![Sample our Computer Science journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5928/TOC-Subject-Sample-2021-Computer-Science.jpg"})
Abstract
Experiment on artificial neurons are usually carried in ideal electronic environment aimed to minimise the disturbance generated by electronic noise. On the contrary, biological neurons are actively using the noise naturally present in their environment. Synaptic processes are noise-tolerant and in some cases noise-assisted. Enhancement of spike transmission assisted by noise has been proposed for arrays of artificial neurons, to emulate behaviour of biologically-plausible spike transmission among real neurons. We employ a biophysically meaningful hardware spiking neuron model reproducing bursting discharges of action potentials to experimentally realise noise-assisted spike transmission. The circuit consists of an array of artificial neurons equipped with Gaussian noise sources capable to deliver the noise current at the inter-neuron nodes. We simulate and experimentally demonstrate the enhancement of spike transmission by tuning Gaussian noise added at the input node of hardware neurons. The circuit provides experimental demonstration of noise-assisted enhancement of spike transmission along an array of weakly coupled neurons.
Graphical Abstract
Experimental demonstration of noise-assisted spike transmission in an electronic circuit representing an array of artificial membranes/neurons. In a chain of 3 neurons coupled by inhomogeneous resistors, spikes were not transmitted over the chain (a) if no noise were give, whereas spikes were travelled over the chain when the amount of the noise was increased (b)(c).
Acknowledgements
The authors thank Raffaella Burioni (Universitá degli Studi di Parma, Italy) for useful discussions.
Notes
No potential conflict of interest was reported by the authors.