摘要
分析了现有主动式恢复方法的实现方式,并通过连续时间马尔可夫链(CTMC)对端到端恢复和本地恢复两种方式进行了建模和分析。在理论分析的基础上提出一种基于最短恢复路径的本地恢复的故障恢复方法,在单链路和单节点故障两种情形下,均可利用无环路的最短恢复路径重新连接因故障分离的子树。仿真结果表明,方法的故障恢复时间与现有"冗余树"和"双树"方法相比,分别减少了56.3%和35.1%左右,而故障恢复后组播树的代价与现有方法相当。
This paper systematically presented possible implementation approaches on multicast proactive recovery schemes and classified them into two catalogs： end-to-end recovery and local recovery. It modeled and analyzed these implementation approaches in continuous-time Markov chains （CTMC）. Based on theoretical results, this paper proposed an novel fast faultrecovery scheme of muhicast routing, which could effectively calculate a shortest and loop-free backup path to reconnect the isolated sub-trees due to fault, whether single link or single node failure cases. Simulation experiments show that failure recovery time of our scheme decreased by 56.3 % and 35.1% compared to ＂redundant tree＂ scheme and ＂dual-tree＂ scheme, the scheme cause less multicast tree cost increase after restoration than those schemes proposed previously.
出处
《计算机应用研究》
CSCD
北大核心
2010年第10期3950-3954,共5页
Application Research of Computers
基金
国家"863"计划资助项目(2007AA01Z2a1)
