Invariant-driven intrusion detection system (IDS) for cyber-physical systems

网络物理系统（CPS）是与其物理环境密切交互的计算机系统。由于 ** 相互关联的性质和相对缺乏保护，CPS一直受到 ** 恶意攻击者的安全攻击。例如，汽车、智能电网和胰岛素泵都是CPS被成功攻击的例子。在我们以前的工作中，我们开发了ARTINALI，一个基于不变量的入侵 ** 检测系统（IDS），用于检测攻击CPS使用自动化的方法。我们在两个小型CPS系统上演示了 **ARTINALI，并表明它在假 ** 阳性和假阴性方面都优于以前的IDS。在这个项目中，我们计划将ARTINALI扩展到包括数千个信号和事件的大型CPS系统 **，以及具有实时操作限制的系统，这是我们的合作伙伴加拿大通用动力陆地系统公司（GDLS）感兴趣的。

Cyber-physical systems (CPS) are computer systems that closely interact with their physical environments. Due**to their interconnected nature and relative lack of protection, CPS have been subject to security attacks by**malicious attackers. For example, automobiles, smart grids and insulin pumps are examples of CPSes that have**been attacked successfully. In our prior work, we developed ARTINALI, an invariant-based Intrusion**Detection System (IDS) for detecting attacks on CPS using an automated approach. We demonstrated**ARTINALI on two small CPS systems, and showed that it outperformed previous IDS in terms of both false**positives and false negatives. In this project, we plan to extend ARTINALI to large-scale CPS systems**comprising thousands of signals and events, and to systems with real-time operating constraints, which are of**interest to our partner, General Dynamics Land Systems (GDLS), Canada.