A green's function approach to the analysis of non volatile memory device variability as a function of individual trap position

This paper is aimed at exploring efficient approaches for the simulation of Random Telegraph Noise (RTN) in variability analysis of advanced floating gate non-volatile memories. RTN is traced back to randomly occupied localized traps located close to the Si/SiO₂ interface. While the effect of traps has been investigated previously by means of time-consuming Monte Carlo simulations [1], in this work we try to exploit an efficient Green Function based analysis, akin to the one implemented in Synopsys SDevice for Random Doping Fluctuations (RDF) [2].