A synaptic circuit of a pulse-type hardware neuron model with STDP

Abstract

A number of recent studies of neural networks have been conducted with the purpose of applying engineering to the brain. We have been attempting to construct artificial neural networks. In the present study, we focus on STDP (spike timing dependent synaptic plasticity) and construct a neural network from a pulse-type hardware neuron model (P-HNM) with STDP. In addition, we propose a simple synaptic circuit of the P-HNM with STDP, which is composed of MOSFETs, and construct a simple neural network in order to verify the robustness with respect to phase fluctuation. We show that a pulse-type hardware neural network with STDP synapses has a learning function that has robustness with respect to phase fluctuation.