【作者】 孟维嘉；

【导师】 庞伟正；

【作者基本信息】 哈尔滨工程大学 ， 通信与信息系统， 2007， 硕士

【摘要】 随着互联网用户和流量的爆炸性增长，网络资源不足及网络拥塞问题日益严重，而各种新的多媒体业务的出现又对网络的服务质量提出了新的要求。流量工程就是为了解决这些问题而提出的，它希望在己有网络资源的基础上，通过优化网络资源利用，来达到减小拥塞、提高网络性能的目的。而MPLS被公认为是目前实现流量工程的最有力工具。本文的研究结论和主要创新：本论文主要研究了如何在MPLS网络中实现流量工程，并着重于研究路由问题。网络资源分配的不合理会导致网络节点状态发生变化，当这种变化超过一定范围时，选路算法所选择的路径就会产生错误更加剧了网络性能的恶化，从而导致拥塞产生。因此，本文主要从路由算法对网络资源的利用角度进行研究，通过为业务合理分配网络资源来保证网络资源的合理利用。(1)提出了一种基于量子遗传算法的优化网络资源利用的多约束QoS路由选择算法。本文将一种新的优化算法—量子遗传算法应用于解决QoS路由问题，并提出了将网络资源消耗和负载分布作为目标函数进行优化。此目标函数包含了带宽，跳数，费用，时延等QoS参数，力求在消耗网络资源最小的基础上，使负载尽量均衡分布，达到合理利用网络资源，降低网络的拥塞，提高网络吞吐量的目的。仿真计算的结果表明，该算法在解决多目标规划问题上是一个有效、可行的方法。(2)论文利用OPNET网络仿真软件设计了一个基于MPLS网络的流量工程分析试验，包含了从拓扑结构，业务源，链路到路由节点的所有模块。仿真比较了进行MPLS流量工程前后的网络模型。从结果中可以看到MPLS技术在网络的流量管理、负载均衡等方面都有着很大的效率，比“尽力而为”型的传统IP网络在使用网络资源、处理网络新业务上表现出了更大的优势。更多还原

【Abstract】 With the rapid growth of the network users and traffic, the problem of network resource lack and network congestion is more and more serious. But new multimedia applications require the network to provide quality of service guarantees. So traffic engineering is proposed. Its purposes are to minimize the congestion and improve the network performance through optimizing resource utilization of the existing networks. MPLS has been regarded as a powerful tool to traffic engineering.The results of study and main idea in the paper are given as follows:The paper mainly studies how to realize traffic engineering in the MPLS network, and emphatically in route question. The network resource distribution unreasonable can cause the node condition to change, when this kind of change surpasses certain scope, the routing algorithm chooses the way could make the mistake to intensify the network performance worsening, thus caused the jam production. Therefore, this article mainly research the route algorithm from the point of view of network resources use, which guarantee reasonable use of the network resources through rational distribution network resources.(1) Proposed one kind of QoS routing algorithm that based on the quantum genetic algorithm and optimizing network resource utilization. This article uses one kind of new optimized algorithm - quantum genetic algorithm to solve the QoS route question. We also proposed that the network’s resource expenditure and load distributed are optimized as object function, which include QoS parameters such as bandwidth, hop counts, cost, delay etc. The target is to balance load, make good use of network resource, reduce congestion of network, and increase throughout with minimized network resource. Simulation results indicated that the algorithm is efficient and reasonable.(2) The paper has designed one examination using the OPNET network simulation software based on the MPLS network traffic engineering. It has contained all modules from the topology, the service source, the link to route node. The simulation compared network model has carried on the MPLS traffic engineering and hasn’t. We can see from the result that MPLS network is more efficient than traditional IP network in traffic management and QoS implementation. MPLS can optimize performance of network.更多还原