无线Mesh网络负载均衡技术研究

发布时间：2018-02-03 10:10

本文关键词： 无线Mesh网络负载均衡网关部署路由协议安全机制入侵检测　出处：《吉林大学》2016年博士论文　论文类型：学位论文

【摘要】：作为下一代重要的无线网络接入技术,无线Mesh网络得到了快速的发展。然而,随着无线Mesh网络规模和应用领域的不断扩大,人们开始对无线Mesh网络的性能和网络安全性提出了越来越高的要求,逐渐成为国内外科研机构和广大学者研究的热点问题。负载均衡技术是提升无线Mesh网络吞吐量和Qo S性能的关键技术之一。特殊的网络结构决定了无线Mesh网络的负载均衡既包括网关节点的负载均衡也包括Mesh路由器间的负载均衡,以及入侵检测系统中检测点间的流量负载均衡。因此本文以优化、提升无线Mesh网络的性能为目标,以无线Mesh网络的负载均衡技术为主线,分别从无线Mesh网络的网关部署、路由协议和入侵检测系统三个方面展开相应的研究工作。针对无线Mesh网络的网关负载均衡问题,本文将数据空间聚类思想和遗传算法引入网关部署过程中,分别从无线Mesh网络的设计阶段和运行阶段对无线Mesh网络的网关部署策略进行设计和优化,分别提出了基于数据空间聚类的网关部署算法和基于遗传算法的网关部署优化策略。针对无线Mesh网络的路由器间的负载均衡问题,本文研究了实现Mesh路由器间流量负载均衡的分层混合路由协议,不同的分层采用不同的路由方式,并将节点的负载值作为路由判据,以此适应无线Mesh网络分层结构的特点,并能充分发挥各种路由方式的优点。针对无线Mesh网络入侵检测系统的负载均衡问题,本文分别从入侵检测系统的模型设计和入侵检测系统的分布式架构两方面着手,提出了基于支持向量机的入侵检测模型和基于检测点流量负载均衡的分布式入侵检测系统。从提高系统的检测精度和实现入侵检测系统的检测点间的流量负载均衡两个方面展开研究,以此来解决现有检测系统样本训练时间长、检测精度低和检测节点间负载失衡的问题,从而提升无线Mesh网络入侵检测系统的效率。仿真实验结果验证了上述几个算法的正确性及有效性,能够有效地从整体上提高无线Mesh网络的性能。
[Abstract]:As the next generation of important wireless network access technology, wireless Mesh network has been rapid development. However, with the expansion of wireless Mesh network scale and application field. People begin to put forward higher and higher requirements for the performance and network security of wireless Mesh network. Load balancing technology is to improve the throughput and QoS of wireless Mesh networks. The special network structure determines that the load balancing of wireless Mesh network includes not only the load balance of gateway node but also the load balance between Mesh routers. Therefore, this paper aims at optimizing and improving the performance of wireless Mesh network, and takes the load balancing technology of wireless Mesh network as the main line. The corresponding research work is carried out from three aspects: gateway deployment, routing protocol and intrusion detection system of wireless Mesh network, aiming at the problem of gateway load balancing in wireless Mesh network. This paper introduces the idea of data space clustering and genetic algorithm into the process of gateway deployment. The gateway deployment strategy of wireless Mesh network is designed and optimized from the design stage and running phase of wireless Mesh network. The gateway deployment algorithm based on data space clustering and the optimization strategy of gateway deployment based on genetic algorithm are proposed, respectively. The load balancing problem between routers in wireless Mesh networks is discussed. In this paper, the hierarchical hybrid routing protocol is studied to realize the traffic load balance between Mesh routers. Different layers adopt different routing methods, and the load value of nodes is taken as the routing criterion. In order to adapt to the characteristics of wireless Mesh network layered structure, and can give full play to the advantages of various routing methods, aiming at the wireless Mesh network intrusion detection system load balancing problem. This paper starts from two aspects: the model design of intrusion detection system and the distributed architecture of intrusion detection system. This paper presents an intrusion detection model based on support vector machine and a distributed intrusion detection system based on load balancing of detection points. The research is carried out in two aspects. In order to solve the problems of long training time, low detection accuracy and load imbalance between the existing detection system. In order to improve the efficiency of wireless Mesh network intrusion detection system, the simulation results verify the correctness and effectiveness of the above algorithms, and can effectively improve the performance of wireless Mesh network as a whole.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TN929.5

【相似文献】