학술논문
Self-learning perceptron using a digital memristor emulator
Document Type
Conference
Source
2019 8th International Conference on Modern Circuits and Systems Technologies (MOCAST) Modern Circuits and Systems Technologies (MOCAST), 2019 8th International Conference on. :1-4 May, 2019
Subject
Language
Abstract
In this work we present a neuron model using a digital memristor emulator that mimics the spike time dependent plasticity (STDP). The emulator has been implemented in VHDL, and it is completed with a digital circuitry to obtain a leaky integrate and fire spiking neuron model (LIF). Our model presents the principal characteristics of a biological neuron: inhibitory and excitatory synapses, synaptic weights, excitation threshold and refraction period. The results show that our final circuit has the capability to reproduce the OR function after the time required to do the self-learning process.