the topological structure of the group as well as the availability of the resources into consideration and decides the best solution at any particular time instance. We consider a clustering approach where a cluster head coordinates the usage of resources among the cluster members. We evaluate the energy-accuracy trade-off of the proposed algorithm for various fixed sampling intervals. The evaluation is based on the 2D position tracks of 40 nodes generated using Reynolds' flocking model. For a given energy budget, the proposed algorithm reduces the mean tracking error by up to $20\%$ in comparison to the existing energy-efficient cooperative algorithms. Moreover, the proposed algorithm is as accurate as the individual-based tracking while using almost half the energy.