An integrated method for extracting the statistical distribution of RTN time constants

Abstract

Modelling Random Telegraph Noise (RTN) is a challenging task and its accuracy is generally unknown. The Monte Carlo modelling in the time domain requires the statistical distribution of the capture and emission time (CET) constant of traps. Although a lot of efforts were made in early works to extract the CET of individual traps, the number of traps measured is generally too limited to establish the statistical distribution of CET reliably and there are disagreements on the statistical models of CET. Two models proposed by early works are Log-normal and Log-uniform distributions, which gives very different predictions for the RTN and this difference increases as the time window becomes wider. As an accurate modelling of RTN cannot be achieved without a trustable statistical distribution of CET, it is important to find a method that allows extracting the CET distribution reliably. In contrast with early works that focus on measuring the CET of individual traps, this work proposes an integrated method for extracting the statistical distribution of CET.