Coupled lasers and neuromorphic photonics

Dynamics of coupled lasers has been studied since many years providing a rich phenomenology and interesting physics to be understood. Indeed the study of many coupled nonlinear oscillators exhibits a vast range of interesting phenomena including synchronization, chimera states and more.
Neuromorphic photonics is an emerging paradigm which tries to reproduce optically the dynamics of brain especially promoting brain-inspired computational and information processing models but also studying optical analogous of the type of cell constituting brain: the neuron.
Indeed one key feature of the neuron is excitability: the neuron transmits an electric signal (it is excited) only when, it is triggered by an input signal/perturbation that is strong enough: its response has a threshold. Furthermore the magnitude of the neuron response (the amplitude of the signal which propagates along the axon when the neuron is excited) is also independent or weakly dependent on the magnitude of the input stimulus (provided of course that the threshold is crossed as mentioned before).

A variety of optical systems, such as lasers and amplifiers exhibit excitability in certain particular configurations. An example, is the laser with saturable absorber: let’s consider the laser below threshold, then in this case no light is emitted because of cavity losses and absorber dissipating all the energy from spontaneous emission. Now let’s imagine that some perturbation act on the laser (fluctuations injected from outside of the laser cavity) then if the perturbation is strong enough to survive the losses, it will grow very fast: it will stimulate the lasing process depleting the gain: a giant spike-like pulse will be emitted. After that, the laser has to wait a bit until the upper state of the atomic transition is repopulated, such time interval of quiet is called the refractory time. After that the process can repeat…

I’m interested in investigating collective properties of populations of coupled lasers in particular (why not?) when they are excitable. Excitable lasers exhibit interesting dynamics which may sound predictable such as synchronization, but also some unexpected ones like array enhanced coherence resonance (the increment of a single oscillator regularity or coherence proportionally to the number of coupled elements in the population) and spatial localization of excitability -à la Anderson but with dissipation- in the array mediated by disorder (randomness in the respective coupling strength).
Engineering of the couplings opens up new ways to be explored both in the excitable regime and in the non excitable one.

My works in this field: