Performance of delay-based Trojan detection techniques under parameter variations

Various schemes utilizing path delay information have been proposed to determine the authenticity of a given integrated circuit. However, the performance of these techniques under parameter variations must be evaluated, as delay variations may mask the delay impact of design alterations, such as hardware Trojan horses. This paper examines how path delay characterization performs under extensive delay variations, and results reveal that it remains a powerful tool for detecting design alterations by leveraging statistical techniques.