Prediction of Single Event Upset-Induced Drain Peak Current Using Machine Learning

Le résumé fourni par la source

This study proposes a machine learning-based framework for predicting the drain peak current induced by a Single Event Upset (SEU) in a Buried Channel Array Transistor (BCAT) structure. Transient current responses were obtained through TCAD simulations under various radiation conditions, with linear energy transfer (LET) values ranging from 1 to$\text{1 0 0 ~ M e V} \cdot \text{cm}^{2} / \text{m g}$and trap densities from$1 \times 10^{13}$to$1 \times 10^{14} \text{~ c m}^{-3}$. These data were used to train and validate a regression-based learning model capable of quantitatively estimating the drain peak current without additional physical simulations. The trained model achieved a determination coefficient ($R^{2}$) of 0.989 and a mean squared error (MSE) of$3.29 \times 10^{-8}$, demonstrating high predictive accuracy. The machine learning model successfully captured the nonlinear interactions between LET and trap density that govern SEU-induced transient behavior. This approach provides a computationally efficient surrogate to conventional TCAD analysis, enabling rapid evaluation of radiation-induced effects in advanced DRAM devices and offering practical insights for optimizing SEU-tolerant semiconductor design.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.