Cyber-attack and defense simulation framework

Abstract

Various papers on cyberwarfare in virtual environments and cybersecurity in intelligent systems have been published. Work has focused on the integration of cyberwarfare communication effects into a live–virtual–constructive (LVC) environment in order to better represent a network centric battlespace subject to cyber-attack, at a large force level. In addition, virtual cyber ranges have been developed. A virtual cyber range is a portable modeling and simulation framework that provides a real-time, hardware-in-the-loop capability for simulation of cyber threats to the entire net-centric infrastructure. The framework enables interoperability with LVC simulations, providing training and assessment of human-in-the-loop performance. This work builds on previous work and shows the need for a framework to support the modeling and simulation of cyber-attack and defense for training and assessment. This work does focus on cybersecurity for wireless communications, but the focus is on large network-centric systems at force level. Furthermore, there are published descriptions of research focused on the use of multi-agent systems in vehicle-to-vehicle communication and autonomous driving systems, such as Google’s driverless car. While the vehicles are commercial by nature, there are many similarities to military autonomous vehicles (MAVs), in their system components and architecture as well as the cybersecurity threats that they face. While these works are related to the work presented in this paper, with their focus on synthetic environments and commercial autonomous systems, they are not closely related. The focus of this paper is on small point to point and point to multipoint wireless networks, comprised of only a few nodes, with the nodes representing MAVs and controllers. The US Army, Air Force, and Navy are continuing to increase their use of these systems, and are now focusing on cybersecurity. Due to the lagging focus on wireless networks and autonomous vehicles, there is not much published work that is related. Current unmanned systems were not built with cybersecurity considerations taken into account, and are thus vulnerable to cyber-attack. The objective of this paper is to present and describe the need for a cyber-attack and defense simulation framework to support the modeling and simulation for cybersecurity of autonomous vehicle systems used by US Armed Forces. These autonomous vehicle systems include unmanned aerial systems and unmanned ground systems. The paper describes a notional framework to support this type of modeling, as well as a detailed use case and example cyber-attack simulation system.