A Coupled Schmitt Trigger Oscillator Neural Network for Pattern Recognition Applications

Abstract

This paper demonstrates a coupled Schmitt trigger oscillator based oscillator neural network (SMT-ONN) for pattern recognition applications. Unlike previous ONN models, the SMT-ONN can be easily realized in both hardware and software levels. A mathematical model of the Schmitt Trigger Oscillator as well as the corresponding CMOS circuit are presented to validate the mathematical model. The SMT-ONN can realize the pattern recognition task by considering the convergence time and frequency as the recognition indicators. A Kuramoto model based frequency synchronization approach is utilized, and simulation results indicate less than 160 ms convergence time and close frequency match for a simplified pattern recognition application.