Target control based on edge dynamics in complex networks

Structural controllability of general edge dynamics in complex network

Dynamic processes that occur on the edge of complex networks are relevant to a variety of real-world systems, where states are defined on individual edges, and nodes are active components with information processing capabilities. In traditional studies of edge controllability, all adjacent edge states are assumed to be coupled. In this paper, we release this all-to-all coupling restriction and propose a general edge dynamics model. We give a theoretical framework to study the structural controllability of the general edge dynamics and find that the set of driver nodes for edge controllability is unique and determined by the local information of nodes. Applying our framework to a large number of model and real networks, we find that there exist lower and upper bounds of edge controllability, which are determined by the coupling density, where the coupling density is the proportion of adjacent edge states that are coupled. Then we investigate the proportion of effective coupling in edge controllability and find that homogeneous and relatively sparse networks have a higher proportion, and that the proportion is mainly determined by degree distribution. Finally, we analyze the role of edges in edge controllability and find that it is largely encoded by the coupling density and degree distribution, and are influenced by in- and out-degree correlation.

Nodes with the highest control power play an important role at the final level of cooperation in directed networks

Controllability of complex networks aims to seek the lowest number of nodes (the driver nodes) that can control all the nodes by receiving the input signals. The concept of control centrality is used to determine the power of each node to control the network. The more a node controls the nodes through connections in the network, the more it has the power to control. Although the cooperative and free-rider strategies and the final level of cooperation in a population are considered and studied in the public goods game. However, it is yet to determine a solution to indicate the effectiveness of each member in changing the strategies of the other members. In a network, the choice of nodes effective in changing the other nodes' strategies, as free-riders, will lead to lower cooperation and vice versa. This paper uses simulated and real networks to investigate that the nodes with the highest control power are more effective than the hubs, local, and random nodes in changing the strategies of the other nodes and the final level of cooperation. Results indicate that the nodes with the highest control power as free-riders, compared to the other sets being under consideration, can lead to a lower level of cooperation and are, therefore, more effective in changing the strategies of the other nodes. The obtained results can be considered in the treatment of cancer. So that, destroying the tumoral cells with the highest control power should be a priority as these cells have a higher capability to change the strategies of the other cells from cooperators to free-riders (healthy to tumoral).