The Effect of Clustering-Based and Degree-Based Weighting on Robustness in Symmetrically Coupled Heterogeneous Interdependent Networks

We studied robustness of symmetrically coupled interdependent networks against load failures when the flow dynamics is dominated by global redistribution based on weighted betweenness centrality. Three kinds of weighting schemes, which are random, clustering-based, and degree-based, respectively, were compared. We demonstrated that both the clustering-based and degree-based weighting scheme outweigh random weighting on robustness against the cascade of load failures in the symmetrically coupled interdependent networks. In a limited range of weighting parameters, degree-based weighting scheme outweigh clustering-based scheme. Outside of that range, clustering-based weighting scheme outweigh degree-based scheme. And it will be easier for the symmetrically coupled interdependent network to achieve robustness against one-node removal and better cost configuration against the cascade of load failures when relative interconnection intensity was weaker under degree-based and clustering-based weighting schemes. Our findings might have great generality for characterizing load-failure-induced cascading dynamics in real-world weighted interdependent networks.