摘要
虚拟网络映射问题是网络虚拟化研究中的核心问题之一,其主要目标是将虚拟网络请求高效地映射到底层物理网络上.针对面向多拓扑类型请求的虚拟网络映射问题进行研究,提出了节点连通性模型和通用底层物理网络节点评价模型用以判断节点映射的优先次序,在此基础上设计了一种复合型虚拟网络映射算法,在映射过程中通过识别虚拟网络请求的拓扑特征调用相应的映射子算法完成网络映射.仿真结果表明,该复合型映射算法获得了较高的虚拟网络请求接受率和网络收益开销比,整体上提高了虚拟网络映射性能.
As a core issue of network virtualization,virtual network embedding / mapping problem focus on allocate the virtual network requests onto the shared substrate network. Focusing on the virtual network embedding problem with multi-topology virtual network request,a node connectivity model and general substrate node measurement was proposed to rank the nodes in network. On the basic of node ranking,a new complex algorithm with topology recognition has been proposed,including several sub-algorithm for specific topology feature. The simulation results show that the complex algorithm improve the performance of embedding by increasing both the acceptance ratio of requests and revenue / cost ratio.
出处
《北京邮电大学学报》
EI
CAS
CSCD
北大核心
2015年第3期88-93,共6页
Journal of Beijing University of Posts and Telecommunications
基金
国家重点基础研究发展计划(973计划)项目(2012CB315801)
国家自然科学基金项目(61302089)
中央高校基本科研基金项目(2013RC0113)
关键词
网络虚拟化
虚拟网络映射
拓扑识别
拓扑特征
network virtualization
virtual network embedding
topology recognition
topology feature
