Predictive Modeling Of Erroneous Software Behavior In Embedded Digital Systems Due To Extreme Electromagnetic Interference

Predicting the behavior of digital electronic and embedded systems subjected to extreme electromagnetic interference (EEMI) is a growing concern for military and civilian systems operating in ever denser ambient electromagnetic (EM) or Radio Frequency (RF) environments. While the interaction of the EM/RF stimulus with the system circuitry occurs at the physical or hardware level (“glitch”), the manifestation of this interaction is often observed as an erroneous behavior in the software state of the system under EM/RF stress. In this paper, we present ongoing progress in developing Markov State Vector Machine (MSVM) modeling frameworks for describing the software state of a microcontroller-based digital electronic system subjected to an externally applied glitch. We also describe our ongoing work in developing predictive MSVM models for the erroneous responses of complex software scripts when the underlying digital electronic system is subjected to EEMI glitches. Our predictive capability will help EMI/EMC engineers develop quick assessment tools for modeling the behavior of higher complexity digital electronic and embedded systems exposed to extreme electromagnetic environments.