Authors:	R. Van Assche, S. Masaad, E.J.C. Gooskens, S. Sackesyn, J. Van Kerrebrouck, X. Yin, P. Bienstman
Title:	Real-time All-optical Signal Equalisation with Silicon Photonic Recurrent Neural Networks
Format:	International Conference Proceedings
Publication date:	9/2025
Journal/Conference/Book:	European Conference on Optical Communication
Volume(Issue):	p.M.03.03.2
Location:	Copenhagen, Denmark
Citations:	Look up on Google Scholar
Download:	(1.5MB)

Abstract

We experimentally demonstrate real-time all-optical equalisation of nonlinear fibre distortions
using a silicon photonic recurrent neural network. A 28 Gbps OOK signal is equalised with BER below
the GFEC threshold 5.8 × 10−5 for fibre links up to 50 km.

Related Research Topics

• Photonic Reservoir Computing

Related Projects

• H2020: NEBULA (Neuro-augmented 112Gbaud CMOS plasmonic transceiver platform for Intra- and Inter-DCI applications)
• Horizon Europe: PROMETHEUS (PROgraMmable integrated photonic nEuromorphic and quanTum networks for High-speed imaging, communications and sEcUrity applicationS)
• CHIPS-JU: Nehil 
• EIC Pathfinder: NEHO