Research On The Invulnerability Of Space Information Networks Topology

With the increasing reliance on network communication technologies,space information networks(SINs)are gradually becoming a gladiatorial arena for various countries.The SINs are composed of space-segment(various spacecraft such as satellites),near-space segment(various near vehicles such as UAVs),and groundsegment(various gateways and terminals).In SINs,the number of nodes is huge,the movements of nodes are very dynamic,and the connections between nodes are diverse,resulting in the complex and changeable topology of the network.And the invulnerability of the network is much important,so the paper conducts invulnerability research on the topology of SINs.The study of the invulnerability of network topologies often starts from two aspects: first,invulnerability assessment metrics,and second,invulnerability optimization.In the paper,research is conducted around these two directions in the context of SINs.The main work is as follows:(1)An invulnerability assessment metric for SINs based on natural connectivity is proposed.Natural connectivity is an invulnerability evaluation metric in complex networks by measuring network redundancy,and its calculation formula is simple and has low complexity.However,because of the nodes moving quickly in the SINs,the topology changes frequently,so the natural connectivity cannot be used directly to measure the destructiveness.Therefore,the paper considers the proportion of link connection time of two satellites in one cycle as the weighted adjacency matrix and proposes weighted natural connectivity for SINs by combining the concept of natural connectivity.Simulation proves that the weighted natural connectivity is strictly monotonic,and can stably and correctly measure the invulnerability of SINs.(2)An invulnerability optimization model is developed based on weighted natural connectivity.The purpose of researching network invulnerability is to build a destructible network,so the paper from the weighted natural connectivity,and then establishes a multi-objective invulnerability model,which combined with constraints such as communication efficiency and maximum degree limitation in SINs.(3)Improved the algorithm of NSGA-?.According to the established model constraints,the paper improves the initial population and uses the VIKOR method to select the final solution from the Pareto optimal solution set.Simulation experiments show that the improvement of the initial population can reduce the blind search in the early stage of the algorithm,greatly reduce the number of iterations,and increase the convergence speed;the proposed model and algorithm are suitable for SINs and can be maintained in the excellent performances of static time slices and dynamic time periods;VIKOR can also accurately screen out the final solution from the Pareto optimal solution set.