Temporal Centrality-Balanced Traffic Management for Space Satellite Networks

Structural centrality metrics of the networks are important features to identify central nodes or links, to assess network vulnerability against removal of nodes or links, and to manage or control the network traffic to reduce the network congestion. Classic graph-based centrality metrics are not effective in dynamic space satellite networks due to the changing connectivity and they cannot effectively capture important temporal properties. In this paper, we employ a new technique called time-varying graphs (TVGs) to express temporal concepts and definitions of space satellite networks and propose several temporal centrality metrics in order to account for the temporal features of the networks. To calculate these temporal centrality metrics, we adopt a modified matrix multiplication algorithm to find the shortest journeys between nodes in the TVG. A temporal centrality-balanced traffic management scheme is further developed to enhance the network performance, which is integrated into a temporal centrality-balanced routing algorithm. To show the performance of the proposed centrality metrics and centrality balanced routing algorithm, we conduct some experiments performed on a scenario of the LEO satellite network. The simulation results show that the proposed centrality metrics are more powerful to analyze the evolution of time-varying centrality and vulnerability of space satellite networks than topological adaptation of vanilla centrality methods. Besides, the proposed centrality-balanced routing algorithm reaches the superior quality values of the network performance (in terms of several factors) in comparison with those attained by the traditional shortest journeys routing algorithm and proves to be more powerful to support the high traffic load.