User scheduling and cluster formation in fog computing based radio access networks

Fog computing based radio access networks (FRANs) are potential paradigms for further improving throughput and decreasing latency. By fully utilizing the caching and signal processing capabilities of fog computing based access points (FAPs), the burden on the fronthaul can be alleviated. However, when the spectrum is shared among FAPs, the inter-FAP interference can be severe, which degrades the performance of F-RANs. In this paper, to mitigate inter-FAP interference, a distributed cluster formation algorithm is proposed, which considers the fronthaul cost induced by fetching contents missed in the cache. Meanwhile, a distributed user scheduling algorithm is designed to help each FAP schedules the proper user, aiming at optimizing the system throughput. By simulations, the benefit of enabling caching is validated, and the effectiveness of the proposed scheduling algorithm is shown.