Abstract
Unmanned Aerial Vehicles (UAVs) have been extensively utilized in numerous scenarios where freshness of information is paramount. Due to the limited endurance of a single UAV, multiple UAVs can ensure the freshness of sensor node information while completing data collection tasks. Therefore this paper focuses on the multi-UAV-aided collaborative data collection problem, aiming to enhance the freshness of information. We use Age of Information (AoI) to measure information freshness, primarily including the sensors’ data aggregation time within clusters, the UAVs’ data collection time and flight time. Firstly, we employ a simulated annealing-based algorithm to construct data aggregation trees within clusters, determining the scheduling order of sensor nodes. Secondly, we assign tasks to the UAVs and establish the association between clusters and UAVs. Finally, based on the results of the previous two steps, we optimize the flight 1 trajectory of UAVs using two schemes, 2-opt Heuristics via Deep Reinforcement Learning and greedy-based multi-trajectory planning. Simulation results demonstrate that this strategy can optimize the average AoI of sensor networks and improve the information freshness.