Semiconductor Electronic Concepts for Neural Network Emulation

D. D. Coon
Department of Physics, University of Pittsburgh, Pittsburgh, PA 15260

A. G. U. Perera
Department of Physics and Astronomy, Georgia State University, Atlanta, GA 30303

ABSTRACT
The existence of neuron-like transient phenomena in simple circuits containing silicon p⁺-n-n⁺ diodes indicates that such devices could provide a natural hardware basis for the fabrication of neural networks. Development of this hardware basis is pursued through exploratory work on circuits which exhibit some basic features of biological neural networks. The features discussed include action potentials, refractory periods, excitation, inhibition, summation over synaptic inputs, synaptic weights, temporal integration, memory, network connectivity modification based on experience, pacemaker activity, firing thresholds, coupling to sensors with graded signal outputs and the dependence of firing rate on input current. © 1987 International Journal Electronics

Return to Dr. Perera's Homepage
Return to Physics & Astronomy Homepage