The design of neuron-bipolar junction transistor (vBJT) cellular neural network (CNN) structure with multi-neighborhood-layer templates

A compact and efficient neuron-bipolar junction transistor (vBJT) cellular neural network (CNN) structure with multi-neighborhood-layer templates is proposed and analyzed. Using the proposed structure, the coefficients of the templates with two neighborhood layers are fully realizable. However, those with more than two neighborhood layers are constrained. As the demonstrative examples on the applications of the proposed vBJT CNNs, the functions of both de-blurring and Muller-Lyer arrowhead illusion functions have been successfully realized and verified by HSPICE simulation.