Distributed Cluster-Based Fault-Tolerant Topology Control for Space Information Networks

Space information networks are a new type of self-organizing networks constituted by communication and information systems of land, sea, air and space. The performances of space information networks are closely related to their distinguishing characteristics such as large scale, high component complexity and high link redundancy. The paper proposed a distributed cluster-based and fault-tolerant topology control algorithm to construct a more efficient and failure-resistant topology for space information networks. Within each cluster, nodes discover local topology and compute appropriate transmitted powers reaching all nodes in their close vicinity through k optimal vertex-disjoint paths to achieve k-connectivity of the whole cluster. Between each pair of adjacent clusters, k disjoint links connecting the two clusters are maintained by finding out optimal matching of weighted bipartite graphs formed by topology of border nodes and their connections, which arranges inter-cluster relaying optimally. In the third phase, nodes adjust their transmitted power according to the logical topology generated by intra-cluster and inter-cluster topology control. Simulation results show performances such as invulnerability, reliability and validity are enhanced under the proposed topology control algorithm.