Effect of Preferential Resource Allocation on Epidemic Spreading in Correlated Multiplex Networks

Based on the global information of the structure of complex networks, various optimization algorithms has been proposed to optimally allocate the limited public resource to suppress the disease spreading. However, generally we can only have the local information about the connections among individuals in the complex network, and the infected individuals can only depend on the help from their local friends other than the public resource invested to support their recovery. To this end, we propose a model that the process of resource allocation and disease spreading coupled in a two-layer multiplex network. Through extensive simulations, we find that when the two layers of the network are randomly correlated, the resources of healthy nodes should be preferentially allocated to small degree nodes in order to better suppress the disease spreading, which is in contrast to the intuition that the hub nodes should be protected. Moreover, we find that there exists an optimal strategy, under which the disease could be suppressed to the most extend. When the two layers are negatively correlated, the resources of the healthy nodes should allocated preferentially to the large degree nodes, while it's the opposite when the two layers are positively correlated. More interestingly, we find that the phase transition type of the fraction of infected nodes changes from continuous transition to hybrid transitions with the increase of a bias parameter.