Multilevel Runtime Verification for Safety and Security Critical Cyber Physical Systems from a Model Based Engineering Perspective
Doctor of Philosophy
Electrical & Computer Engineering
Dr. Carl Elks
Advanced embedded system technology is one of the key driving forces behind the rapid growth of Cyber-Physical System (CPS) applications. CPS consists of multiple coordinating and cooperating components, which are often software-intensive and interact with each other to achieve unprecedented tasks. Such highly integrated CPSs have complex interaction failures, attack surfaces, and attack vectors that we have to protect and secure against. This dissertation advances the state-of-the-art by developing a multilevel runtime monitoring approach for safety and security critical CPSs where there are monitors at each level of processing and integration. Given that computation and data processing vulnerabilities may exist at multiple levels in an embedded CPS, it follows that solutions present at the levels where the faults or vulnerabilities originate are beneficial in timely detection of anomalies.
Further, increasing functional and architectural complexity of critical CPSs have significant safety and security operational implications. These challenges are leading to a need for new methods where there is a continuum between design time assurance and runtime or operational assurance. Towards this end, this dissertation explores Model Based Engineering methods by which design assurance can be carried forward to the runtime domain, creating a shared responsibility for reducing the overall risk associated with the system at operation. Therefore, a synergistic combination of Verification & Validation at design time and runtime monitoring at multiple levels is beneficial in assuring safety and security of critical CPS. Furthermore, we realize our multilevel runtime monitor framework on hardware using a stream-based runtime verification language.
© The Author
Is Part Of
VCU University Archives
Is Part Of
VCU Theses and Dissertations
Date of Submission