Probabilistic Modelling of Variation in FGMOSFET

Abstract

The analytical probabilistic modelling of random variation in the drain current of a Floating-Gate MOSFET (FGMOSFET) induced by manufacturing process variations has been performed. Both triode and saturation region operated FGMOSFETs have been considered. The results have been found to be very efficient since they can accurately fit the probabilistic distributions of normalized random drain current variations of the candidate triode and saturation FGMOSFETs obtained using the 0.25μm level BSIM3v3 based Monte-Carlo SPICE simulations, where the variation of the saturation FGMOSFET has been found to be more severe. These results also satisfy the goodness of fit test at a very high level of confidence and more accurately than the results of the previous probabilistic modelling attempts. Using our results, many statistical parameters, probabilities and the objective functions, which are useful in statistical/variability aware analysis and design involving FGMOSFETs can be formulated. The impact of drain current variation upon the design trade-offs can be studied. It has been found that the occurrence of the drain current variation is absolutely certain. Moreover, the analytical probabilistic modelling and computationally efficient statistical/ variability aware simulation of FGMOSFET based circuits can also be performed.