UAV-Assisted Timely Data Collection and Enhanced Information Coverage in IoT Networks

The high flexibility and low operational cost of unmanned aerial vehicles (UAVs) have made it feasible to collect data from internet of things (IoT) devices in remote areas. However, achieving wide-ranging data collection in dynamic environments under limited energy constraints presents significant challenges. This paper investigates the real-time data collection problem in UAV-assisted IoT networks, where UAV gathers data from edge data collection node (DN). Each DN has a specific coverage range for data collection. To address the problem of age measurement under varying regional data coverage rates, we propose age of regional synchronization (AoRS). We formulate the AoRS minimization problem as a Markov decision process and adopt multi-task transfer deep reinforcement learning approach to optimize UAV flight trajectories. Simulation results demonstrate that the proposed algorithm outperforms baseline strategies in terms of adaptability to environmental changes.