C-RAN Uplink Optimization Using Mixed Radio and FSO Fronthaul

Abstract

Cloud radio access networks (C-RANs) are a promising architecture for 5G systems in which simple radio units (RUs) fronthaul signal to a central processor (CP) for joint decoding. Although the C-RAN has reduced cost and complexity, high data rate fronthaul links are necessary. In this paper, we investigate the joint design of wireless fronthaul networks using both radio frequency (RF) and radio-over-free-space optical (RoFSO) links in the uplink of a C-RAN. Unlike earlier work which focuses on performance characterization of RF/FSO fronthaul networks, this paper presents a novel optimization approach to jointly design the quantizers for the RF fronthaul links and the amplifier gains of the RoFSO fronthaul links which suffer from clipping distortion. A subset of RUs fronthaul data via radio links using Wyner-Ziv source coding subject to a shared sum capacity constraint, while other RUs employ RoFSO fronthaul which converts the incoming RF receptions to optical signals by analog modulation of a laser. The optimization problem jointly designs both RF front- haul and RoFSO fronthaul links to maximize the weighted sum user rates. Simulation results of a simple C-RAN using measured weather data for two locations demonstrate that adding RoFSO links results in drastic improvements in end user rates but requires careful design of RF and RoFSO fronthaul links.