Exploring the effectiveness of node attacks based on combined centrality measures in scale-free networks

Abstract

Abstract This article introduces a novel method for targeting complex networks that involves using a hybrid centrality score to rank nodes and carry out attacks. Unlike previous studies that have focused on using individual centrality measures, this approach takes into consideration the varying significance of nodes across different centrality measures. The study utilizes simulations on scale-free networks to demonstrate that the proposed strategy can be highly effective in inducing network failure, and certain combinations of centrality measures can result in greater attack severity than using individual measures alone. Overall, the research offers valuable insights into improving node-attack strategies for complex networks, which are typically resilient to random failures but susceptible to targeted attacks.