A solution for on-line trust validation

Abstract

Summary form only given. We show that pre-deployment checks are unlikely to guarantee the detection of Trojans using a time-bomb mechanism for activation. We describe a new approach to on-line TRUST validation based on extensions to an existing technology for in-system silicon validation. The basis is a reconfigurable infrastructure platform added to the RTL model of the target SoC. The platform can be configured to implement design-for-enabling-security (DEFENSE) logic consisting of security monitors and logic to provide countermeasures to detected attacks.