SMART-5G: Secure Multichannel Automated operRations Through 5G Networks

International Business Machines Corporation, Reston

Investigators

Abstract

The objective of the SMART-5G project or Secure Multichannel Automated opeRations Through 5G Networks, is to develop a suite of technologies, tools and operational principles which will allow for the secure operation of military communications on civilian 5G infrastructure. We will employ the capabilities of Multi-access Edge Computers or MEC, for secure communication. By leveraging military-provided servers that are deployed within the 5G network infrastructure plus servers at the tactical edge and in the backend infrastructure, we show how a more secure operational environment can be obtained with a variety of techniques including: application of AI to network traffic inspection, policy-driven separation of traffic along a low-bandwidth secure communication path and a high-bandwidth insecure communication path, virtual multi-function authentication leveraging mobile edge computing mechanisms, and self-generation of security policies for automated operations. To support the scenarios to be developed as part of the convergence research project, the team will develop a suite of technologies to enable military personnel to operate through the insecure 5G network without compromising security requirements, and group these technologies into three different, but inter-related thrusts: multi-channel exploitation, network situation awareness, and automation of security workflows. Multi-channel exploitation includes the use of multiple channels with different performance and security properties to provide the required environment for mission-critical applications. Network situation awareness provides information to users through monitoring and analysis of the traffic and activities in the 5G network; and, automation of security workflows use automation of security protocols to improve responsiveness to threats. This multidisciplinary effort brings together research fields such as computer science, machine learning, network operations and management, and social science. Potential threats that can be launched through social engineering will be identified, and countermeasures proposed for detection and mitigation.