Noise Considerations in Submicron Devices

Abstract

There currently is little understanding of noise processes in very-small semiconductor devices. We consider here factors that arise in these structures due to temporal and spatial correlation of fluctuations. In this approach, time correlation functions determine macroscopic transport properties, and when energy-momentum conserving models are used as a basis for noise modeling several two-point spatial correlation functions, which describe the correlations between energy and momentum at two points in the device, are important. Ensemble Monte Carlo techniques are used to study Gauss’s law and Poisson’s equation are solved concurrently to provide an accurate relation between the position dependent carrier and field profiles. As the ensemble Monte Carlo method provides a representation of the carrier distribution function, the spatial correlations can be estimated in a straightforward fashion.