Density-based Clustering Method for Hardware Trojan Detection Based on Gate-level Structural Features

摘要

The outsourcing of IP cores raises serious concerns about the security of ICs due to the possible malicious modifications of circuits. The malicious modifications of ICs called hardware Trojans can change the system behavior, cause malfunctions of chips or leak information to a third party. This paper presents a hardware Trojan detection approach for gate-level netlists using a density-based clustering method. The unsupervised density-based clustering method can identify HTs based on the gate-level structural features, avoiding deliberate threshold setting for different features and feature information loss. We conducted experiments on the TrustHub benchmarks to verify the validity of this approach. The results demonstrate that our method can improve the accuracy rate up to 5 times compared to the existing state-of-the-art methods.