Glitch Variability Aware Power Side-Channel for Detecting Hardware Trojans
Ibrahim Farah Ahmed, Fakir Sharif Hossain · 2024
Addressing the challenge of high detection sensitivity in the presence of process variation is crucial for hardware Trojan detection through post-silicon side-channel analysis. In this work, we introduce an effective and efficient approach for Trojan detection, even in the face of increased systematic and random process variations. To enhance detection sensitivity, our method involves: i) Activating small regions to amplify Trojan-to-circuit activity, ii) comparing power levels from neighboring regions within the same chip, ensuring a consistent trend in both interdie and intra-die systematic process variations and iii) selecting random variation-tolerant test patterns by analyzing glitch effects through Monte Carlo simulation. We propose a fine-grain, low-cost circuit partitioning algorithm and suggest an evaluation method for detection sensitivity based on a glitch variation-aware detection threshold. Our approach is evaluated on ISCAS'89 benchmark circuits, encompassing both combinational and sequential Trojan types. Through an analysis of variation thresholds and experiments with Trojan-inserted circuits across diverse process variation levels, we demonstrate high detection sensitivity.