Factors That Affect the Centrality Controllability of Scale-Free Networks
HU Dong1, SUN Xian1, LI Ping1**, CHEN Yan1, ZHANG Jie2**
1Center for Intelligent and Networked Systems, School of Computer Science, Southwest Petroleum University, Chengdu 610500 2Center for Computational Systems Biology, Fudan University, Shanghai 200433
Abstract：The influence of a node in a network can be characterized by its macroscopic properties such as eigenvector centrality. An issue of significant theoretical and practical interest is to modify the influence or roles of the nodes in a network, and recent advances indicate that this can be achieved by just controlling a subset of nodes: the so-called controllers. However, the relationship between the structural properties of a network and its controllability, e.g., the control of node importance, is still not well understood. Here we systematically explore this relationship by constructing scale-free networks with a fixed degree sequence and tunable network characteristics. We calculate the relative size (nC*) of the minimal controlling set required to control the importance of each individual node in a network. It is found that while clustering has no significant impact on nC*, changes in degree–degree correlations, heterogeneity and the average degree of networks demonstrate a discernible impact on its controllability.
2015, Vol. 32 
