Add to BibSonomyAbstract
We present an efficient scheme to upgrade DNN-based optical monitors collaboratively trained through multiple network operators without revealing each confidential data, applying federated learning with pre-model size reduction based on transferable lottery ticket hypothesis.
© 2023 The Author(s)
PDF ArticleMore Like This
Network for AI: Communication-Efficient Federated Learning with MST-based Scheduling and Multi-Aggregation over Optical Networks
Ruikun Wang, Jiawei Zhang, Memedhe Ibrahimi, Zhiqun Gu, Yuming Xiao, Francesco Musumeci, Massimo Tornatore, and Yuefeng Ji
Tu3B.3 Optical Fiber Communication Conference (OFC) 2024
High-Precision Edge-Cloud Collaboration with Federated Learning in Edge Optical Network
Chao Li, Hui Yang, Quiyan Yao, Zhengjie Sun, and Jie Zhang
W6A.11 Optical Fiber Communication Conference (OFC) 2021
Scalable Federated Learning over Passive Optical Networks
Jun Li, Lei Chen, and Jiajia Chen
W6A.36 Optical Fiber Communication Conference (OFC) 2021