Weighted Networks Model Based on Traffic Dynamics with Local Perturbation

In the study of weighted complex networks, the interplay betweentraffic and topology have been paid much attention. However, thevariation of topology and weight brought by new added vertices oredges should also be considered. In this paper, an evolution modelof weighted networks driven by traffic dynamics with local perturbation is proposed. The model gives power-law distribution of degree, weight and strength, as confirmed by empirical measurements. By choosing appropriate parameters W and δ, the exponentsof various power law distributions can be adjusted to meet real world networks. Nontrivial clustering coefficient C, degreeassortativity coefficient r, and strength-degree correlation arealso considered. What should be emphasized is that, with theconsideration of local perturbation, one can adjust the exponent ofstrength-degree correlation more effectively. It makes our modelmore general than previous ones and may help reproducing real worldnetworks more appropriately.