Enhanced F-RANs by Utilizing Reconfigurable Intelligent Surfaces and UAV Smart Helpers

In traditional fog-radio access networks (F-RANs), deploying many enhanced remote radio heads (eRRHs) faces challenges due to site limitations and fronthaul link costs. To tackle this issue, we introduced smart helpers (SHs) in our earlier study. SHs mainly listen to communications between eRRHs and users, smartly cache popular content, and serve users without having a fronthaul link to the macro base station (MBS). In this paper, we explore using an unmanned aerial vehicle (UAV) as the SH to assess the potential benefits of integrating UAVSHs into F-RANs. Because UAVs typically have limited battery life and cover smaller areas, we leverage a reconfigurable intelligent surface (RIS) to extend the service area of the UAVSH. To assess the performance of the considered UAVSH-aided F-RAN, we formulate the problem of average delivery delay minimization. To tackle the problem, we develop an algorithm to optimize user and RIS scheduling and caching decisions using multiagent reinforcement learning (MARL). The simulation results numerically prove that integrating a UAVSH into a RIS-assisted F-RAN leads to significant improvements in delivery delay, fronthaul load, and cache hit rate metrics.