arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2009.05348eess.SPcs.LGmath.OC
收藏

基于分布式系统建模的多跨系统功率演化预测与优化

Power Evolution Prediction and Optimization in a Multi-span System Based on Component-wise System Modeling

Technical University of Denmark(丹麦技术大学)

Metodi P. Yankov, Uiara Celine de Moura, Francesco Da Ros

摘要:

在单个物理设备上训练的基于机器学习的EDFA增益模型和完全可微分的受激拉曼散射光纤模型的级联被用于预测和优化实验性多跨满载C波段光通信系统的输出处的功率分布。

英文摘要:

Cascades of a machine learning-based EDFA gain model trained on a single physical device and a fully differentiable stimulated Raman scattering fiber model are used to predict and optimize the power profile at the output of an experimental multi-span fully-loaded C-band optical communication system.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑