Securing the Timing of Cyber-Physical Systems
pdf
This project aims to build a framework for identifying, analyzing and defending against timing attacks in cyber-physical systems. Timing attacks, where attackers attempt to compromise the timing of computation or communication operations, can be particularly destructive in a CPS because the correctness of system functionality is affected not only by the data values of operations but also by the time those operations are conducted. To address these types of attacks in building the proposed framework, this project includes three closely-related research thrusts: (1) Thrust A: identifying and analyzing timing-based attack surface and strategies; (2) Thrust B: conducting cross-layer analysis of the impact of timing attacks on system-level properties; and (3) Thrust C: designing and implementing control-based and cyber-security defense techniques for timing attacks. By addressing critical yet little-studied timing-based security challenges for CPS, this project should discover new timing attack surface and threat models, develop novel cross-layer methodologies for analyzing the impact of timing attacks on system properties, and correspondingly develop novel run-time detection and mitigation techniques as well as design-time strategies.
The project focuses on two important CPS application domains, vehicle networks and multi-agent robotic systems (e.g., a group of UAVs or ground robots). Systems in these two domains can both be viewed as multi-agent intelligent systems and share some common elements. Yet, they typically use different computation and communication components and have different algorithms running at application level. The project will address the specific challenges in both domains, and also evaluate the general applicability of our methodology to other CPS domains as well.
In addition to disseminate our results through publications, workshops and release of the framework, we are closely collaborating with our industry partners, in particular the Toyota InfoTechnology Center, on applying our findings to practical industrial systems. The research findings will also be integrated into the undergraduate and graduate curriculum at Northwestern University and University of California, Riverside.
Submitted by Qi Zhu
on
This project aims to build a framework for identifying, analyzing and defending against timing attacks in cyber-physical systems. Timing attacks, where attackers attempt to compromise the timing of computation or communication operations, can be particularly destructive in a CPS because the correctness of system functionality is affected not only by the data values of operations but also by the time those operations are conducted. To address these types of attacks in building the proposed framework, this project includes three closely-related research thrusts: (1) Thrust A: identifying and analyzing timing-based attack surface and strategies; (2) Thrust B: conducting cross-layer analysis of the impact of timing attacks on system-level properties; and (3) Thrust C: designing and implementing control-based and cyber-security defense techniques for timing attacks. By addressing critical yet little-studied timing-based security challenges for CPS, this project should discover new timing attack surface and threat models, develop novel cross-layer methodologies for analyzing the impact of timing attacks on system properties, and correspondingly develop novel run-time detection and mitigation techniques as well as design-time strategies.
The project focuses on two important CPS application domains, vehicle networks and multi-agent robotic systems (e.g., a group of UAVs or ground robots). Systems in these two domains can both be viewed as multi-agent intelligent systems and share some common elements. Yet, they typically use different computation and communication components and have different algorithms running at application level. The project will address the specific challenges in both domains, and also evaluate the general applicability of our methodology to other CPS domains as well.
In addition to disseminate our results through publications, workshops and release of the framework, we are closely collaborating with our industry partners, in particular the Toyota InfoTechnology Center, on applying our findings to practical industrial systems. The research findings will also be integrated into the undergraduate and graduate curriculum at Northwestern University and University of California, Riverside.