A complex networks framework for cyber-physical power system critical node identification considering various communication structures and coupling relationships

Cyber-physical power systems (CPPS) require robust identification of critical nodes due to their complex cyber-power interdependencies. This paper proposes a complex networks framework combining: community-based centrality for structural role analysis, extended centrality incorporating electrical properties, and communication architecture sensitivity analysis. The latter evaluates diverse communication topologies, Rewiring, Barabasi-Albert scale-free, Watts-Strogatz small-world, and Erdos-Renyi random networks, while investigating four one-to-one connection techniques and partially dependent connections between power and communication layers. Validated on IEEE 30-bus and 118-bus systems, the methodology demonstrates that coupling relationships significantly impact node criticality rankings. Key findings reveal that while the most critical nodes remain consistent, secondary critical nodes exhibit notable variations across communication architectures and coupling modalities. Unlike conventional methods that analyze topological, electrical, or cyber layers in isolation, the proposed framework integrates these dimensions to identify vulnerabilities that may otherwise remain overlooked. This work provides a practical approach for optimizing protection strategies, enhancing resilience against cascading failures, and guiding communication network design in CPPS.