A NONLINEAR EQUALIZATION METHOD USING DEEP LEARNING TO IMPROVE ROF TRANSMISSION QUALITY OF A FREQUENCY MODULATED TWO-CHANNEL CRAN CONNECTION

Abstract

Radio over Fiber (RoF) represents an innovative technology for upcoming wireless networks, particularly within fifth-generation Cloud-Radio Access Networks (C-RAN). Additionally, given the widespread use of deep learning in various sectors like communication and data processing...This study examines the nonlinear impact of a fronthaul interface, demonstrated through numerical simulations, for two wireless signal channels operating in the VHF frequency band via continuous-phase frequency-shift keying (CPFSK) modulation. Furthermore, this research proposes the deployment of a nonlinear equalizer based on a deep neural network (DNN) to alleviate nonlinear impairments during long-distance data transmission. Experimental results involving a 50-km transmission demonstrate that employing a DNN with six hidden layers can effectively minimize nonlinear distortion.