A memristive model of spiking neuron

Compared to traditional artificial neural networks, spiking neural networks (SNNs) have a better energy and computational efficiencies in applications of the autonomy of machines in the real world. Memristors, with memory function and biological dynamics, are emerging building blocks of spiking neuron. In order to prepare for further study on spiking neural networks at the hardware level, here, we propose a memristive model of spiking neuron. The model reaches the leakage, integration, and firing functions of Leaky integrate-and-fire (LIF) spiking neuron, which satisfies the demand of SNNs. This work contributes to the development of fully memristive SNNs.