Monitoring social centrality for peer-to-peer network protection

Resilience of networking infrastructures is crucial for ensuring protection and readiness in the case of an emergency. Enabling Peer-to-Peer (P2P) communication is one way of alleviating potential outcomes of depending on a centralized system with a single point of failure, which would be hard to recover in the case of cyberattack. In this article, we consider P2P-enabled networks and discuss how it is possible to benefit from social centrality metrics to prevent catastrophic spreading of malicious content. Using classical Susceptible-Infected- Recovered (SIR) simulation models designed for disease propagation in a social community on two different reallife P2P network topologies, we demonstrate that the nodes that are ranked highly by social centrality metrics should be better protected to prevent malicious content propagation. Our results suggest that nodes that are central are significantly more effective than random nodes in spreading malicious content across the network and among different centrality metrics, Eigenvector centrality is more useful for identifying the nodes that should be better protected in P2P networks.