Structure of flow and noise on functional scale-free networks

Abstract

Topological heterogeneity of structured networks effects course of dynamic processes on them by causing uneven function of nodes and/or links. Within the model of information traffic we demonstrate how dynamically relevant topology emerges on cyclic scale-free Web graph. The dynamic centrality measures are quantitatively characterized through analysis of traffic flow and multichannel noise. The dynamically generated heterogeneity of links is represented by maximum spanning tree. Fluctuations of noise at individual nodes in traffic on the original graph and on its maximum spanning tree (optimal transport) reveal universal scaling features that are related to the underlying link structure and driving conditions. Furthermore, we give some evidence for stochastic ergodicity breaking in traffic on networks.