Add to BibSonomyAbstract
We propose data augmentation of monitoring information using VAE and GAN to reduce the amount of required soft-failure training data. Results show that only 5 samples per failure type are needed for F1-scores above 0.9.
© 2024 The Author(s)
PDF ArticleMore Like This
GAN Based Soft Failure Detection and Identification for Long-haul Coherent Transmission Systems
Huazhi Lun, Xiaomin Liu, Meng Cai, Yiwen Wu, Mengfan Fu, Lilin Yi, Weisheng Hu, and Qunbi Zhuge
Th4J.2 Optical Fiber Communication Conference (OFC) 2021
Data Augmentation to Reduce Computational Complexity of Neural-Network-Based Soft-Failure Cause Identifier
Lareb Zar Khan, Pedro J. Freire, João Pedro, Nelson Costa, Antonio Napoli, and Nicola Sambo
M3G.3 Optical Fiber Communication Conference (OFC) 2023
Data Augmentation to Improve Machine Learning for Optical Network Failure Management
Lareb Zar Khan, Joao Pedro, Nelson Costa, Antonio Napoli, and Nicola Sambo
We4B.3 European Conference and Exhibition on Optical Communication (ECOC) 2022