Empirical models for complex network dynamics: A preliminary study

Abstract

Network analysis has draw a considerable amount of attention in the last decade, especially after the discovery of common topological characteristics such as Small World or a Power Law degree distribution. Recently our understanding of complex networks has been augmented with the inclusion of a local view of patterns of connectivity, such patterns that are present more often in real networks than in randomized ones have been called motifs. These global and local perspectives equip us with powerful tools to understand the behavior of many real networks. However, an important aspect of complex network analysis is often neglected: the dynamics of the information flows. The structural elements of the network topology are very important, but to fully understand the dynamics of these networks we need to take a closer look at the dynamics of the information flow in a self-regulation perspective. For example, we know that the performance and reliability of a compute network is likely influenced by the dynamics of the packet flows, as much as it is influenced by the network topology. In a biological regulatory network we need to understand the dynamics that control the excitation and the suppression of gene activity and other transcription factors. In this work we introduce a preliminary simulation study of the flow of information in networks with different topological properties and activation functions. The goal is to approach the analysis of network dynamics from a data-driven approach, using simulations to capture, understand, and possibly model the overall dynamics of the network in a self-regulated perspective. © 2014 Springer International Publishing Switzerland.