Optimal Capacity Allocation and Caching Strategy for Multi-UAV Collaborative Edge Caching

Abstract

UAVs-assisted content caching has emerged as a promising content delivery paradigm to effectively reduce the user equipments (UEs) access delay caused by the ever-increasing data traffic in current networks. Existing research on UAVs caching mainly focuses on stand-alone UAVs, without considering the complicated service interactions among UAVs in a cooperative manner. To this end, we focus on the UAVs collaborative edge caching scheme with one more link to retrieve the target files from one UAV to another. We propose a three-stage sequential optimization model to capture the complicated interactions involved in such a system. Specifically, in Stage I, we aim to maximize the benefit of the Internet Content Provider, by optimizing the number of UAVs and the cache capacity of UAVs to be deployed. In Stage II, given the UAVs’ deployment, we optimize the service scheduling for UAVs to balance the service energy and communication energy. In Stage III, we assign a caching strategy to maximize the number of UEs served by UAVs. Extensive simulation results show that our proposed collaborative edge caching model significantly outperforms existing edge caching strategies in terms of the number of UEs served by UAVs, the UAV cache hit rate, the average downloading latency, and the diversity of files.