Analysis of RWA in WDM optical networks using machine learning for traffic prediction and pattern extraction

Machine learning (ML) has attracted researchers to discover numeral solutions in the field associated with optical networking problems. In this paper, ML discipline procedures have been discussed to proficiently execute Routing and Wavelength Assignment (RWA) for contribution in traffic calculation in the Wavelength Division Multiplexing (WDM) optical set-up. The growing demand for data transport through WDM networks has created problems related to the search for routes and the assignment of wavelengths in these networks. In optical networks, RWA is a well-known problem. To address this problem, researchers have proposed simple to complex heuristic machine learning algorithms. This paper describes how machine learning support can be used and shared in optical networks and the assessment of transmission quality (QoT), data traffic patterns, and crosstalk detection to help route and distribute resources. The RWA algorithm assessments rely on performance measures such as the blocking probability, network utilization, etc. The paper summarizes future research trends for the use of routing and distribution of resources in machine learning processes in optical networks with the results obtained.