WDM业务疏导网络中多节点启动恢复机制研究
发布时间:2018-07-13 13:49
【摘要】:波分复用(Wavelength Division Multiplexing,WDM)技术的提出和发展使得网络带宽资源得到了充分的利用。业务量疏导技术将多个低速的业务请求汇聚到高速的波长通道上进行传输,有效地提高了网络资源利用率,同时降低了网络阻塞率。由于WDM业务量疏导网络承载了巨大的业务量,一旦网络发生突发故障,便会导致大量业务的中断,带来严重的后果。因此对WDM业务疏导网络生存性问题的研究极为重要。为了解决WDM业务疏导网络的生存性问题,本文研究了WDM业务疏导网络的动态恢复机制。完整的动态恢复机制包括重路由算法和恢复协议。本文首先研究了基于链路捆绑辅助图LBAG的动态疏导算法DGA L,并将其作为动态恢复机制的重路由算法;其次,研究并设计了双节点启动恢复协议BNRP和多节点启动恢复协议MNRP,将这两种恢复协议应用到疏导网络中,完整地设计了单链路故障双节点启动恢复机制。按照恢复粒度要求的不同,设计分析了连接级和光路级双节点启动恢复机制,C-BNRM和L-BNRM。之后,对C-BNRM进行扩展,设计了单链路故障多节点启动恢复机制S-MNRM;再次,考虑到随着网络规模的日益扩大,网络中存在发生双链路故障的可能性,本文针对网络发生双链路故障的情况分别设计了双链路故障双节点和多节点启动恢复机制,D-BNRM和D-MNRM。论文最后对设计的WDM业务疏导网络的动态恢复机制进行了仿真,仿真结果表明,C-BNRM相对于L-BNRM灵活性较强,C-BNRM的恢复成功率和资源利用率高于L-BNRM,且C-BNRM的故障恢复时间整体上短于L-BNRM;将C-BNRM进行扩展设计了S-MNRM,将这两种动态恢复机制与源节点启动恢复机制SIRM及传统的无疏导的动态恢复算法DPR进行比较,结果表明,S-MNRM的恢复性能最好,S-MNRM增加了恢复过程启动节点的个数,缩短了故障消息到达启动节点的传输时延,从而缩短了故障恢复时间。此外,业务量疏导技术的应用有助于提高动态恢复机制的故障恢复率,同时降低故障恢复时间;将D-BNR RM和D-MNRM进行比较的结果表明,D-MNR M在不增加故障恢复率、资源利用率及网络阻塞率的前提下,有效地缩短了故障恢复时间,具有良好的恢复性能。
[Abstract]:With the development of wavelength Division Multiplexing (WDM) technology, the network bandwidth resources are fully utilized. Traffic grooming technology converges multiple low-speed traffic requests to high speed wavelength channels for transmission, which effectively improves the utilization of network resources and reduces the network blocking rate. As the WDM traffic grooming network carries a huge amount of traffic, once the network has a sudden failure, it will lead to a large number of business disruption, which will bring serious consequences. Therefore, it is very important to study the survivability of WDM traffic grooming networks. In order to solve the survivability problem of WDM traffic grooming networks, the dynamic recovery mechanism of WDM traffic grooming networks is studied in this paper. The complete dynamic recovery mechanism includes rerouting algorithm and recovery protocol. In this paper, the dynamic grooming algorithm DGA Lbased on link binding auxiliary graph LBAG is studied, and it is regarded as the rerouting algorithm of dynamic recovery mechanism. A two-node startup recovery protocol (BNRP) and a multi-node start recovery protocol (MNRP) are studied and designed. The two recovery protocols are applied to the grooming network and a two-node startup recovery mechanism for a single link fault is designed. C-BNRM and L-BNRM are designed and analyzed according to different granularity requirements. After that, C-BNRM is extended to design a single-link fault multi-node startup recovery mechanism S-MNRM.The third, considering the possibility of double-link failure in the network with the increasing scale of the network, In this paper, the dual-link fault dual-node and multi-node startup recovery mechanisms, D-BNRM and D-MNRMrespectively, are designed for the case of double-link failure in the network. Finally, the dynamic recovery mechanism of WDM grooming network is simulated. The simulation results show that C-BNRM is more flexible than L-BNRM and C-BNRM has higher recovery success rate and resource utilization than L-BNRM.The fault recovery time of C-BNRM is shorter than that of L-BNRM.The C-BNRM is extended to design S-MNRMand the two dynamic recovery mechanisms and source sections are designed. The point start recovery mechanism (SIRM) is compared with the traditional ungrooming dynamic recovery algorithm (DPR). The results show that S-MNRM has the best recovery performance. S-MNRM increases the number of starting nodes in the recovery process, shortens the transmission delay of the fault message to start node, and thus shortens the time of fault recovery. In addition, the application of traffic grooming technique can improve the recovery rate of dynamic recovery mechanism and reduce the recovery time. The comparison between D-BNR RM and D-MNRM shows that D-MNR M does not increase the recovery rate. Under the premise of resource utilization and network blocking rate, the fault recovery time is shortened effectively and the recovery performance is good.
【学位授予单位】:东北大学
【学位级别】:硕士
【学位授予年份】:2014
【分类号】:TN929.1
[Abstract]:With the development of wavelength Division Multiplexing (WDM) technology, the network bandwidth resources are fully utilized. Traffic grooming technology converges multiple low-speed traffic requests to high speed wavelength channels for transmission, which effectively improves the utilization of network resources and reduces the network blocking rate. As the WDM traffic grooming network carries a huge amount of traffic, once the network has a sudden failure, it will lead to a large number of business disruption, which will bring serious consequences. Therefore, it is very important to study the survivability of WDM traffic grooming networks. In order to solve the survivability problem of WDM traffic grooming networks, the dynamic recovery mechanism of WDM traffic grooming networks is studied in this paper. The complete dynamic recovery mechanism includes rerouting algorithm and recovery protocol. In this paper, the dynamic grooming algorithm DGA Lbased on link binding auxiliary graph LBAG is studied, and it is regarded as the rerouting algorithm of dynamic recovery mechanism. A two-node startup recovery protocol (BNRP) and a multi-node start recovery protocol (MNRP) are studied and designed. The two recovery protocols are applied to the grooming network and a two-node startup recovery mechanism for a single link fault is designed. C-BNRM and L-BNRM are designed and analyzed according to different granularity requirements. After that, C-BNRM is extended to design a single-link fault multi-node startup recovery mechanism S-MNRM.The third, considering the possibility of double-link failure in the network with the increasing scale of the network, In this paper, the dual-link fault dual-node and multi-node startup recovery mechanisms, D-BNRM and D-MNRMrespectively, are designed for the case of double-link failure in the network. Finally, the dynamic recovery mechanism of WDM grooming network is simulated. The simulation results show that C-BNRM is more flexible than L-BNRM and C-BNRM has higher recovery success rate and resource utilization than L-BNRM.The fault recovery time of C-BNRM is shorter than that of L-BNRM.The C-BNRM is extended to design S-MNRMand the two dynamic recovery mechanisms and source sections are designed. The point start recovery mechanism (SIRM) is compared with the traditional ungrooming dynamic recovery algorithm (DPR). The results show that S-MNRM has the best recovery performance. S-MNRM increases the number of starting nodes in the recovery process, shortens the transmission delay of the fault message to start node, and thus shortens the time of fault recovery. In addition, the application of traffic grooming technique can improve the recovery rate of dynamic recovery mechanism and reduce the recovery time. The comparison between D-BNR RM and D-MNRM shows that D-MNR M does not increase the recovery rate. Under the premise of resource utilization and network blocking rate, the fault recovery time is shortened effectively and the recovery performance is good.
【学位授予单位】:东北大学
【学位级别】:硕士
【学位授予年份】:2014
【分类号】:TN929.1
【相似文献】
相关期刊论文 前10条
1 罗海丽;;含错误恢复机制的预测分析法的构造[J];阴山学刊(自然科学版);2007年01期
2 罗海丽;;含错误诊断恢复机制的算符优先法的构造[J];阴山学刊(自然科学版);2007年02期
3 徐洁;郭星廷;夏梦芹;;一种基于分段共享的预测式错误恢复机制[J];计算机科学;2003年12期
4 熊余;林艳芳;王汝言;;一种端节点启动的光网络快速恢复机制[J];微计算机信息;2008年18期
5 匡兴红;李军军;梁拥成;;WSN跟踪应用中一种新的目标恢复机制研究[J];科学技术与工程;2010年34期
6 刘铸,李钢,姚益平,季利春;联盟执行的保存恢复机制及其实现[J];计算机仿真;2004年08期
7 罗海丽;;语法分析中错误恢复机制的构造[J];内蒙古科技与经济;2007年14期
8 海深;陆阳;袁菲;;嵌入式存储系统恢复机制的设计与实现[J];计算机工程;2006年24期
9 刘方方;余洁;吴靖U,
本文编号:2119582
本文链接:https://www.wllwen.com/kejilunwen/wltx/2119582.html