| Many complex systems are characterized by collective activity over a broad range of scale, for example in neural networks. Studying synchronization in networks of coupled oscillators can shed light on the emergence of such complex activity. Here we characterize complexity of synchronization patterns in complex networks. A measure of complexity is small for both weak and strong couplings, and it displays a maximum at intermediate coupling, reflecting the dynamical segregation and integration simultaneously. We show that networks with communities possess high complexity in a broad range of coupling. Then we study the complexity and modularity in the cat corticocortical network by modeling the network activity using neurophysiologically realistic neural mass oscillators. We find that the dynamical complexity is reduced when the original network is rewired to increase or decrease the modularity. Thus the real brain network is optimized in the light that it enables both functional segregation and integration manifested by the maximal complexity. |
|