Abstract
Flexible-grid optical transport networks (OTNs) are currently undergoing a massive scale upgrade to bandwidth variable transceivers (BVTs) to cater to the exponential Internet traffic growth. Network operators seeking to plan future capacity in a scalable manner can also add additional bands of operations or additional fibers in the OTN. We propose two strategies for planning multiband upgrades in OTNs, which use an integer linear programming (ILP) solution and a heuristic solution with a reinforcement learning (RL)-based extension. Network simulation studies on three differently sized OTNs show that the heuristic and its RL-based extension perform extremely close to the optimal ILP-based solution while reducing execution times drastically. Finally, a techno-economic study based on what we believe to be a novel cost model highlights the cost-per-bit-per-second savings through carefully planned upgrades.
© 2024 Optica Publishing Group
Full Article | PDF ArticleMore Like This
Mario Wenning, Maria Samonaki, Sai Kireet Patri, Tobias Fehenberger, Helmut Griesser, and Carmen Mas-Machuca
J. Opt. Commun. Netw. 16(5) B81-B90 (2024)
Shrinivas Petale, Aleksandra Knapińska, Egemen Erbayat, Piotr Lechowicz, Krzysztof Walkowiak, Shih-Chun Lin, Motoharu Matsuura, Hiroshi Hasegawa, and Suresh Subramaniam
J. Opt. Commun. Netw. 16(9) E48-E60 (2024)
Soheil Hosseini, Ignacio de Miguel, Noemí Merayo, Juan Carlos Aguado, Óscar González de Dios, and Ramón J. Durán Barroso
J. Opt. Commun. Netw. 15(11) F22-F35 (2023)