Statistical Model Selection for TID Hardness Assurance

Abstract

We investigate model dependence of bounding estimates or TIO degradation as a function of sample size and statistical model and develop a method for selecting the model with greatest predictive power.

System Health Awareness in Total-Ionizing Dose Environments

“…The TID delay per gate increases exponentially with TID, which is consistent with the exponential increase in leakage current shown for the PIC16 part in Fig. 6 and other studies of STI leakage [18]. In Figure 24, the exponential model is shown to fit the data, allowing comparisons between different microcontrollers and forecasts of remaining useful life in the field.…”

Statistical Modeling for Radiation Hardness Assurance: Toward Bigger Data

“…To facilitate accessibility, we have used widely available, open-access tools. Most of the plots in this section are created using the statistical coding language R [20], along with the graphics add-on ggplot2. [21].…”

Inclusion of Radiation Environment Variability in Total Dose Hardness Assurance Methodology

“…Difficulties can arise because environment uncertainties are dependent on the particular orbit, mission duration and environment dynamics. In addition the piece part failure distribution can vary significantly from a mean failure level and is not accounted for by equation 1 [18].…”