Numerical study of the noise power of a carbon nanowire network

Abstract

A thin film made of a carbon nanowire network can be mapped into a resistor network containing tunnel junctions that are randomly switched. In such a network the variance of individual resistance is infinity so the perturbative analysis must be applied on conductances. We study the relationship between the noise power and the network morphology through a Monte Carlo simulation of the conductance and the $1/f$ noise spectrum. We find that the noise power scales with the average current in a power law $Spropto I^{-omega}$ where $omega$ is a function of the network morphology. The noise spectrum is studied in detail and we give a simple explanation for the observed relation between total noise power and the conductance.