Performance Analysis of Service Caching in Large-Scale UAV-Enabled Wireless Networks with Coupled Uplink and Downlink

In this paper, we investigate the service caching performance in a large-scale unmanned aerial vehicle (UAV)-enabled wireless network, in which multiple UAVs equipped with caching and computation capabilities are deployed to support multiple types of tasks generated by the terrestrial user equipment (UEs). We consider a practical-yet-challenging assumption where the uplink transmission for task offloading and the downlink transmission for results feedback are coupled, and derive analytical expressions of the successful transmission probability and successful service probability, demonstrating how they are determined by the UAV's density and operating height, as well as the service caching strategy. Through numerical results, we verify the correctness of our analysis and show that the SSP of our system can be significantly enhanced by optimizing the density and operating height of the UAVs.