Transport optimization on complex gradient networks

Abstract

We present a strategy of the optimization of transport on complex gradient networks by introducing the correlations between the gradient field and the local topology of networks. In this strategy, the value hi of the gradient field at node i is determined by the degree kj of its own neighborsj, that is, hi=∑jkj−α(α is control parameter). We find that there exists the optimal values of α, where the structural correlated gradient field can strongly reduce jamming in both random and scale-free networks. Furthermore, the optimal scale-free networks are less congested than optimal random networks for average degree〈k〉>2. In addition, by comparing the transport capacity of network under different routing rules ranging from random diffusion to gradient-driven transport, we find that gradient-driven transport on the scale-free networks can hold the maximal transport capacity when the gradient field is optimal correlated. All these observations present the evidence supporting the idea that the scale-free topology and gradient-driven transport mode are beneficial to improve the transport capacity of networks.