Integrated Photonic Reservoir Computing with All-Optical Readout

TL;DR

This paper introduces a fully-integrated all-optical photonic reservoir computing system with an optical readout, demonstrating real-time processing and online training capabilities at 20 Gbps for complex tasks.

Contribution

It presents a novel integrated coherent photonic reservoir chip with an optical readout and a new weight update strategy for all-optical online training.

Findings

Successfully recognized 3-bit headers at 20 Gbps

Performed delayed XOR tasks in real-time

Achieved robust, fabrication variation-insensitive operation

Abstract

Integrated photonic reservoir computing has been demonstrated to be able to tackle different problems because of its neural network nature. A key advantage of photonic reservoir computing over other neuromorphic paradigms is its straightforward readout system, which facilitates both rapid training and robust, fabrication variation-insensitive photonic integrated hardware implementation for real-time processing. We present our recent development of a fully-optical, coherent photonic reservoir chip integrated with an optical readout system, capitalizing on these benefits. Alongside the integrated system, we also demonstrate a weight update strategy that is suitable for the integrated optical readout hardware. Using this online training scheme, we successfully solved 3-bit header recognition and delayed XOR tasks at 20 Gbps in real-time, all within the optical domain without excess delays.