摘要
为减小相互依存网络间的相继故障,在对归一化度、网络间的匹配系数及网络间的簇系数定义的基础上,借鉴网络间相似拓扑耦合思想,相互提出一种网络间同地位节点耦合的拓扑构建方法,该方法以核心节点作为搜索源节点采用广度优先搜索算法,逐级搜索并最大化网络同地位节点对的匹配,以提高相似匹配度和扩展应用场景。并以相互依存的随机网络和相互依存的无标度网络作为实例进行仿真,实验表明:此拓扑连接方法下,网络间故障渗流相变从一维非连续相变转变为二维连续相变到;相比于随机拓扑耦合网络在随机攻击、目的攻击及防御情况下,该拓扑耦合下的相互依存网络的鲁棒性均明显增强。
To reduce cascading failures of interdependent networks, it introduces a topological coupling strategy that a network connects another network with same position nodes, by drawing on the experience of inter-similarity coupling, after defining normalized degree, inter-assortativity coefficient and inter-clustering coefficient. Adopting breadth first search algorithm and taking hub node as initial search node, the strategy can improve the matching degree of inter-similarity and extend the application scene. Interdependent ER networks and SR networks are taken as examples and simulated, the result of the experiment implies that the coupling algorithm leads to change from a first to second order percolation transition, and can improve robustness of interdependent networks compared to random coupling algorithm under targeted attacks, random attacks and targeted defenses.
出处
《计算机工程与应用》
CSCD
2014年第11期85-89,共5页
Computer Engineering and Applications
基金
国家自然科学基金(No.61201290)
湖南省自然科学基金项目(No.13JJ6072
No.12JJ4060)
湖南省教育厅科学研究项目(No.12C0584)
益阳市科技计划项目(No.2011JZ45)
关键词
相继故障
相互依存网络
匹配系数
故障渗流
随机网络
无标度网络
cascade of failures
interdependent networks
assortativity coefficient
percolation of failures
Erdos-Renyir networks
scale-free networks
