Adaptive neural network equalisation using skip connections for future 100 Gbit/s/λ passive optical networks
Loading...
Files
Accepted Version
Date
2023
Authors
Murphy, Stephen L.
Jamali, Fariba
Townsend, Paul D.
Antony, Cleitus
Journal Title
Journal ISSN
Volume Title
Publisher
Institution of Engineering and Technology
Published Version
Abstract
We present a novel equaliser based on a neural network with skip connections for 100 Gbit/s PAM4 SOA-preamplfied PONs which can converge within 2000 symbols on a burst-by-burst basis, and effectively compensate SOA patterning and 81.6 ps/nm fiber dispersion, enabling 21 dB system dynamic range.
Description
Keywords
Optical fibre dispersion , Equalisers , Optical fibre networks , Passive optical networks , Pulse amplitude modulation
Citation
Murphy, S. L., Jamali, F., Townsend, P. D. and Antony, C. (2023) 'Adaptive neural network equalisation using skip connections for future 100 Gbit/s/λ passive optical networks', 49th European Conference on Optical Communications (ECOC 2023), Glasgow, UK, 1-5 October, pp. 1422-1425. https://doi.org/10.1049/icp.2023.2580
Link to publisher’s version
Copyright
© 2023, Institution of Engineering and Technology. This paper is a postprint of a paper submitted to and accepted for publication in 49th European Conference on Optical Communications (ECOC 2023), and is subject to Institution of Engineering and Technology Copyright. The copy of record is available at the IET Digital Library.