Abstract

Reservoir computing is a brain-inspired approach for information processing, well suited to analog implementations. We report a photonic implementation of a reservoir computer that exploits frequency domain multiplexing to encode neuron states. The system processes 25 comb lines simultaneously (i.e., 25 neurons), at a rate of 20 MHz. We illustrate performances on two standard benchmark tasks: channel equalization and time series forecasting. We also demonstrate that frequency multiplexing allows output weights to be implemented in the optical domain, through optical attenuation. We discuss the perspectives for high-speed, high-performance, low-footprint implementations.

© 2022 Optica Publishing Group

Deep photonic reservoir computer based on frequency multiplexing with fully analog connection between layers

Alessandro Lupo, Enrico Picco, Marina Zajnulina, and Serge Massar
Optica 10(11) 1478-1485 (2023)

Multi-task photonic time-delay reservoir computing based on polarization modulation

Long Huang and Jianping Yao
Opt. Lett. 47(24) 6464-6467 (2022)

Optical phase encoding in a pulsed approach to reservoir computing

Johan Henaff, Matthieu Ansquer, Miguel C. Soriano, Roberta Zambrini, Nicolas Treps, and Valentina Parigi
Opt. Lett. 49(8) 2097-2100 (2024)

Adaptive time-delayed photonic reservoir computing based on Kalman-filter training

Jiaoyang Jin, Ning Jiang, Yiqun Zhang, Weizhou Feng, Anke Zhao, Shiqin Liu, Jiafa Peng, Kun Qiu, and Qianwu Zhang
Opt. Express 30(8) 13647-13658 (2022)

Integrated photonic reservoir computing based on hierarchical time-multiplexing structure

Hong Zhang, Xue Feng, Boxun Li, Yu Wang, Kaiyu Cui, Fang Liu, Weibei Dou, and Yidong Huang
Opt. Express 22(25) 31356-31370 (2014)