NSF Convergence Accelerator Track G: Proactive End-to-End Zero Trust-Based Security Intelligence for Resilient Non-cooperative 5G Networks

Regents Of The University Of Michigan - Dearborn, Dearborn MI

Investigators

Abstract

Military and tactical units are required to operate in regions where a dedicated and trusted communications infrastructure may not be available. To communicate effectively, efficiently, and securely at all times, it is critical to have the ability of operating through untrusted indigenous infrastructure and at the same time, seamlessly and securely connect with devices on trusted networks. With the growing threat landscape and attack surfaces in 5G communication systems, a promising approach to tackle the security challenges is to adopt a holistic Zero Trust approach based on the principle of “never trust, always verify”. The goal of this project is to achieve trustworthy communications over untrusted and potentially compromised 5G networks by creating end-to-end security intelligence. The project team comprising of experts from academia, industry, and academic-industry partnerships, will accelerate the development of a software-based solution for automated and trustworthy network orchestration allowing military and critical infrastructure operators to securely communicate using commercial 5G networks. The project aims to develop capabilities that enable military and critical infrastructure personnel to securely communicate using both military and modern commercial 5G infrastructure anywhere in the world, without the risk of being breached or hacked. The developed tools will enhance the capabilities of tactical networks in enforcing Zero Trust and policy-based access management of tactical units. Furthermore, the convergence team will develop a cross-disciplinary curriculum for training and education of next-generation workforce on Zero Trust for 5G networks. Special emphasis will be placed on recruiting under-represented minority students in the research and development of secure 5G networks. This project will integrate cross-disciplinary expertise from trustworthy system design, machine learning/artificial intelligence, and 5G networks & edge computing to accelerate the automated creation of an overlay of proactive, end-to-end zero-trust security and resilience mechanism over the tactical 5G network cloud infrastructure. The convergence research is centered around three main thrusts. Thrust 1 will focus on the dynamic and proactive trust evaluation of tactical devices in the network. A distributed and autonomous approach will be used to collect information about device interactions, leverage triggers, and indicators to make quantitative trust assessments. Thrust 2 will use the trust information obtained in Thrust 1 to orchestrate network resources according to tactical mission scenarios and the requirements of different end-to-end paths. It will be achieved using a deep reinforcement learning approach by iteratively obtaining access permissions from a novel trust management engine and providing it with resource selection decisions. Thrust 3 will automate the access management system to proactively thwart unqualified access to critical resources and prevent lateral movement of attacks and malware. The tools developed as a result of this project will be oriented toward operational scenarios in tactical missions. 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.