IP网络路由交换与流量优化研究

发布时间：2018-05-19 01:33

本文选题：数据中心网络 + 多度量　；参考：《北京邮电大学》2014年硕士论文

【摘要】：IP数据中心网络是一个独立于运营商现有基础网络的专线网络,在这种专线网络与基础网络并存的新网络体系结构中,单纯采用Shortest-Path与Best-Effort模式可能带来丢包、延迟、抖动等问题,无法满足高Qos应用业务流。本文针对数据中心网络侧的路由交换与流量优化问题,在充分研究现有的基于网络参数信息存储方式的路由算法和多路径流量调度算法的基础上,提出了基于链路混合度量参数的多路径流量优化算法——MMSA (maximize mixing metric scheduling algorithm)算法和基于AHP的多度量参数加权流量优化算法——HP-WSA(Analytic Hierarchy Process Weighted Scheduling Algorithm)。文中主要研究两种业务数据流量：聚合流量(Aggregate Traffic)和显式流量(Explicit Traffic)。MMMSA主要用于聚合流量,针对ECMP(Equal-Cost Multipath Routing)的局限性(只对不同路径上流的数目进行平衡,而不是对可能引起超额链路出现拥塞的比特速率进行平衡),充分考虑流的大小,将流分为大型流与小型流,并基于混合度量参数在多路径之间实现调度优化；AHP-WSA主要用于显式流量,采用多属性决策的层次分析法,提出了基于多个度量的分布式多路径搜索算法,来发现源节点和宿节点之间的多条路径,通过两次运用层次分析法,根据多度量参数,从业务源节点到业务目的节点顺序建立多条不同权重的路径,将不同业务类型的流量调度到相应权重的路径上。两种方法分别在OPNET中进行了仿真验证,结果表明,数据业务流采用这两种方法后,端到端时延降低,吞吐量提高。可见本文提出的方法有效的减小了网路开销,并提高了网络性能,对IP数据中心网络路由交换与流量调度问题的进一步研究有一定价值。
[Abstract]:The IP data center network is a special line network independent of the existing basic network of the operators. In the new network architecture, which is coexisting with the basic network and the basic network, the simple use of Shortest-Path and Best-Effort patterns may bring about the problems of packet loss, delay, jitter and so on, which can not meet the high Qos application traffic.
Based on the existing routing algorithms and multipath traffic scheduling algorithms based on the existing network parameter information storage, this paper proposes a multi-path traffic optimization algorithm, MMSA (maximize mixing metric schedul), on the basis of the existing routing and multipath traffic scheduling algorithms based on network parameter information storage. Ing algorithm) algorithm and AHP based multi parameter weighted flow optimization algorithm HP-WSA (Analytic Hierarchy Process Weighted Scheduling Algorithm).
In this paper, we mainly study two kinds of traffic data traffic: Aggregate Traffic (Explicit Traffic) and explicit traffic (Explicit Traffic).MMMSA for aggregated traffic, for the limitation of ECMP (Equal-Cost Multipath Routing) (only the number of streams in different paths is balanced, not the bit speed that may cause congestion in the excess link. " The rate is balanced), the flow is fully considered, the flow is divided into large flow and small flow, and the scheduling optimization is realized between multi paths based on mixed metric parameters. AHP-WSA is mainly used for explicit flow, and the hierarchical analysis method of multi attribute decision making is used to find the distributed multi-path search algorithm based on multiple measures to discover source nodes and lodging. The multiple paths between nodes, through the two use of AHP, set up a number of different weight paths from the business source node to the business destination node in accordance with the multi measure parameters, and schedule the traffic of different business types to the corresponding weight paths. The two methods are simulated in OPNET respectively. The results show that the data are data. After the use of these two methods, the end to end delay is reduced and the throughput is improved. The proposed method can effectively reduce the network overhead and improve the network performance. It has some value for further research on Routing and traffic scheduling in IP data center network.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.06

【参考文献】