Reorganizations of complex networks: Compounding and reducing

Real networks interact with each other by different kinds of topological connections, which are usually demonstrated by linking nodes of different networks. Simple connection, such as one-to-one corresponding, random connection and similar connection are adopted for studying the interacted networks. Practical interrelations established between the two networks are ignored. In this study, a generalized framework of multi-subnet composited complex network that allowed us to investigate interrelations among several subnets is developed. Based on that, reorganizations of networks: compounding (compound subnets into a "bigger" one) and reducing (obtain a "smaller" network from a "bigger" one) are proposed. As an empirical evidence, influence of compounding on traffic dynamics is discussed. And the properties of nodes linking two networks are also considered. Onset of compounding between two networks is revealed. Numerical simulations on artificial networks as well as real bus and tube networks of Qingdao in China agree well with our analysis, which show validity of our model.