A Distributed Approach in Uplink Device-to-Device Enabled Cloud Radio Access Networks

Device-to-device (D2D) enabled cloud radio access networks (C-RANs) are potential solutions for further improving spectral efficiency (SE) and decreasing latency by allowing direct communication between two user equipments. Due to the acquirement of global channel state information (CSI) and the execution of centralized algorithms in the uplink D2D enabled C-RANs, heavy burdens are put on fronthaul and the baseband unit pool. To tackle this challenge, a game-theoretic approach to mode selection and resource allocation for potential D2D pairs is proposed with a distributed manner, in which pairs are endowed with decision-making capabilities. The proposal is categorized into three stages: communication mode and subchannel selection, remote radio head (RRH) association, and reinforcement learning based strategy update. The core idea is that D2D pairs autonomously optimize the mode selection and resource allocation without global CSIs under several practical constraints. Simulation results show that enabling D2D can significantly improve SE for C-RANs. Furthermore, the performance gain is mainly determined by the fronthaul capacity and the distance between D2D transmitters and RRHs.