Community robustness and its enhancement in interdependent networks

Abstract The complex network attracts increasing attention in scientific researches and daily applications, and there has been an explosion of studies on the network theory. Concentrating on the functional information behind network structures, the community robustness has been proposed and dissected on single networks in recent studies. Currently, existing studies on the robustness of interdependent networks are limited to structural connections. As indicated by several practical dilemmas, the analyses on the invulnerability of communities and the maintenance of functional clusters in interdependent networked systems should be considered urgently. Therefore, in this paper, we first design a measure for numerically evaluating the community robustness of interdependent networks; and then guided by this measure, a memetic optimization algorithm, termed MA-CR inter , is proposed to successfully enhance the community robustness of various synthetic and real-world networks through rewiring topologies. Furthermore, two non-rewiring optimization strategies have also been modified and tested to make comparisons. The results show that these methods are valuable for enhancing networks’ community robustness as well, and different optimization strategies can be selected based on actual restraints and requirements Aiming at improving the attack tolerance of communities in interdependent networks, the obtained results facilitate the information mining from networked systems and may provide potential solutions to realistic optimization problems.