Traffic-driven epidemic spreading in finite-size scale-free networks

 

SANDRO MELONI

DEPARTMENT OF INFORMATICS AND AUTOMATION, UNIVERSI

 

Scale-free networks have been shown to reproduce the structural form of many social, technological and biological systems. The study of this kind of networks has given many unexpected results. One of them is the absence of an epidemic threshold for diseases spreading in infinite-size scale-free networks. Nowadays, one of the open problems in epidemic modeling on complex networks is how to deal with irregular interaction between nodes. The vast majority of current theoretical approaches assumes that infections are transmitted as a reaction process from nodes to all neighbors. Here we try to reproduce irregular interactions by means of traffic flows. Specifically, we propose a disease spreading model in which traffic flows drive interactions and contagion between nodes. Both analytical predictions and numerical simulations show that the epidemic incidence is shaped by traffic conditions and that the epidemic threshold is related to the first and second moments of the betweenness centrality distribution for a given routing protocol. In particular the threshold value depends on the traffic and decreases as flow increases. This conceptual framework opens also a new way to model complex interaction in dynamical processes in complex networks.