Title
Impairment Identification for PAM-4 Transceivers and Links Using Machine Learning
Abstract
We demonstrate simultaneous TDECQ estimation and impairment identification using convolutional neural networks. Robust training with representative impairments yielded identification accuracy of 100% for TDECQ >2.6 dB, when considering limited bandwidth, signal compression and SNR degradation. (C) 2021 Stephen E. Ralph
Year
Venue
DocType
2021
2021 OPTICAL FIBER COMMUNICATIONS CONFERENCE AND EXPOSITION (OFC)
Conference
Citations 
PageRank 
References 
0
0.34
0
Authors
7