Frangibility of Complex Networks Based on Flow Attack and Edge Failure
基于流量攻击和边失效的复杂网络脆弱特性

Attacks based on flow may bring tremendous damage to complex networks. In existing works, the cases of nodes attacking are mainly concerned, however, few work is involved to the edges. In this paper, the frangibility of complex networks is discussed in the case of some edges being deleted. Additionally, the effects of time strategy and network size are also discussed. By analyzing the load and degree of complex networks, it is demonstrated that the complex networks possess a high heterogeneous distribution of loads, which is caused by the power-law degree distribution, and the heterogeneity makes the networks particularly vulnerability to attacks. The analytic results show that complex networks exhibit strong error tolerance to random failures of edges, but a large-scale cascade of node failure can be triggered by disabling several key edges, which may result in the collapse of networks.