High-Speed Time Series Prediction with a GHz-rate Photonic Spiking Neural Network built with a single VCSEL

TL;DR

This paper demonstrates a GHz-rate photonic spiking neural network built with a single VCSEL, achieving high-speed, efficient multi-step time-series prediction, showcasing the potential for ultrafast neuromorphic photonic hardware.

Contribution

It introduces the first application of a single VCSEL-based photonic spiking neural network for complex time-series prediction at GHz rates.

Findings

Achieved accurate multi-step-ahead predictions of chaotic time-series.

Demonstrated ultrafast operation with only two inputs.

Showcased potential for hardware implementation in neuromorphic computing.

Abstract

Photonic technologies hold significant potential for creating innovative, high-speed, efficient and hardware-friendly neuromorphic computing platforms. Neuromorphic photonic methods leveraging ubiquitous, technologically mature and cost-effective Vertical-Cavity Surface Emitting Lasers (VCSELs) are of notable interest. VCSELs have demonstrated the capability to replicate neuronal optical spiking responses at ultrafast rates. These characteristics have triggered research into applying these key-enabling devices in spike-based photonic computing. Here, a GHz-rate photonic Spiking Neural Network (p-SNN) using a single VCSEL is reported, and its application to a complex time-series prediction task is demonstrated for the first time. The VCSEL p-SNN combined with a technique to induce network memory, is applied to perform multi-step-ahead predictions of a chaotic time-series. By providing the feedforward p-SNN with only two temporally separated inputs excellent accuracy is experimentally demonstrated over a range of prediction horizons. VCSEL-based p-SNNs therefore offer ultrafast, efficient operation in complex predictive tasks whilst enabling hardware implementations. The inherent attributes and performance of VCSEL p-SNNs hold great promise for use in future light-enabled neuromorphic computing hardware.