Modeling of Neuron Based on Single Electron Transistor
Peng Qiu, Guanglong Wang, Jianglei Lu, Shuang Lei Feng · 2009
As the basic unit of cell neural network (CNN), neuron has been the focus of the research of CNN. This paper introduces a modeling method of neuron that is based on single electron transistor (SET). SET is a kind of nano electronic device, which has come to be considered candidate as the basic element for future low power, high density integrated circuits, and the quantum effect of SET shows two basic characteristics: Coulomb oscillation and Coulomb blockade. According to the nonlinear equation of neuron, a neuron can be divided into three modules: cell module, feedback template module and control template module. An equivalent structure of neuron based on these three modules is put forward. The modeling circuits that match the requirement of neuron equations are designed with SETs in the end of this paper.