Towards Cross-Domain Resilience in Interdependent Power and ICT Infrastructures: A Failure Modes and Effects Analysis of an SDN-enabled Smart Power Grid

The adoption of cloud-native technologies like the Software Defined Networking (SDN) paradigm, into the management of Critical Cyber-Physical System (CCPS)'s monitoring and control functions, leads to the emergence of complex interdependencies between the cyber and physical domains, which would increase the risk of cascading failure, especially in the cyberdomain represented by edge Data Center (DC) networks. These Edge DCs host critical software services characterized by high dependability and performance requirements. The downtime of such services has a considerable impact that may destabilize socioeconomic well-being. In this work, we provide a failure modes analysis of an SDN-enbaled Smart Power Grid (SD-SPG) with a focus on the subsystems involved in cross-domain failure propagation. The objective of the analysis is to establish the causal effect between subsystem failure modes that may lead to cross-domain failure cascades. Then, we focus on the evaluation of Steady State Availability (SSA) metric under different interaction scenarios between the power and telecommunication subsystems. To this end, we propose a hierarchical modeling framework combining continuous-time Markov chains (CTMCs) and Reliability Block Diagram (RBD)s to capture both, subsystems and complex systems' steady-state behavior.

Keywords: NFV, SDN, Smart grid, Dependability evaluation, Failure mode analysis, Hierarchical modeling, Markov chain, Reliability block diagrams.

Download PDF