Robustness of Coherence in Noisy Scale-Free Networks and Applications to Identification of Influential Spreaders

Abstract

This brief investigates robustness of the consensus characterized as coherence in the noisy scale-free networks under average degree, random nodal failures, and target attacks. A new centrality index named as leader centrality is proposed to identify more influential spreaders based on the coherence. It is shown that the leaderless and leader–follower consensus is better at the same network size with higher average degree. The obtained results reveal that the consensus becomes worse under the target attacks where the nodes are removed by the descending order of degree, closeness and eigenvector centrality. It is numerically demonstrated that the leader centrality is more effective to identify important spreaders by simulating the spreading process of an SIR epidemic model.