A Novel Framework to Introduce Hardware Trojan Monitors using Model Checking Based Counterexamples: Inspired by Game Theory

Abstract

Interest in hardware Trojan detection has grown tremendously over the last decade. Several methods for detecting hardware Trojan have been proposed. Due to enormity of possible Trojans, researchers believe that runtime detection is required as a last line of defense. However, to cater different types of hardware Trojans, several detection techniques should be concurrently applied. In this paper we propose a framework for optimizing runtime hardware Trojan detection monitors. We propose a formal verification approach to identify the vulnerable behavior of the hardware using the counterexamples generated by model checker. We devised a heuristic to understand the relationship between the counterexamples and required detection unit. To optimize the number of detection techniques, we leveraged Game Theoretic models to obtain a set of optimized design choices.