Speaker: Chenglu Jin, Smart Cities Postdoctoral Associate at New York University
Title: Enhancing Cyber-Physical Systems Security with Cryptography and Hardware Security Primitives
Abstract: Since the introduction of Stuxnet, security research of industrial control systems, or cyber-physical systems (CPSs) in general, has become a rapidly growing area. In order to achieve security properties for CPS in practice, solutions based on cryptographic primitives are generally considered to be too computationally expensive. In this talk, we show that, if designed wisely, a CPS can make use of cryptographic primitives efficiently and achieve cryptographically strong security guarantees. We leverage garbled circuits and fault-tolerant algorithms for developing the first multi-sensor fusion system that can formally defend against data pollution attacks while preserving the privacy of individual sensor data. During the second part of the presentation, we address hardware security primitives (as opposed to cryptographic primitives), which can be used to enhance the security of cyber-physical systems and computer systems. We detail Physical Unclonable Functions (PUFs) and explain the full spectrum of secure PUF design ranging from a state-of-the-art silicon PUF design withstanding all known (machine learning based) attacks to the hardware/ software interfaces of PUFs.
Bio: Chenglu Jin is a smart cities postdoctoral associate at New York University. He got his Ph.D. degree and Master degree in Computer Engineering at the University of Connecticut and New York University, respectively. His research interests are in hardware security, cyber-physical system security, and applied cryptography.
Seminar Host: Degang Chen