Controlling Mobile Sink Trajectory for Data Harvesting in Wireless Sensor Networks

This paper proposes a mathematical optimization model to maximize the lifetime of wireless sensor networks through determining the optimal trajectory (OT) of mobile sink (MS). We address deadline and event based applications where by capturing an event, a sensor node has to send its data to MS in a restricted time slot defined as a deadline. We demonstrate that the addressed problem is in NP-hard form and then by dividing the problem into two phases, we propose a heuristic approach based on mathematical optimization. In the first phase, the trajectory of MS is determined through proposing a convex mathematical optimization model; in this step, we specify an optimal line as OT with respect to the current location and constant velocity of MS; moreover, the volume of captured data by sensor nodes, deadline and geographical locations of sensor nodes are taken into account. We extend our work in the second phase by proposing a mixed integer linear programming (MILP) model to relax the constant velocity assumption of MS. To obtain an optimal solution of MILP, subsequently a tabu-based algorithm is proposed. The effectiveness of our approach is validated via the extensive number of simulation runs and comparison with other proposed algorithms.