Abstract
We use a neural network trained jointly by multi-task learning on datasets acquired at multiple wavelengths to mitigate the impact of chromatic dispersion in 4×200Gb/s CWDM4 PAM4 transmission. By sharing a single set of weights among all involved wavelengths, while keeping the biases reconfigurable, we enable logic simplification of multipliers in the VLSI implementation of the neural network. Results show that the neural network equalizer achieves a similar BER compared with a Volterra equalizer with 71% reduction in hardware area.
© 2023 The Author(s)
PDF Article | Presentation VideoMore Like This
Christian Bluemm, Bo Liu, Bing Li, Talha Rahman, Md Sabbir-Bin Hossain, Maximilian Schaedler, Ulf Schlichtmann, Maxim Kuschnerov, and Stefano Calabrò
Tu3C.4 European Conference and Exhibition on Optical Communication (ECOC) 2022
Georg Böcherer, Florian Strasser, Elias Arnold, Youxi Lin, Johannes Schemmel, Stefano Calabrò, and Maxim Kuschnerov
W4E.1 Optical Fiber Communication Conference (OFC) 2023
Xizi Tang, Ji Zhou, Mengqi Guo, Jia Qi, Tiantian Zhang, Zhenshan Zhang, Yueming Lu, and Yaojun Qiao
W2A.62 Optical Fiber Communication Conference (OFC) 2018