Significantly Improves Detection of Hardware Trojans and Reduces Test Time
This malicious hardware detection system identifies hardware Trojans present in system-on-chips that can lead to undesired information leakage and system malfunction. System-on-chips are critical components of nearly all modern electrical systems, and the system-on-chip market should reach $205 billion globally by 2023. Insuring the integrity of system-on-chips is a crucial step in quality and security assurance of larger electronics. Current methods for detecting Hardware Trojans in system-on-chips can be time-consuming and are limited due to the high computational complexity of the methods, inability to detect stealthy Trojans, and inability to distinguish Hardware Trojan activity from background activity.
Researchers at the University of Florida have designed a malicious hardware detection system that is more effective at detecting Hardware Trojans and less time-consuming than available detection methods. This system promotes system-on-chip security and mitigates unwanted information leakage and larger system malfunction.
Application
Detects malicious Hardware Trojans in system-on-chips more accurately and more quickly than available Hardware Trojan detection systems.
Advantages
- Detects Hardware Trojans in system-on-chips, preventing information leakage and wider system malfunction
- Distinguishes Hardware Trojan activity from background activity, enabling more-accurate, rapid Hardware Trojan detection
- Uses reinforcement learning algorithm, generating Hardware Trojan detection tests rapidly and improving overall detection speeds.
Technology
This system detects malicious Hardware Trojans in system-on-chips to prevent information leakage and wider system failure. This system detects Hardware Trojans more accurately than other detection systems by using critical path analysis to increase side-channel sensitivity and better distinguish Hardware Trojan activity from background activity. This system also detects Hardware Trojans more quickly than other systems by applying reinforcement learning to rapidly and autonomously generate tests that detect Hardware Trojans using delay-based analysis.
Brochure
