Performance Analysis of Hybrid UAV Networks for Probabilistic Content Caching

Caching content in small-cell networks can reduce the traffic congestion in backhaul. In this paper, we develop a hybrid caching network comprising of unmanned aerial vehicles (UAVs) and ground small-cell base stations (SBSs), where UAVs are preferred because of their flexibility and elevated platform for line-of-sight. First, we derive the association probability for the ground user affiliated with a UAV and ground SBS. Then, we derive the successful content delivery probability by considering both the inter-cell and intra-cell interference. We also analyze the energy efficiency of the hybrid network and compare it with the separate UAV and ground networks. We further propose the caching scheme to improve the successful content delivery by managing the content popularity, where the part of the caching capacity in each UAV and ground SBS is reserved to store the most popular content (MPC), while the remaining stores less popular contents. Numerical results unveil that the proposed caching scheme has an improvement of 26.6% in content delivery performance over the MPC caching which overlooks the impact of content diversity during caching.