适于LEO卫星IP网络的源组播算法

为了解决LEO卫星IP网络中现有源组播算法的信道资源浪费问题，提出了一套新的组播算法。即基于核心群的特定源组播（CSSM）算法和加权的CSSM（ω-CSSM）算法。CSSM算法以源节点作为初始核心群，通过核心群和剩余组成员的最短路径方法逐步扩展直至整棵组播树构建完成，这样所得的树代价最小，从而大大提高了网络的传输带宽利用率和有效传输容量。在ω-CSSM算法中，所提出的加权因子可以调整组播树的树代价和端到端传播时延之间的折衷程度，因此，可以通过调整加权因子柬适度增大树代价、降低端到端传播时延，从而支持某些有严格端到端时延要求的实时组播业务。通过与LEO卫星IP网络中典型源组播算法MRA的仿真比较，可以看出CSSM和ω-CSSM算法的树代价性能比MRA有显著改善，而端到端传播时延略高。

Source-Based Multicast Algorithms for LEO Satellite IP Networks

To resolve the channel resources waste problem of current source-based multicast routing algorithms in low earth orbit(LEO) satellite IP networks,a new source-specific multicast(SSM) algorithm called core-cluster-based source-specific multicast(CSSM) algorithm and its improved one(i.e.the weighted CSSM algorithm) are proposed.The CSSM algorithm takes source node as initial core cluster,and extends it to construct entire multicast tree with the lowest tree cost step by step by a shortest path scheme between newly-generated core cluster and surplus group member nodes,which can greatly improve transport bandwidth utilization and efficient transport capacity.In the w-CSSM algorithm,a weighted approach is proposed to make tradeoff between tree cost and end-to-end propagation delay so that end-to-end propagation delay can be decreased to meet strict end-to-end delay requirements of some real-time multicast applications by adjusting a weighted factor at the expense of a bit increase of tree cost.Compared with a typical SSM algorithm in LEO satellite networks,i.e.multicast routing algorithm(MRA),tree cost performance of the CSSM and w-CSSM algorithms is better than that of MRA at the expense of a bit higher end-to-end propagation delay.