Optimal Offloading for Interdependent Tasks of Unmanned Aerial Vehicle in Edge Server aided Low Power Wide-Area Networks

In this work we study the optimal offloading problem for an unmanned aerial vehicle (UAV) connected to edge servers aided low power wide-area network (LPWAN). The UAV, served as a device, a gateway or a base station in the LPWAN, has some computation-intensive tasks with interdependence to be accomplished during its given path with fixed start and end points. By allowing the offloading to edge servers in the LPWAN, we aim at minimizing the total processing time of all tasks by jointly designing its trajectory and offloading decision. An analytically optimal trajectory is proposed and the optimal offloading strategy is obtained by solving an integer programming. Numerical results show that the proposed algorithm significantly reduces the total processing time compared to the other benchmarks.