Resilience of space information network based on combination of complex networks and hypergraphs

Le Zhang,Ye Du

DOI: https://doi.org/10.1016/j.comcom.2022.08.016

IF: 5.047

2022-11-01

Computer Communications

Abstract:The space information network provides global users with high-speed information services that can be accessed anytime, anywhere. Its enormous economic, social and military value makes it vulnerable to large-scale targeted cyber attacks. Therefore, enhancing the resiliency of services is critical. We propose to solve the above problems from the perspective of complex networks and hypergraph fusion. A space information network model integrating complex networks and hypergraphs is constructed. Edges represent physical links between entities, and hyperedges represent information services logically composed of multiple entities. In the new model, the network efficiency within each service is comprehensively considered, and a more reasonable new measure for evaluating network service capability is proposed, called service efficiency. In the vulnerability stage of resilience, a local edge efficiency-betweenness removal strategy is proposed to identify important edges in the network, which can preserve the most performance by hardening them before an attack occurs. When the network suffers an inevitable attack, the recovery strategies of the central node based on global information and community information are proposed to restore the network performance with the fewest components. The former considers the degree of nodes in the entire hypergraph, and the latter considers the more closely connected hypergraph community structure. Simulation results under five sets of network parameters show that all our schemes apply in all cases. Hardening the links elected by the local edge efficiency-betweenness strategy in advance can enhance the network’s survivability. When the network is damaged, the recovery strategy of the central node based on global information is better under single-step recovery, while the recovery strategy of the central node based on community information is better under synchronous recovery. Using the proposed hardening and recovery strategies together, resiliency can be improved by up to 160%.

computer science, information systems,telecommunications,engineering, electrical & electronic