Search papers, labs, and topics across Lattice.
This paper introduces a novel method for early detection of hardware Trojans using Neural Controlled Differential Equations (NCDEs) and power trace analysis, addressing the limitations of existing techniques that only detect Trojans post-activation. By training the NCDE model on Trojan-free data, the approach effectively learns nominal power behavior and employs a Linear Discriminant Analysis (LDA) classifier to differentiate between scenarios of no Trojan, dormant Trojan, and active Trojan. Experimental results indicate that this method outperforms traditional machine learning techniques, particularly in detecting dormant Trojans, thereby enhancing security in digital systems.
Early detection of dormant hardware Trojans is now possible, significantly reducing the risk of security breaches before activation.
Evolving Hardware Trojans pose a serious threat to modern digital systems by evading traditional detection through stealthy, adaptive behavior. Even recent methods that leverage advances in machine learning can only detect them after activation, leaving a critical window for potential security breaches. To address this gap, we propose a novel approach for hardware Trojan detection and prediction using Neural Controlled Differential Equations (NCDEs) and analysis of power traces. Our method leverages an NCDE model trained exclusively on Trojan-free data to learn nominal power behavior, combined with a Linear Discriminant Analysis (LDA) classifier calibrated on labeled data, to distinguish between three scenarios: no Trojan, dormant Trojan, and active Trojan. Our method uses a sliding window to process side-channel measurements, enabling detection of subtle power consumption deviations that indicate Trojan presence, even when dormant. Experimental results demonstrate that the proposed NCDE-based method achieves superior accuracy compared to traditional machine learning approaches, with the additional advantage of handling dormant Trojans above a sensitivity threshold. We validate our approach on standard hardware Trojan benchmarks, showing robust detection and prediction performance.
