基于分层攻击的社团结构稳定性研究

Stability of Community Structure Based on Hierarchical Attack Strategy

针对复杂网络的社团结构稳定性问题,基于节点的聚类系数,在传统的蓄意攻击和随机攻击的基础上提出了分层攻击策略;对比社团结构稳定性测度量,运用测度E来衡量攻击网络前后其社团结构划分的差异性。分别用分层攻击、蓄意攻击和随机攻击对网络进行打击,对比研究测度E的大小,仿真试验结果表明社团结构在分层攻击和蓄意攻击下表现出脆弱性,而在随机攻击下表现出稳定性;同时,结果显示,对于无标度特性不明显的网络,分层攻击比蓄意攻击和随机攻击对网络的攻击性更强。

According to the problem about the stability of community structure,the hierarchical attack strategy based on node clustering coefficient was proposed on the basis of traditional deliberate attack and random attack.Contrasted with the stability measures of community structure,the measure E was used to measure the differences of community structure partition.While separately attacking the networks under hierarchical attack,deliberate attack and random attack,the quantitative value of E was comparatively studied.The results of simulation experiment show that the community structure is fragile under hierarchical attack and deliberate attack,but it is stable under random attack.It also proves that hierarchical attack is more effective than deliberate attack and random attack on networks that do not have obvious scale-free features.