Detection of Information-Leak Hardware Trojan in AES Cryptographic Circuits

Modern Integrated Circuits (ICs) are becoming increasingly vulnerable to malicious inclusions and alterations (hardware Trojans) due to the globalization of semiconductor design and fabrication processes, which brings in the requirement for reliable detection of hardware Trojans made in an untrusted fabrication facility during post-manufacturing test. In this paper, we present a new method for Trojan detection that calculates temporal features of IDDT and IDDQ signals, then applies mRMR (minimum-redundancy maximum-relevance) algorithm to select feature subset, and finally uses Mahalanobis distance calculation to detect Trojan circuits. Experimental results featuring an information-leak hardware Trojan inserted into a 128-bit AES cryptographic circuits on Xilinx Spartan-3E FPGAs demonstrate the simplicity and effectiveness of the proposed approach for Trojan detection.