Minimizing Age of Information in UAV-Aided IoT Networks via Hierarchical Optimization of Cluster Heads and UAV Trajectories

초록

This paper addresses the problem of minimizing the average age of information (AoI) in an Internet of Things (IoT) network where a single unmanned aerial vehicle (UAV) is employed to collect data from clustered nodes. We propose a system model that incorporates node-specific data sizes, an efficient data aggregation and compression model, and a dynamic time model that captures the trade-off between UAV flight time and intracluster data aggregation time. The resulting NP-hard optimization problem involves the joint selection of cluster heads (CHs) and the determination of the UAV's trajectory. To solve this, we propose a hierarchical optimization scheme combining a genetic algorithm (GA) in the outer loop for global CH selection and a greedy algorithm in the inner loop for efficient UAV trajectory planning. The GA's fitness function is the inverse of the total mission time, which directly minimizes the average AoI by accounting for the complex interplay among flight efficiency, aggregation efficiency, and upload efficiency. Comprehensive simulations demonstrate that the proposed scheme significantly outperforms two benchmarks by effectively balancing these competing objectives through an optimized intra-cluster scheduling policy, leading to a near-optimal solution that minimizes the overall AoI of the entire network.

키워드