Heuristics for the Efficient Spectrum Assignment of Services in Elastic Optical Networks
Faced with the constant growth of data traffic, primarily caused by social networks, remote work, voice and video on demand, and online games, among others, Optical Networks are fundamental. They play an important role in transporting data and guaranteeing the quality of the services offered. With the technology of Elastic Optical Networks (EON)s, there are new solutions for better bandwidth optimization in the optical spectrum without the need to deploy new optical fibers in existing networks. EONs differ from traditional Wavelength Division Multiplex (WDM)s networks by controlling spectral granularity, increasing optical spectrum efficiency compared to existing networks. Due to optical modulation technologies: Orthogonal Frequency Division Multiplexing (OFDM), and Nyquist-Wavelength Division Multiplex (N-WDM), it is possible to control the spectral granularity for a very small spectral width. EONs are managed by Routing and Spectrum Assignment (RSA) algorithms, which determine routes and spectral position for the attribution of the requested demand, maintaining the contiguity and continuity restrictions imposed by the EONs. This work aims to evaluate, with the use of simulation, new RSA heuristics developed to improve the use and occupation of the optical spectrum. With better spectrum occupation algorithms, the assignment of services and the organization of slots of the optical spectrum using the C-band (4.4THz) are performed. These algorithms allow better use and flexibility of the spectrum because it tends to present non-contiguous bandwidth spaces over time, increasing the Blocking Probability (BP) and the fragmentation of the optical network. The new heuristics developed showed better performance results of EON, minimizing the computational load on the Network Management System (NMS) in the three topologies evaluated, compared to some algorithms present in the scientific literature.