Graph Neural Networks (GNNs) have shown great success in learning on graph-structured data, such as social networks, recommendation systems, and protein-protein interactions. In our work, we leverage the fact that circuits can be represented naturally as graphs and we employ GNNs to learn the properties of circuits. Through our graph-based learning on circuits, we are able to identify critical security vulnerabilities in the implementations of various design-for-trust solutions aimed to achieve hardware security.
- Our paper has received the 2020 Best Paper Award for IEEE Transactions on Emerging Topics in Computing
- TinyML Talks: Energy-Efficiency and Security for TinyML and EdgeAI: A Cross-Layer Approach
- Graph Neural Networks (GNNs)
- Cybersecurity implications of Construction 4.0
- Change Your Pa55w0rd – FragAttacks: Clarifying Some Aspects