Self-Consistency and Consistency-Based Detection and Diagnosis of Malicious Circuitry

Hardware Trojans (HTs) have become a major concern in the modern integrated circuit (IC) industry, especially with the fast growth in IC outsourcing. HT detection and diagnosis are challenging due to the huge number of gates in modern IC designs and the high cost of testing. We propose a scalable and efficient HT detection and diagnosis scheme based on segmentation and consistency analysis of gate-level properties. In addition, we employ a self-consistency-based approach, where we conduct variable elimination and create subsegments from a fixed set of power measurements of the entire IC to minimize the number of power measurements. We evaluate our HT detection and diagnosis schemes on a set of ISCAS and ITC benchmarks.