Abstract
We experimentally demonstrate the use of convolutional neural networks to accelerate TDECQ assessments for 400G direct-detect transmitter qualification. The method estimates TDECQ from static eye-diagrams ~1000 times faster than conventional methods with <0.25dB mean discrepancy.
© 2020 The Author(s)
PDF ArticleMore Like This
Lorenzo De Marinis, Federico Nesti, Marco Cococcioni, and Nicola Andriolli
PsTh1F.3 Photonics in Switching and Computing (PS) 2020
Mario Miscuglio, Zibo Hu, Shurui Li, Puneet Gupta, Hamed Dalir, and Volker J. Sorger
SpM5C.2 Signal Processing in Photonic Communications (SPPCom) 2021
Daniel Lippiatt, Alirio Melgar, Varghese A. Thomas, Siddharth Varughese, Pavel Zivny, Shane Hazzard, and Stephen E. Ralph
W7A.5 Optical Fiber Communication Conference (OFC) 2021