Ph.D. Thesis: AI-Powered System-Scientific Defense for High-Confidence Cyber-Physical Systems: Modeling, Analysis, and Design

You are here

Inside Signal Processing Newsletter Home Page

Top Reasons to Join SPS Today!

1. IEEE Signal Processing Magazine
2. Signal Processing Digital Library*
3. Inside Signal Processing Newsletter
4. SPS Resource Center
5. Career advancement & recognition
6. Discounts on conferences and publications
7. Professional networking
8. Communities for students, young professionals, and women
9. Volunteer opportunities
10. Coming soon! PDH/CEU credits
Click here to learn more.

News and Resources for Members of the IEEE Signal Processing Society

Ph.D. Thesis: AI-Powered System-Scientific Defense for High-Confidence Cyber-Physical Systems: Modeling, Analysis, and Design

By: 
Linan Huang

Thesis Title: AI-Powered System-Scientific Defense for High-Confidence Cyber-Physical Systems: Modeling, Analysis, and Design
Ph.D. Thesis Author: Linan Huang | Email
Advisor: Quanyan Zhu
Granting Institution: New York University
Access Full Thesis | Presentation video

Abstract: This dissertation develops Defense through AI-powered SYstem-scientific methods (DAISY) for high-confidence Cyber-Physical Systems (CPSs).  We start by designating five generations of Security Paradigms (SPs) that have evolved since the birth of the Internet. Positioned as the foundation of 5G-SP, DAISY enables the following six dimensions of the evolution of security solutions, i.e., from empirical to theoretical, technical to socio-technical, single-agent to multi-agent, secure to resilient, add-on to built-in, and reactive to proactive security. This dissertation have a multitude of impacts in both theory and practice. First, the established models and frameworks enable quantitative design and circumvent laborious trial-and-error design procedures. Second, leveraging a broad range of system science tools and AI, this dissertation lays solid theoretical foundations to characterize fundamental limits and tradeoffs, discover security principles and laws, and design strategic security mechanisms. Finally, we develop efficient and scalable algorithms to create implementable technologies for critical CPS application fields.


 
Cyber-Physical Systems (CPSs) are smart systems that include engineered interacting networks of physical, digital, and computational components. It is critical to build high-confidence CPSs that behave in a well-understood, predictable, and justifiably trusted fashion to fulfill life-critical tasks such as medical surgery, autonomous driving, and nuclear power plant control. The status quo, however, is remote from this objective due to the challenges that arise from the distinctive features of CPSs, e.g., diversity and heterogeneity, large-scale connections and complex interdependence, human-in-the-loop, and openness and dynamic properties. Environmental (e.g., cascading failures), accidental (e.g., human errors), and deliberate threats (e.g., advanced persistent threats) have attested to the inadequacy of the off-the-shelf defense mechanisms.
 
This dissertation focuses on developing Defense through AI-powered SYstemscientific methods (DAISY) for high-confidence CPSs. To this end, the dissertation starts by delineating a brief history of security technologies by designating five generations of Security Paradigms (SPs) that have evolved since the birth of the Internet. They include the first-generation SP (1G-SP) of laissez-faire security, the 2G-SP of perimeter security, the 3G-SP of reactive security, the 4G-SP of proactive security, and the 5G-SP of federated security. DAISY addresses central challenges of the current security landscape, including lack of security standards and metrics, human-targeted and human-induced attacks, strategic and intelligent attacks, imperfect security, piecemeal design of CPS, and defenders’ time, space, information, and cooperation disadvantages. Positioned as the foundation of 5G-SP, DAISY enables the following six dimensions of the evolution of security solutions, i.e., from empirical to theoretical, from technical to socio-technical, from singleagent to multi-agent, from secure to resilient, from add-on to built-in, and from reactive to proactive security. Read more...
 
To learn more about Linan Huang, visit his LinkedIn and Google Scholar pages.
 

 

SPS on Twitter

  • DEADLINE EXTENDED: The 2023 IEEE International Workshop on Machine Learning for Signal Processing is now accepting… https://t.co/NLH2u19a3y
  • ONE MONTH OUT! We are celebrating the inaugural SPS Day on 2 June, honoring the date the Society was established in… https://t.co/V6Z3wKGK1O
  • The new SPS Scholarship Program welcomes applications from students interested in pursuing signal processing educat… https://t.co/0aYPMDSWDj
  • CALL FOR PAPERS: The IEEE Journal of Selected Topics in Signal Processing is now seeking submissions for a Special… https://t.co/NPCGrSjQbh
  • Test your knowledge of signal processing history with our April trivia! Our 75th anniversary celebration continues:… https://t.co/4xal7voFER

IEEE SPS Educational Resources

IEEE SPS Resource Center

IEEE SPS YouTube Channel