A New Methodology for Two-Level Random-Telegraph-Noise Identification and Statistical Analysis

We present a two-level random telegraph noise (RTN) characterization technique which allows the estimation of RTN model parameters with statistical validity, thus acquiring a deeper insight into the physics of noise-generating process. A two-state Markov chain is used to model the stochastic behavior of RTN. In this way, an easy implementation of Monte Carlo performance and reliability simulations of large populations of electron devices can be performed. The robustness of the methodology has been investigated under different inherent noise conditions.