CAREER: Secure Timing Architecture for Untrusted Edge Systems

University Of Massachusetts Amherst, Amherst MA

Investigators

Abstract

A reliable notion of time has always been critical for the security of national and industrial infrastructure, such as smart grids and industrial control systems. With the emergence of human-in-the-loop systems at the network edge, such as virtual reality, autonomous vehicles, mobile health, and smart financial systems, the integrity of timing primitives is more important than ever for user security and safety. Although there have been advances in isolated domains such as trusted clock sources in shielded hardware systems, secure time synchronization protocols, and resilient high-resolution timers, realizing secure time architectures for widespread adoption in zero-trust edge environments is hindered by two key challenges. First, trusted processors do not provide secure and precise access to their timekeeping source. Second, there is no universal time transfer technique that is resilient to delay attacks in the network, and hence, existing approaches are not suitable for adoption over heterogeneous and large-scale edge deployments. The proposed work is the first effort to provide an end-to-end, secure time architecture for edge systems with diverse clocks, for timestamping and time transfer. This project is also the first to study active and passive sensing mechanisms for secure time coordination. The proposed work also investigates challenges to secure time-sensitive networking in 5G networks, distinguished by support for emerging time-critical applications at large scales. This project will advance the state-of-the-art in secure timing with new clock sources, time transfer channels, and network protocols. The proposed cross-stack approach reimagines core concepts in timing -a fundamental primitive used by all networked systems- and will create a foundation for designing safe, secure, and efficient systems. The diverse and secure clock sources will enable researchers and industry partners to support emerging applications with strict timing requirements over real-world edge deployments. The project also creates opportunities for education and outreach: The project will incorporate research on time security into new and existing graduate and advanced undergraduate classes at the PIs' home institution as part of the Cybersecurity Institute. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.