Overview of Security for Smart Cyber-Physical Systems

Abstract

The tremendous growth of interconnectivity and dependencies of physical and cyber domains in cyber-physical systems (CPS) makes them vulnerable to several security threats like remote cyber-attacks, hardware, and software-based side-channel attacks, especially in safety-critical applications, i.e., healthcare, autonomous driving, etc. Though traditional software or hardware security measures can address these attacks in the respective domains due to enormous data and interdependencies of the physical-world and cyber-world, these techniques cannot be used directly. Therefore, to address these challenges, machine learning-based security measures have been proposed. This chapter first presents a brief overview of various security threats at different CPS layers, their respective threat models, and associated research challenges towards developing robust security measures. Towards the end, we briefly discuss and present a preliminary analysis of the state-of-the-art online anomaly detection techniques that leverage the machine learning algorithms and property-specific language, respectively.