SHF: Small: System-Level Design of Attack-Resistant Safety-Critical Systems

University Of Nebraska-Lincoln, Lincoln NE

Investigators

Abstract

This project takes a design automation approach to the protection of safety/security-critical cyber-physical (CPS) and Internet-of-Things (IoT) systems against attacks originating from the Internet or from other inappropriate use. CPS and IoT systems connect sophisticated computer networks to physical world objects and machines. This combination poses new concerns over information security and physical safety than are posed by traditional information technology (IT) systems. This project will develop methodologies, algorithms, and tools for the design of safety-critical systems, including both multiprocessor systems-on-chips (MPSoCs) and networked control systems. Threats may come from information security vulnerabilities as well as safety-related faults. The methodologies and tools developed by this project will enable system designers to harden their designs against known attacks. These methods can also be used after system deployment to redesign system components to reduce their vulnerability to emergent threats. The project would also help develop new curriculum and train US workforce in the important area of information security and would be inclusive of the women, minority, and underrepresented groups in this endeavor. This project develops a new methodology for attack-aware design for safety-critical cyber-physical and Internet-of-Things (IoT) systems. A formal, multilayer model is used to describe the system. New design-time re-synthesis methods analyze the system for potential attacks and modify the system to mitigate the attacks. The methodology considers both functional attacks and timing-oriented quality-of-service attacks. Mitigations may be applied at multiple levels: middleware, operating system, and hardware. Attack mitigation is posed as an optimization problem. A set of mitigations is selected based on cost and on the number of mitigations applied for each valid computation. The methodology is applicable to networked systems at multiple scales, including both multiprocessor systems-on-chips used in smartphones and networked control systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.