A Verification Framework For Trojan Detection (U. of Kansas, U. of Florida)
Source
Published
TL;DR
AI GeneratedResearchers from the University of Kansas and the University of Florida have published a technical paper titled "COVERT: Trojan Detection in COTS Hardware via Statistical Activation of Microarchitectural Events." The paper addresses the challenge of detecting hardware Trojans in Commercial Off-The-Shelf (COTS) components, which are often used in system design but can be susceptible to malicious logic. The COVERT framework aims to efficiently test for hardware Trojan implants in COTS microprocessors by identifying rare microarchitectural events and leveraging Large Language Models to generate test programs. The framework has been evaluated on open-source RISC-V COTS microprocessors, demonstrating high coverage in activating Trojan triggers and verifying the trustworthiness of these components.