Abstract
As researchers and developers move from digital to cyber-physical systems, a gap is emerging that is revealing challenges to performance and security in many different cyber-physical system domains. In particular, cybersecurity protections in the digital domain provide desirable protection and verification mechanisms that currently have no analog in the physical domain, limiting verification capabilities in cyber-physical systems. To illustrate this gap, suppose two devices are deployed in the same physical space where they should be allowed to coordinate despite any previous physical or digital relationship. Instead of relying on risky human involvement or specialized hardware, automating the pairing process using measurable and verifiable details of common context should allow them to bootstrap a trust relationship. Our project goals to enable this verification based on context measurement will fill an important gap in cyber-physical systems.
The concept of "context fingerprinting" is offered as an approach to bridge this gap without the need for computationally intensive cryptography. It is a technique to allow two dissimilar devices to observe events,
process measurement data, and create and exchange contextual fingerprints to verify a shared property. This project explores the CPS foundations for context fingerprinting of devise. The concept will be empirically evaluated, showing how to facilitate usable and
secure bootstrapping of trust among IoT devices.
Performance Period: 01/01/2017 - 12/31/2019
Institution: Carnegie-Mellon University
Sponsor: National Science Foundation
Award Number: 1645759