Talk: Safeguarding CPS/IoT Systems in the AI Era

As Cyber-Physical Systems (CPS) and Internet of Things (IoT) devices become increasingly integrated into smart homes, smart grids, and smart cities, they generate vast amounts of network traffic data. Unfortunately, this data can expose sensitive user information, making it a target for AI-powered privacy attacks. Existing defense mechanisms often fall short in addressing these evolving threats. In this talk, I will present a data-driven, privacy-friendly IoT management framework designed to help users understand and mitigate information leakage risks. First, I will introduce IoT Traffic Exposure Monitoring Toolkit (ITEMTK)—a low-cost, open-source system that includes AI-based privacy attack models to uncover vulnerabilities in CPS/IoT systems. Next, I will discuss my recent advancements in traffic reshaping techniques, including PrivacyGuard, VoiceAttack, and PAROS, which provide novel, low-cost, and distributed solutions to enhance user privacy. Finally, I will outline future research directions in developing next-generation defense mechanisms to secure CPS/IoT systems against emerging AI-driven threats.

 

Dong Chen is a tenure-track Assistant Professor in the Department of Computer Science, Colorado School of Mines (Mines). He received his Ph.D. in Electrical and Computer Engineering from the University of Massachusetts Amherst in 2018, and Ph.D. in Computer Science in 2014 from Northeastern University, China. His research lies at the intersection of Cyber-Physical Systems, the Internet of Things, AI@Edge, Distributed Machine Learning, and Security and User privacy. He directs Cyber-Physical Systems Laboratory at Mines, where he conducts experimental CPS system research with a focus on improving Cybersecurity, User Privacy and Sustainability of Cyber-Physical Systems and the Internet of Things at different system scales. His research has been published in top-tier CPS/IoT conferences and journals, including IPSN, BuildSys, ICDCS, PerCom, IoTDI, EWSN, CNS, BigData, SECON, MASS, ICCCN, TOSN, TMC, TIOT, TOIT, etc. He is the recipient of the Best Paper Award at ACM BuildSys’20 and was recognized as the Best Paper Runner-Up at ACM EWSN’24. Specially, on his recent CPS/IoT privacy systems research, he received the NSF CAREER Award from the Computer Systems Research program in the Division of Computer and Network Systems in 2023. He has also served on the organizing committee (such as General Chair and Publications Chairs) and technical program committee member for multiple ACM/IEEE CPS/IoT leading conferences and the organizing chair for two IEEE CPS workshops. His recent research has been supported by NSF CSR Program, NSA, Cyber Florida Seed Program, NSF RET Program, Google Education, NOAA, and NEXUS Seed Grants.