EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Hybrid Explainable Edge Intelligence Using Electromagnetic Interference Prediction for Signal Integrity

A. N. Gnana Jeevan

Journal of Intelligent Decision Making and Information Science · 10.10.2026 · 3 (16s) · 54-70

DOI: 10.17762/jidmis.v3.4883 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Intelligent Decision Making and Information Science, online erschienen am 10.10.2026. Verfasst von A. N. Gnana Jeevan. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

High speed electronics systems or highly populated printed circuit boards (PCBs) pose many challenges in predicting electromagnetic interference (EMI) and assessing its impact on signal integrity (SI) challenge with low computational latency and sufficiently interpretable. Most existing solutions focus on one of the following aspects of EMI prediction, SI monitoring, explainability or edge intelligence, with little integration of these features in a single framework. A Hybrid Explainable Edge Intelligence (EW) framework for EMI prediction and SI assessment is developed in this study, where an EMI measurement is combined with circuit simulation, data preprocessing, Random Forest (RF), Long Short-Term Memory (LSTM) artificial intelligence, SHAP based explainability, and edge deployment. To develop a dataset of electrical, operating and PCB parameters, simulation and experimental measurements are combined, followed by 70/15/15 partition of training, validation and testing sets, respectively, and median imputation, Min-Max normalization, feature engineering. The hybrid RF-LSTM model gives an RMSE of 2.18 dBµV, which is 30.8% and 22.7% lower than that of RF and LSTM, respectively. Switching frequency is the key factor for SHAP with positive value of +5.2 dBµV followed by PCB length (+3.4 dBµV) and input voltage (+2.1 dBµV). The edge pipeline gives a total inference time of 5.5ms and the jitter slowly increased from 0.17ns at +40dBµV to 0.87ns at +80dBµV. The framework is therefore designed to incorporate correct prediction, interpretable EMI analysis, and low latency edge-based SI monitor – enabling intelligent electronic-design verification and real-time EMI-aware SI assessment.

Zitieren

APA

Jeevan, A. N. G. (2026). Hybrid Explainable Edge Intelligence Using Electromagnetic Interference Prediction for Signal Integrity. Journal of Intelligent Decision Making and Information Science, 3(16s), 54-70. https://doi.org/10.17762/jidmis.v3.4883

Vancouver

Jeevan ANG. Hybrid Explainable Edge Intelligence Using Electromagnetic Interference Prediction for Signal Integrity. Journal of Intelligent Decision Making and Information Science. 2026;3(16s):54-70. doi:10.17762/jidmis.v3.4883

BibTeX
@article{Jeevan2026,
  author = {Jeevan, A. N. Gnana},
  title = {Hybrid Explainable Edge Intelligence Using Electromagnetic Interference Prediction for Signal Integrity},
  journal = {Journal of Intelligent Decision Making and Information Science},
  year = {2026},
  volume = {3},
  number = {16s},
  pages = {54--70},
  doi = {10.17762/jidmis.v3.4883}
}
Kostenlos starten →