Machine Learning-Based Digital Pre-Distortion Scheme for RoF Systems and Experimental 5G mm-waves Fiber-Wireless Implementation

The advent of the 5th generation mobile networks brought a large number new use case and applications to be supported by physical layer (PHY), which must more flexible than all previous radio access (RAN). concept centralized RAN (C-RAN) allows baseband processing performed in central office, simplifying network deployment also allowing operators dynamically control PHY according with requirements. radio-frequency (RF) signal generated C-RAN can transported remote unit (RRU) using over fiber (RoF) system. In this paper, we propose two RoF approaches for composing transport next-generation systems. first investigation relies on implementation machine learning-based digital pre-distortion (DPD), designed second approach, implement an system characterize optical electrical power levels aiming reduce non-linear distortions. overall link performance is evaluated measuring error vector magnitude (EVMRMS) 590 Mbit/s achieved EVMRMS as low 4.4% 10 m reach cell.