Spiking Excitable Semiconductor Laser as Optical Neurons: Dynamics, Clustering and Global Emerging Behaviors

Abstract : Within the framework of this two-year project, detailed experimental and numerical studies have been performed, focusing on the interplay of noise and nonlinear dynamics. Specifically, we used a method of time-series analysis, referred to as symbolic ordinal analysis, to demonstrate that serial correlations present in the output intensity of a semiconductor laser with optical feedback operating in the low-frequency fluctuations regime share common features with serial correlations present in the inter-spike-intervals (ISIs) of biological neuronal systems. The symbolic dynamics underlying the sequence of inter-dropout-intervals in the laser intensity has the same statistical features, in terms of distribution of symbolic patterns, as in ISI sequences of biological neurons. Therefore, semiconductor laser-based optical neurons could provide a novel, inexpensive and controllable experimental set up that could allow for improving our understanding of neuronal activity. By establishing a direct connection between these different dynamical systems our research offers new perspectives, both, in photonics and in neuroscience. For example, the optical setup could be used to analyze the role of external forcing in neuronal spike sequences. On the hand, optical neurons constructed from inexpensive semiconductor lasers could lead to the development of novel neuro-inspired optical computing devices (threshold detectors, logic gates, signal recognition, etc.).