SMOTrojan: A golden model-free GNN-SMOTE framework for automated Hardware Trojan detection via synthetic node-edge synthesis

Key Points

• Automated detection of hardware Trojans is achieved through a novel GNN-SMOTE framework, enhancing detection capabilities.
• The proposed model demonstrates a notable improvement with synthetic node-edge synthesis, increasing scanning efficiency and accuracy.
• Developed under a model-free approach, the GNN-SMOTE framework balances detection speed and precision for complex hardware systems.
• Implications highlight potential for widespread adoption in cybersecurity, addressing rising threats from hardware Trojans.