Researchers untangle complex network systems

By exploring the tangled nature of complex network systems, researchers at the Laboratory and the University of Houston may have found a way to help scientists and engineers better understand dynamic processes on complex networks, such as the spread of infectious diseases, cascading massive electrical power failures, sources of vehicle traffic congestion on metropolitan roadways and information flow on the Internet.

In a paper published in a recent issue of the journal Nature, scientists Zoltan Toroczkai and Kevin Bassler explain the connection between the nature of networks and efficiency of flow processing on these structures. That is, how the geometry of interconnections in a network affect the transport processes in them. The researchers show that not all networks are alike and certain types of networks, those called scale-free networks, are more adept for efficient transportation and flow processing, while others -- what they call non-scale-free networks -- will more easily become congested, or jammed.

To reach this conclusion, the researchers made two key observations: the first observation was that large-scale complex networks that have not been globally designed, but rather evolved spontaneously, are more optimal for transport and flow processing. These networks include information networks such as the Internet, electronic mail networks, social networks, and some biological networks, like cell metabolic pathways.

The second observation was that many transport processes are induced by the existence of local gradients, such as chemical potential, temperature or concentration, and that the presence of these gradients will naturally generate, or influence, flows in complex networks. Combining these observations, the researchers show that scale-free networks, networks that follow a power-law degree, rather than a Poisson, or bell curve, distribution, have more efficient processing of gradients-induced flows.

The value of this discovery is perhaps greatest to network analysts, designers and engineers who could use the information to gain a better understanding of network transport efficiencies. Toroczkai and Bassler have already begun research into understanding how the principles of scale-free networks can be incorporated into the design of existing infrastructure networks to improve their efficiency of transport and robustness.

Toroczkai is acting deputy director of the Center for Nonlinear Studies (T-CNLS). The research at Los Alamos was part of a complex networks research project supported by Laboratory-Directed Research and Development (LDRD) funds. LDRD funds basic and applied research and development focusing on creative concepts selected at the discretion of the Laboratory director.

-- Todd Hanson

Other Headlines