利用多目标优化方法实现的GSON网拓扑构建算法  

Overlay Construction and Optimization Method with MOO

作　　者：赵星[1] 

机构地区：[1]陕西财经职业技术学院,陕西咸阳712000

出　　处：《新技术新工艺》2015年第3期54-58,共5页New Technology & New Process

摘　　要：覆盖网(Overlay)虚拟化能有效分离网络应用服务与底层网络基础设施,提升服务质量和用户体验,设计了一种利用多目标优化(multi-objective optimization,简称MOO)的通用覆盖网(general service overlay network,简称GSON)网拓扑构建算法(MOO-GSON)。该算法利用多目标优化模型提出了以服务节点和物理链路重用为约束,从而提高了虚拟拓扑与物理网络匹配度,最小化虚拟链路代价。降低了网络总体开销。采用随机网络拓扑进行试验仿真的结果表明,在满足各类约束的条件下,本算法能有效地优化时间和成本。Considering that the Overlay （network） virtualization can effectively separate network application service and the underlying network infrastructure, improve service quality and user experience. A way was designed by using a multi- objective optimization （multi - objective optimization, MOO）, a general network （GSON） network topology construction al- gorithm （MOO-GSON）, the algorithm using the multi-objective Optimization model is put forward to the service nodes and the physical link reuse as constraint, thus improve the virtual topology and the physical network compatibility, minimized the virtual link cost, reduces the network overhead as a whole. Using random network topology experiment simulation, the results showed that under the condition of satisfying various constraints, the algorithm can effectively optimize the time and costs.

关 键 词：异构网络 QOS参数 参数映射 映射模型 映射方法 Omnet 

分 类 号：TP393.2[自动化与计算机技术—计算机应用技术]