Recovery of complex network after intentional attack

• Main Author: Wang, Hongbo
• Other Authors: Xiao Gaoxi
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/152507
• Institution: Nanyang Technological University

Partially due to the fast developments of the network technology, the complex networks have found their wide applications in many different areas. One of the most significant observations is that many real-life systems can be resembled as scale-free networks, where the nodal degrees of the networks approximately follow a power-law distribution. In such networks, a limited number of nodes, known as hubs or hub nodes, have very high degrees. It is known that when these hub nodes are broken down or under attack, the connectivity of the networks may be badly destroyed. Repairing or recovering of the networks under or after intentional attack therefore becomes an important topic that has been attracting intensive research interests. Existing studies have typically considered the cases where system repairing could be carried out on global basis or the attacked nodes can partially or totally recover their function, both of which may not necessarily be the case in many real-life applications. In this study, we mainly focus on the cases where system recovery is mainly done by network link compensation, and more specifically, when such link compensations are mainly carried out based on local system information. Considering that the average shortest path length indicates the average cost of inter-nodal communications in the network and the largest cluster size indicates whether the network is still largely connected after the attack and repair, in our study we mainly adopt these two metrics to reflect the survivability of the network. Extensive numerical simulations are conducted to compare and evaluate a few different link compensation schemes under different attack methods. Best repairing methods under different types of intentional attacks are identified. The evaluation of these methods, as we believe, shall lay a basis for further studies and future developments of more effective system repairing schemes.