Performance enhancement of FSO communication system using machine learning for 5G/6G and IoT applications

Abstract

6G networks will provide extremely high capacity and will support a wide range of new applications in the future, but the existing frequency bands may not be sufficient. Furthermore, because traditional wireless communications are incapable of providing high-speed data rates, 6G enables superior coverage by integrating space/air/underwater networks with terrestrial networks. 5G-and-beyond (5 GB) and 6G networks have been mandated as a paradigm shift to take the enhanced broadband, massive access, and ultra-reliable and low latency services of 5G wireless networks to an even more advanced and intelligent level, to meet the ever-growing quantities of demanding services. In 5G and 6G wireless communication systems, artificial intelligence (AI), particularly machine learning (ML), has emerged as an essential component of fully intelligent network orchestration and management. 5 GB and 6G communication systems will also rely heavily on a tactile Internet of Things (IoT). The diverse nature of heterogeneous traffic and the established service quality parameters in 5 GB networks will present numerous challenges. Many other wireless technologies, including free space optics (FSO), look promising for meeting the demands of 5 GB systems. FSO has been identified as a promising technology for achieving higher data rates while consuming less power. However, attenuation due to weather, pointing errors, and turbulences limits its performance. Traditional Maximum likelihood decoding techniques require prior channel information to decode the signals. in this paper, first time we proposed a novel decoding technique for decoding on–off keying (OOK) modulated FSO signals using support vector machines (SVM). The model is tested under various atmospheric weather conditions such as fog, rain, and snow, as well as turbulence and pointing errors. Simulated numerical results demonstrate that the proposed SVM-based decoding schemes are capable of mitigating attenuation, pointing error, and turbulent channel impairments.