An Entropic Topology Reconfiguration to Improve the Robustness of Complex Networks under Failures and Targeted Attacks
نویسندگان
1 Faculty of Computer Science and Engineering, Shahid Beheshti University, Tehran, Iran
2 School of Computer Science, Institute for Research in Fundamental Sciences (IPM), Tehran, Iran
doi
چکیده
One of the defensive strategies to increase the resilience of the networks is the rewiring mechanism in which the affected nodes are disintegrated from the faulty nodes and connect to a profitable node with a given probability. Inthis article, based on the concept of Shannon’s entropy, a rewiring method is proposed to modify the configuration of the complex networks and improve their resilience. We measure the network robustness based on a spectrum ofdegree distribution, heterogeneity, and in addition, the average size of the largest connected cluster during removing the nodes as the consequence of systematic attacks to the centralities (i.e. degree, between ness, and dangalchev’s closeness). We apply the proposed rewiring strategy to six artificial data sets and six real data sets.The experimental results showed that a significant improvement of general resilience can be obtained throughswapping about 30% of communication links.
