Adder Convolutional Neural Network Equalizer for RRM-based O-band Optical Amplification-free 200 GBd OOK Transmission

Konferenz: ECOC 2024 - 50th European Conference on Optical Communication
22.09.2024-26.09.2024 in Frankfurt, Germany

Tagungsband: ITG-Fb. 317: ECOC 2024

Seiten: 4Sprache: EnglischTyp: PDF

Autoren:
Osadchuk, Y.; Li, D.; Ostrovskis, A.; Salgals, T.; Rubuls, K.; Spolitis, S.; Bobrovs, V.; Zibar, D.; Da Ros, F.; Pang, X.; Ozolins, O.

Inhalt:
We experimentally investigate the performance of adder convolutional neural network (Adder- CNN) equalization for ring modulator-based 200 GBd OOK transmission. Replacing multiplications by additions, AdderCNN outperforms the decision-feedback equalizer by 0.8 dB of RoP in 500 m transmission, showing potential for hardware implementation avoiding feedback connections.