Radio over fiber for cellular networks: system identification and pre-distortion strategies

Abstract

5G systems will rely on the use of massive of number of antennas in the base-stations to meet the rate, latency and coverage requirements of next generation cellular networks. This will introduce challenging constraints on the front-haul links connecting the base band processing units (BBUs)s and the radio remote units (RRUs). Digital front-haul transmission strategies like CPRI will need to deliver huge data rates because of the number of antennas and the bandwidth of the radio frequency signals that are used. An alternative strategy using analog radio-over-fiber can relax these constraints in front-haul links. In particular, it allows to deploy dummy remote radio units with reduced complexity, lighter weight and lower power consumption compared to digital solutions. However, the use of radio-over-fiber transmissions introduces non-linear distortions that need to compensated in order to satisfy the requirements on adjacent leakage ratio (ACLR) that are decreed by the 3GPP standard. In order to keep minimal complexity in RRUs, this paper investigates a pre-distorsion approach that is implemented in the BBU. A feedback enabling this pre-distortion and requiring minimal complexity in the RRU is presented. An experimental validation assessing the efficiency of the proposed approach and showing its capability to satisfy 3GPP requirements is presented.