The counterfeiting and recycling of integrated circuits (ICs) have become major problems in recent years, potentially impacting the security of electronic systems bound for military, financial, or other critical applications. With identical functionality and packaging, it is extremely difficult to distinguish recovered ICs from unused ICs. A technique is proposed to distinguish used ICs from the unused ones using a fingerprint generated by a light-weight on-chip sensor. Using statistical data analysis, process and temperature variations' effects on the sensors can be separated from aging experienced by the sensors in the ICs when used in the field. Simulation results, featuring the sensor using 90nm technology, and silicon results from 90nm test chips demonstrate the effectiveness of this technique for identification of recovered ICs.

The counterfeiting of integrated circuits (ICs) has become a major issue for the electronics industry. Counterfeit ICs that find their way into the supply chains of critical applications can have a major impact on the security and reliability of those systems. This paper presents a new method for uniquely identifying ICs through path delay analysis. There is no overhead in terms of area, timing, or power for this method, since it extracts the intrinsic path delay variation information of the IC. Simulation results from 90nm technology and experimental results from 90nm FPGAs demonstrate the effectiveness of our technique.

In the past few years, several Trojan detection approaches have been developed to prevent the damages caused by Trojans, making Trojan insertion more and more difficult. As part of the Embedded Systems Challenge (ESC), we were given two different designs with two different Trojan detection methods, and we tried to design Trojans which could avoid detection. We developed Trojans that remain undetectable by the delay fingerprinting and ring-oscillator monitoring Trojan detection methods embedded into these benchmarks. Experimental results on a Xilinx FPGA demonstrate that most of our hardware Trojans were undetected using the inserted detection mechanisms.