Abstract
Using maximum achievable throughput as an objective, message passing neural networks (MPNN) are applied to larger optical networks (25-100 nodes), enabling physical properties-aware large-scale topology optimisation in record time, reducing computation time by 5 orders of magnitude, with close to perfect throughput correlation (ρ = 0.986).
© 2022 The Author(s)
PDF ArticleMore Like This
Ruijie Luo, Yi-Zhi Xu, Robin Matzner, Georgios Zervas, David Saad, and Polina Bayvel
Th1F.5 Optical Fiber Communication Conference (OFC) 2022
Bowen Bao, Hui Yang, Yu Wan, Qiuyan Yao, Ao Yu, Jie Zhang, Bijoy Chand Chatterjee, and Eiji Oki
F2G.2 Optical Fiber Communication Conference (OFC) 2021
Ruikun Wang, Jiawei Zhang, Shuangyi Yan, Chuidian Zeng, Hao Yu, Zhiqun Gu, Bojun Zhang, Tarik Taleb, and Yuefeng Ji
We4B.4 European Conference and Exhibition on Optical Communication (ECOC) 2022