Excitable interplay between lasing quantum dot states

Refractory times for excitable dual-state quantum dot laser neurons

Excitable photonic systems show promise for ultrafast analog computation, several orders of magnitude faster than biological neurons. Optically injected quantum dot lasers display several excitable mechanisms with dual-state quantum lasers recently emerging as true all-or-none excitable artificial neurons. For use in applications, deterministic triggering is necessary and this has previously been demonstrated in the literature. In this work we analyze the crucially important refractory time for this dual-state system, which defines the minimum time between distinct pulses in any train.

Inhibitory and excitatory integration with a quantum dot laser neuron

Neuromorphic computing has garnered a lot of attention in recent years. Excitable photonic systems in particular demonstrate great potential for ultrafast, controllable spike processing. Optically injected quantum dot lasers display several distinct excitable regimes. We demonstrate here that optically injected dual-state quantum dot lasers can display the classic leaky integrate-and-fire mechanism where the integration of several sub-threshold perturbations can yield an effective supra-threshold perturbation. Intriguingly, a contrasting integrate-and-inhibit mechanism is demonstrated in this work where the integration of two supra-threshold perturbations yields an effective sub-threshold perturbation similar to the pre-pulse inhibition mechanism of biological neurons. This is the first such mechanism in neuromorphic photonics to the best of our knowledge.

Ultrafast neuromorphic photonic image processing with a VCSEL neuron

The ever-increasing demand for artificial intelligence (AI) systems is underlining a significant requirement for new, AI-optimised hardware. Neuromorphic (brain-like) processors are one highly-promising solution, with photonic-enabled realizations receiving increasing attention. Among these, approaches based upon vertical cavity surface emitting lasers (VCSELs) are attracting interest given their favourable attributes and mature technology. Here, we demonstrate a hardware-friendly neuromorphic photonic spike processor, using a single VCSEL, for all-optical image edge-feature detection. This exploits the ability of a VCSEL-based photonic neuron to integrate temporally-encoded pixel data at high speed; and fire fast (100 ps-long) optical spikes upon detecting desired image features. Furthermore, the photonic system is combined with a software-implemented spiking neural network yielding a full platform for complex image classification tasks. This work therefore highlights the potential of VCSEL-based platforms for novel, ultrafast, all-optical neuromorphic processors interfacing with current computation and communication systems for use in future light-enabled AI and computer vision functionalities.

Mapping the Stability and Dynamics of Optically Injected Dual State Quantum Dot Lasers

Optical injection is a key nonlinear laser configuration both for applications and fundamental studies. An important figure for understanding the optically injected laser system is the two parameter stability mapping of the dynamics found by examining the output of the injected laser under different combinations of the injection strength and detuning. We experimentally and theoretically generate this map for an optically injected quantum dot laser, biased to emit from the first excited state and optically injected near the ground state. Regions of different dynamical behaviours including phase-locking, excitability, and bursting regimes are identified. At the negatively detuned locking boundary, ground state dropouts and excited state pulses are observed near a hysteresis cycle for low injection strengths. Higher injection strengths reveal μs duration square wave trains where the intensities of the ground state and excited state operate in antiphase. A narrow region of extremely slow oscillations with periods of several tens of milliseconds is observed at the positively detuned boundary. Two competing optothermal couplings are introduced and are shown to reproduce the experimental results extremely well. In fact, the dynamics of the system are dominated by these optothermal effects and their interplay is central to reproducing detailed features of the stability map.

Optical information processing using dual state quantum dot lasers: complexity through simplicity

We review results on the optical injection of dual state InAs quantum dot-based semiconductor lasers. The two states in question are the so-called ground state and first excited state of the laser. This ability to lase from two different energy states is unique amongst semiconductor lasers and in combination with the high, intrinsic relaxation oscillation damping of the material and the novel, inherent cascade like carrier relaxation process, endows optically injected dual state quantum dot lasers with many unique dynamical properties. Particular attention is paid to fast state switching, antiphase excitability, novel information processing techniques and optothermally induced neuronal phenomena. We compare and contrast some of the physical properties of the system with other optically injected two state devices such as vertical cavity surface emitting lasers and ring lasers. Finally, we offer an outlook on the use of quantum dot material in photonic integrated circuits.

Triple point of synchronization, phase singularity, and excitability along the transition between unbounded and bounded phase oscillations in a forced nonlinear oscillator

We report the discovery of a codimension-two phenomenon in the phase diagram of a second-order self-sustained nonlinear oscillator subject to a constant external periodic forcing, around which three regimes associated with the synchronization phenomenon exist, namely phase-locking, frequency-locking without phase-locking, and frequency-unlocking states. The triple point of synchronization arises when a saddle-node homoclinic cycle collides with the zero-amplitude state of the forced oscillator. A line on the phase diagram where limit-cycle solutions contain a phase singularity departs from the triple point, giving rise to a codimension-one transition between the regimes of frequency unlocking and frequency locking without phase locking. At the parameter values where the critical transition occurs, the forced oscillator exhibits a separatrix with a π phase jump, i.e., a particular trajectory in phase space that separates two distinct behaviors of the phase dynamics. Close to the triple point, noise induces excitable pulses where the two variants of type-I excitability, i.e., pulses with and without 2π phase slips, appear stochastically. The impacts of weak noise and some other dynamical aspects associated with the transition induced by the singular phenomenon are also discussed.

Asymmetric excitable phase triggering in an optically injected semiconductor laser

One of the defining characteristics of excitability is the existence of an excitable threshold: the minimum perturbation amplitude necessary to produce an excitable response. We analyze an optically injected dual state quantum dot laser, previously shown to display a dual state stochastic excitable dynamic. We show that deterministic triggering of this dynamic can be achieved via optical phase perturbations. Further, we demonstrate that there are in fact two asymmetric excitable thresholds in this system corresponding to the two possible directions of optical phase perturbations. For fast enough perturbations, an excitable interval arises, and there is a limit to the perturbation amplitude, above which excitations no longer arise, a phenomenon heretofore unobserved in studies of excitability.

Image edge detection with a photonic spiking VCSEL-neuron

We report both experimentally and in theory on the detection of edge features in digital images with an artificial optical spiking neuron based on a vertical-cavity surface-emitting laser (VCSEL). The latter delivers fast (< 100 ps) neuron-like optical spikes in response to optical inputs pre-processed using convolution techniques; hence representing image feature information with a spiking data output directly in the optical domain. The proposed technique is able to detect target edges of different directionalities in digital images by applying individual kernel operators and can achieve complete image edge detection using gradient magnitude. Importantly, the neuromorphic (brain-like) spiking edge detection of this work uses commercially sourced VCSELs exhibiting responses at sub-nanosecond rates (many orders of magnitude faster than biological neurons) and operating at the important telecom wavelength of 1300 nm; hence making our approach compatible with optical communication and data-centre technologies.

Optical feedback induced oscillation bursts in two-state quantum-dot lasers

We investigate the impact of short optical feedback on a two-state quantum dot laser. A region in the feedback parameter space is identified, where the laser emission periodically alternates between oscillation bursts from the quantum dot ground and excited state, i.e. two-color anti-phase oscillation bursts. We compare these results to the low-frequency fluctuations and regular pulse packages of single-color semiconductor lasers and show via an in-depth bifurcation analysis, that the two-color oscillation bursts originate from a torus-bifurcation of a two-state periodic orbit. A cascade of further period-doubling bifurcations produces chaotic dynamics of the burst envelope. Our findings showcase the rich dynamics and complexity, which can be generated via the interaction of electronic and photonic time scales in quantum dot lasers with optical feedback.