UAV Trajectory Planning for IoT Data Collection and Offloading With Energy Constraints

This work studies unmanned aerial vehicles (UAVs) for data collection and offloading. The UAV collects data from different types of Internet of Things (IoT) devices on the ground and offloads the collected data to their corresponding edge servers. Due to limited energy, the UAV must replace its battery at a battery station before the energy is exhausted. We aim to minimize the total completion time by optimizing the UAV trajectory. The formulated problem is an NP-hard problem, which is intractable in finding optimal solutions in large-scale cases. To solve the problem, we propose an efficient solution with a three-stage heuristic time minimization trajectory planning (TMTP) algorithm. In the first stage, we simplify the problem as a traveling salesperson problem (TSP) with precedence constraints and solve it to find the main visiting order. In the second stage, we propose a dynamic programming-based algorithm to insert battery stations in the trajectory to satisfy the energy constraints. In the last stage, we refine sub-trajectories and reinsert battery stations by an iterative algorithm to further reduce the previous time cost. Simulation results demonstrate that our proposed algorithms are more efficient than the baselines while evaluating the objective and the running time.