Multi-UAV Delay Optimization in Edge Caching Networks: A Mean Field Game Approach

In the edge caching networks (ECN), the request data of a large number of users brings great delay to the base stations/small base stations (BSs/SBSs). Specially, in some special scenarios, such as sports events, it is a big challenge to deploy the BSs/SBSs. This paper investigates a distributed delay optimization approach by using the large number of cache-enable unmanned aerial vehicles (UAVs). The goal of each UAV is to facilitate the BSs/SBSs to minimize the delay of user downloading contents subject to the contents popularity and its own flight strategy. Moreover, we formulate this distributed dynamic flight and caching problem as a mean field game (MFG) by considering the users under the ultra-dense networks. The numerical results show that the download delay is reduced and eventually reaches equilibrium with the limited flight energy consumption.