CISE-MSI: DP: CPS: Statistical and Artificial Intelligence-based Cyberattack Detection Models for Connected Vehicles

Benedict College, Columbia SC

Investigators

Abstract

This award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2). This project is also jointly funded by the CISE MSI Research Expansion Program and the Established Program to Stimulate Competitive Research (EPSCoR). This research is motivated by the necessity to promote U.S. competitiveness and develop talented and skilled young professionals who have in-depth cybersecurity knowledge, focusing on connected and automated vehicles and implementing these methods using advanced computational techniques. Solving tomorrow’s cyberattacks on connected vehicles requires today’s students to learn in multidisciplinary teams about the evolving cyber threats and think broadly about detecting such attacks, which will change daily. This project’s vision is to transform programs at Benedict College using an integrated and scalable approach to produce students capable of innovating in non-traditional, interdisciplinary teams to solve the unique cybersecurity problems. It develops and evaluates detection models, focusing on cyberattacks in the in-vehicle systems and the wireless network connecting the vehicles with other vehicles, infrastructure, and services. The transportation cyber-physical systems consist of mobile edges (e.g., connected vehicles), fixed edge devices, backend servers, and in-vehicle systems, which run different safety, mobility, and environmental applications. Due to the application requirements, the edge devices exchange real-time, heterogeneous data with each other and with the backend servers through different communication options. Attacks on connected vehicles applications can have a significant impact on public safety. For example, once the malware is injected, a critical safety application, such as the collision warning, can malfunction and result in catastrophic consequences. Detecting cybersecurity threats in real-time is challenging because of the dynamic behavior of such attacks, especially in a connected vehicle environment where the vehicles and pedestrians are in motion. To address these, change-point models can potentially detect anomalies in real-time. Also, data-driven artificial intelligence-based models can detect these attacks as these models are adaptable to different attack types with known and unknown patterns. The project tasks include the development and comparison of efficacies of the real-time cyberattack detection strategies, which are based on change-point and artificial intelligence models. 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.