Monte Carlo analysis of noise spectra in self-switching nanodiodes

Abstract

By means of a semiclassical two-dimensional Monte Carlo technique, we analyze current noise spectra of InAlAs/InGaAs-based submicron self-switching diodes. Shot noise (at low bias) and diffusion noise in the series resistance (at high bias) are found to dominate the current noise at lowfrequency (in the plateau beyond the 1/ f range). Two peaks of different origins, which may limit the device performance, are found in the noise spectra at higher frequencies. The dependence of theamplitude and frequency of these peaks on the topology of the diodes is analyzed and discussed. Design indications to improve the noise performance of the devices are provided.