Routing and dynamic core allocation with fragmentation optimization in EON-SDM

Abstract

The increasing need for rapid data transmission in optical networks has made the effective allocation of spectrum resources a crucial necessity. This study presents a new spectrum allocation approach that incorporates path selection and careful consideration of cross-talk (XT) limitations. We propose a novel fragmentation coefficient and dynamic core-changing-based routing, spectrum, and core allocation (FC-DCC-RSCA) technique. The algorithm begins by searching for the shortest paths using the Dijkstra algorithm. It then proceeds to evaluate fragmentation coefficients (FC) using the Continuous Aligned Slot Ratio (CASR) model. Subsequently, the FC undergoes a filtration process wherein only those falling inside predetermined fragmentation ranges (min and maxFC) are retained. Further spectrum allocation has been done based on XT constraints. The algorithm has better proficiency in dynamic flexibility, enabling it to effectively handle varying traffic loads. By conducting a series of rigorous simulations, our proposed approach exhibits a notable decrease in bandwidth blocking probability (BBP) while also opening the door for considerable improvements in spectrum utilization. The algorithm’s proficiency in adapting to changing network circumstances signifies its importance as a helpful thing in enhancing the allocation of resources within optical networks. It is well-positioned to address the developing requirements of contemporary communication systems. The spectrum utilization of the FC-DCC-RSCA increases by 12.5% and 6.25%, exhibits a reduction in its FR by 13.5% and 2.7%, and has a BBP of 0.051 for the traffic load of 450 Erlang, compared to SBP-PMC and FMDE-RSCA. The algorithm that has been provided provides opportunities for more investigation and enhancement. Potential areas for future research include the examination of strategies to improve energy efficiency, the development of scalable solutions for bigger network infrastructures, and the incorporation of machine learning methodologies to facilitate dynamic decision-making processes. Furthermore, it is worth investigating the potential of the method to be applied in developing optical communication technologies, such as space-division multiplexing. These endeavors hold the potential to enhance the current level of knowledge and practices in the field of optical network management and optimization.