Identifying and Protecting Cyber-Physical Systems’ Influential Devices for Sustainable Cybersecurity

Abstract

For sustainable cyber-physical systems (CPS) security, proactive measures to cybersecurity need to be implemented instead of reactive measures. Towards this, we introduce in this paper a proactive methodology implemented in a system called IDI_CPS. It is based on the observation that CPS devices that have LAN-based network sharing (e.g., via Wi-Fi connections) need first to be clustered using some clustering criterion. Then, the influential and central devices in these clusters need to be identified to pay more attention to their file sharing. These influential devices may have network sharing with devices at the WAN level. Therefore, the influential devices at the WAN level that have network sharing with the influential devices in the clusters need also to be identified to pay more attention to their file sharing. We propose novel techniques for: (1) clustering the devices that have LAN-based network sharing using k-clique modeling, (2) employing clustering coefficient-based techniques for identifying the most influential device in each cluster, and (3) employing Independent Cascades model-based techniques for identifying the influential devices at the WAN level that have network sharing with the influential devices in the clusters. We experimentally evaluated our proposed system IDI_CPS and compared it with four comparable methods. Results showed marked improvement.