Tunable presynaptic weighting in optoelectronic spiking neurons built with laser-coupled resonant tunneling diodes

Abstract

Optoelectronic spiking neurons are regarded as highly promising systems for novel light-powered neuromorphic computing hardware. Here, we investigate an optoelectronic (O/E/O) spiking neuron built with an excitable resonant tunnelling diode (RTD) coupled to a photodetector and a vertical-cavity surface-emitting laser (VCSEL). This work provides the first experimental report on the control of the amplitude (weighting factor) of the fired optical spikes directly in the neuron, introducing a simple way for presynaptic spike amplitude tuning. Notably, a very simple mechanism (the control of VCSEL bias) is used to tune the amplitude of the spikes fired by the optoelectronic neuron, hence enabling an easy and high-speed option for the weighting of optical spiking signals in future interconnected photonic spike-processing nodes. Furthermore, we validate the feasibility of this layout using a simulation of a monolithically-integrated, RTD-powered, nanoscale optoelectronic spiking neuron model, confirming the system’s potential for delivering weighted optical spiking signals at very high speeds (GHz firing rates). These results demonstrate the high degree of flexibility of RTD-based artificial optoelectronic spiking neurons and highlight their potential towards compact, high-speed and low-energy photonic spiking neural networks for use in future, light-enabled neuromorphic hardware.

VCSEL-based photonic spiking neural networks for ultrafast detection and tracking

“…phase-change materials [5], optical modulators [6,7], and semiconductor lasers [8][9][10][11], are a few highlight photonic devices that have achieved neuromorphic or neuromimetic operation. Similarly, electro-optic technologies such quantum resonant tunnelling structures [12,13], amongst others (see [14] for a review), have now also been proposed as suitable device platforms for brain-inspired computing hardware components. With one of the major benefits of photonics being the highly-reduced energy requirements, the recent demonstrations of photonic artificial neural networks (pANNs) [15][16][17][18][19] and reservoir computing systems [20][21][22][23] make an exciting case for photonics being the platform at the forefront of potentially sustainable neuromorphic computing.…”