Abstract
Neuron and synapse are the basic constitutional units of biological neural network which has high efficiency in information processing. On account of which, the bio-inspired neuromorphic computing has attracted considerable attention. In this letter, we demonstrate an integration and fire artificial neuron (neuristor) based on threshold switching memristor. The neuristor captures 4 crucial features for action-potential-based computing: all-or-nothing spiking of an action potential, threshold-driven spiking, a refractory period, and the strength-modulated frequency response. A system-level simulation of the neuristor as post-synaptic neuron show that the validity to be used for a pattern recognition application. The experimental and simulation result exhibit that the neurisor can implement the basic functions of spiking neurons and has a great potential for neuromorphic computing.
© 2017 Optical Society of America
PDF ArticleMore Like This
Weikang Zhang, Matěj Hejda, Qusay Raghib Ali Al-Taai, Bruno Romeira, José Figueiredo, Edward Wasige, and Antonio Hurtado
jsiii_5_5 European Quantum Electronics Conference (EQEC) 2023
Rui Yang
ISu4B.3 Information Storage System and Technology (ISST) 2017
Yuchao Yang
ISu4B.1 Information Storage System and Technology (ISST) 2017