Synchronization processes in complex networks (Q858502)

This paper presents a numerical study of the process of synchronization in networks of \(N\) oscillators of the Kuramoto type, described by the system of differential equations \[ \frac{d\theta_i}{dt}=\omega +\sum_{j}K_{ij}\sin(\theta_j-\theta_i),\quad i=1,\dots N. \] The focus is on the way that the network structure, which is determined by the coefficients \(K_{ij}\), determines the gradual synchronization of the system in time: when the network consists of ``communities'' of densely-coupled oscillators, with sparser inter-connections among oscillators in different communities. The process of synchronization reflects this fact: the oscillators in a given community synchronize on a faster time scale, and the different communities synchronize at a later time. Several examples are studied, and several graphical representations are displayed, which show this process. Another manifestation of the network structure, which is discussed in this paper, is the spectrum of the Laplacian matrix of the network. The numerical results also show a resemblance between: (1) the graph displaying the number of synchronize clusters of oscillators versus time, and: (2) the graph showing the inverse of the \(k\)th eigenvalue of the Laplacian matrix versus \(k\). A heuristic explanation for this similarity is given.