Performance investigation of FSO-WDM system employing ML under dynamic environmental conditions

Abstract The primary objective of the present work is to enhance the performance of free space optical – wavelength division multiplexing (FSO-WDM) system under varying atmospheric conditions. The FSO-WDM network leverages 12 continuous wave (CW) laser sources with a channel spacing of 0.5 nm, transmitting data at 5 Gbps per channel using nonreturn-to-zero (NRZ) modulation. The quality factor (Q-factor) of received signal has been ascertained as a function of transmission wavelength at each node considering clear air, fog, and rainy weather conditions. The proposed system demonstrates superior performance with deployment of repeaters at 1 km intervals of the FSO channel resulting in achievement of higher Q factor (35) at a wavelength of 1,550 nm in case of rainy season. A machine learning (ML) model supported by genetic algorithm (GA) has also been implemented to dynamically optimize the performance of the system in terms of Q factor and increase the power efficiency by 5 %.