The research of operation accuracy of a memristor-based artificial neural network with an input signal containing noise and pulse interference

Abstract

This article looks at the issues of calculating the operation accuracy of memristor-based hardware when there are some pulse interference (ε-polluted interference with additive white Gaussian noise) in a square signal. Through the example of operation of the memristor-based artificial neural network synapse, you can see that this type of interference in an input signal of memristor-based hardware causes additional error in the values of their output parameters. It was revealed that there is correlation dependence between the values of parameters of noise components in an input signal of a memristor-based artificial neural network (noise and interference variance, an occurrence probability of pulse interference) and the value of synaptic weight.