Cyber security in Dynamic Navy Multiple Agent Vehicle Networks
Abstract
The current trend towards incorporating powerful sensors and communications to form vehicle networks has tremendous potential to imp,rove the effectiveness of Naval operations. Un-fortunately, this trend also increases the potential for cyber attacks which is a tre,mendous problem for the Navy. This project hopes to show that incorporating powerful sensors and communications to form vehicle netw,orks can actually provide greatly enhanced cybersecurity if these resources are used properly. While autonomous vehicles, are typica,lly tested for deployment just by driving them, this testing alone will not provide suitable information on cyber-attack vulnerabili,ty and it appears infeasible to generate all possible attacks. Attacks on vehicles have already been observed. Developing the theory, of the impact and mitigation of cyber-attacks on networks of autonomous and human driven vehicles is critical and urgent and furthe,r study is greatly needed for the Navy and the general public.The proposed project will seek to develop theory and algorithms for ne,ar optimum low complexity cyber-attack mitigation on sensor-equipped networks of autonomous and human driven vehicles employed for o,bject tracking with a goal of ultimate extension to all relevant tasks. The project intends to build this theory based on rigorous a,nalytical characterization of after attack per-formance for various complexity mitigation approaches under any attacks, regardless o,f if they are on sensors, hardware, software or anything else. Such results do not currently exist. These theoret-ical results will,be employed to develop low complexity near optimum tracking algorithms under attacks with the required scaling and distributed imple,mentations with a goal of ultimate extension to all relevant tasks, including navigation/control. These algorithms will employ unsup,ervised and supervised machine learning and incorporate all relevant information. The results would provide strong contributions to,the Navy and to science in general.The theory and algorithms developed will help protect a large set of systems of great interest to, the Navy, even those which do not involve vehicles. These include systems modeled entirely based on machine learning or by any comb,ination of physics and data driven ideas.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 05, 2022
- Source ID
- N000142212626
Entities
People
- Rick Blum
Organizations
- Lehigh University
- Office of Naval Research
- United States Navy