A Framework for Information Freshness Analysis in UAV-based Sensing and Communications

Mission-critical wireless networks should prioritize improving latency and reliability over data rate, while assuming massive heterogeneous users are served proactively. In this paper, a multi-tier sensing and communications structure is proposed for a network of Unmanned Aerial Vehicles (UAVs) that combines the advantages of both central and distributed configurations. To reduce the impact of packet overloading during the end-to-end delay analysis, optimal server load balancing is achieved by the minimization of a freshness metric known as the Age of Information (AoI). The optimal server load rate, which is the deciding factor of AoI for each UAV in the multi-tier network, is shown to be responsible for maximizing the sensing information transmitted to the central UAV. A reinforcement learning-based solution is proposed that finds the weighted sum of expected AoIs for the targets involved in the successful sensing process. Various performance parameters, such as minimum outage probability and maximum mutual information, are optimized and a comprehensive analysis of data freshness is presented in order to efficiently integrate the sensing data from multiple geographically dispersed targets into the UAV network and finally into the central node.