Dynamics of and on complex networks

Abstract

Complex networks are dynamic, evolving structures that can host a great number of dynamical processes. In this thesis, we address current challenges regarding the dynamics of and dynamical processes on complex networks. First, we study complex network dynamics from the standpoint of network growth. As a quantitative measure of the complexity and information content of networks generated by growing network models, we define and evaluate their entropy rate. We propose stochastic growth models inspired by the duplication-divergence mechanism to generate epistatic interaction networks and find that they exhibit the property of monochromaticity as a result of their dynamical evolution. Second, we explore the dynamics of quantum mechanical processes on complex networks. We investigate the Bose-Hubbard model on annealed and quenched scale-free networks as well as Apollonian networks and show that their phase diagram changes significantly in the presence of complex topologies, depending on the second degree of the degree distribution and the maximal eigenvalue of the adjacency matrix. We then study the Jaynes-Cummings-Hubbard model on various complex topologies and demonstrate the importance of the maximal eigenvalue of the hopping matrix in determining the phase diagram of the model. Third, we investigate dynamical processes on interacting and multiplex networks. We study opinion dynamics in a simulated setting of two antagonistically interacting networks and recover the importance of connectivity and committed agents. We propose a multiplex centrality measure that takes into account the connectivity patterns within and across different layers and find that the dynamics of biased random walks on multiplex networks gives rise to a centrality ranking that is different from univariate centrality measures. Finally, we study the statistical mechanics of multilayered spatial networks and demonstrate the emergence of significant link overlap and improved navigability in multiplex and interacting spatial networks.