Random failures and attacks on Small World Networks

Abstract

Attacks and random failures are powerful damaging tools to know the robustness of a network. Here, a damage analysis was implemented in two different systems. In a first system, we characterized the different node centrality measures for a dolphins’ social network. The results revealed that those nodes with a higher degree have a higher contribution to the exchanging infor-mation capacity of the network than those with higher betweenness centrality. In a second system, attacks and random failures were used in the functional network of two neuronal cultures, one with a homogeneous distribution of neurons and another with an anisotropic, patterned structure. Results showed that both networks were resilient to random failures, but that the patterned network was sensitive to attacks. This suggests that the pattern network is fragiler and more dependent of nodes with high centrality