Machine learning in optical networks enhancement based on channel allocation and spectrum analysis for 5G application

Undoubtedly, Machine Learning (ML) for Optical Communication (OC) has been a popular topic in recent years and is likely to remain so for the foreseeable future. Recent years have seen a tremendous increase in the pace of research in this area. The innovative nature of this study field is related more to the unusual nature of the application domain than to the employment of state-of-the-art ML algorithms in the methodologies used. This study suggests a new method for optimising optical networks for 5G services by using channel allocation and spectral analysis. In this setup, we use a Gaussian reinforcement convolutional learning model to allocate channels for optical network signals, and a multilayer adversarial markov encoder to improve spectrum efficiency. Measurements of spectrum efficiency, bit error rate (BER), channel efficiency, energy consumption, and training correctness are all taken as part of the experimental investigation. According to the findings, the plan was feasible, and decision-making methods based on machine learning algorithms were able to determine the most suitable route of communication.