Optimizing UAV scheduling and trajectory planning: An online auction framework

Unmanned Aerial Vehicles (UAVs) are envisioned to be a critical form of network service provisioning, when the ground infrastructure is vulnerable to disruptions from conflicts and natural disasters. Existing methodologies often fall short in fully optimizing UAV scheduling and resource allocation, leading to suboptimal service performance. This work aims to enhance social welfare through refining UAV scheduling and trajectory planning processes. To address this complex challenge, we first formulate social welfare maximization into a non-traditional integer linear program, and subsequently transform it into its exponential and dual forms. We propose a bifurcated framework called Online Scheduling and Trajectory (OST) which comprises two algorithms: The $A_{OST}$ algorithm is responsible for managing task bids and allocating UAV resources by taking into account bid values, available resources, and task requirements, prioritizing tasks based on their intrinsic value. The $A_{dual}$ algorithm optimizes task selection and UAV trajectory planning by balancing the costs and benefits associated with each task. Theoretical analysis demonstrates that the proposed approach achieves an equilibrium that significantly enhances social welfare by ensuring optimal decisions regarding task allocation and resource distribution. Empirical evaluations corroborate these findings, illustrating notable improvements in network service efficiency and validating the practical applicability of our method in maximizing social welfare.