命名数据网络移动性机制研究

发布时间：2018-03-18 22:21

本文选题：内容中心网络　切入点：命名数据网络　出处：《解放军信息工程大学》2014年硕士论文　论文类型：学位论文

【摘要】：命名数据网络(Named Data Networking,NDN)作为一种典型的以内容为中心的未来互联网发展模式,以其全新的体系架构和独特的设计思路,直接对内容进行命名,实现网络对内容的高效分发,在下一代和未来互联网研究中具有广阔的应用前景。但是,由于NDN与现有TCP/IP互联网从体系结构上有较大差异,现有移动性支持方案难以支持NDN的新型网络环境,主要体现在:1)请求者移动将引入额外通信时延并且可能导致数据应答失败;2)分层的命名结构使数据源移动给网络带来巨大的更新开销;3)缺乏对移动NDN环境下应答包反向路径失效问题的考虑。为此,本文依托国家973计划项目“可重构信息通信基础网络体系研究”,针对命名数据网络中的移动性问题展开研究。本文在深入分析NDN基本工作原理和通信流程的基础上,对NDN的移动主体进行划分。然后,针对请求者的移动性、数据源的移动性以及移动NDN环境下的路由问题分别展开研究,分别提出了相应的移动性支持机制。主要工作和研究成果包括:1.提出了一种基于拓扑势的请求者移动性支持机制(TPB NDN)。该机制通过构建内容副本的拓扑势能,从为请求者选择最优数据源的角度出发,利用节点获取到的拓扑信息,结合缓存策略及拓扑势通告,实现对请求者移动性的支持。性能分析表明,该方法减小了内容请求时延,有效提高了节点的缓存命中率。2.提出了一种基于多代理的动态数据源管理机制(MMA NDN)。该方法将数据源按移动规律进行划分,分别采取不同管理模式在多个代理间协同管理,并合理进行内容缓存,充分利用网络特性减小数据源移动带来的切换时延,优化命名数据网络性能,实现对数据源移动性的支持。仿真结果表明,本方案能有效减小切换时延、降低切换开销和提高传输效率。3.提出了一种移动NDN环境下基于兴趣相似度的概率路由算法(PRAIS NDN)。该算法利用名字路由的特点,在兴趣包传递过程中选择兴趣相似度更高的节点进行转发,提高路由发现效率;在数据包回传的过程中,结合PIT表项信息和概率路由选择转发节点,并将内容在请求者处进行缓存,从而有效减小往返时延、降低网络开销率、提高传递效率和网络数据传输总量,实现NDN在移动环境中的应用。
[Abstract]:Named Data networking (NDN) is a typical content-centric future Internet development model. With its new architecture and unique design ideas, it names the content directly and realizes the efficient distribution of the content on the network. It has a broad application prospect in the next generation and future Internet research. However, due to the great differences between NDN and existing TCP/IP Internet in architecture, it is difficult for the existing mobility support schemes to support the new network environment of NDN. Mainly reflected in: 1) requester mobile will introduce extra communication delay and may cause data reply failure. The hierarchical naming structure causes data source mobility to bring huge updating overhead to the network. (3) lack of response to mobile NDN environment. Consideration of packet reverse path failure. Based on the National 973 Project "Research on Reconfigurable Information and Communication basic Network Architecture", this paper focuses on the mobility problem in named data network. Based on the in-depth analysis of the basic working principle and communication flow of NDN, Then, the mobility of the requestor, the mobility of the data source and the routing problem in the mobile NDN environment are studied. Corresponding mobility support mechanisms are proposed respectively. The main work and research results include: 1. A mobility support mechanism for requesters based on topological potential is proposed. This mechanism is based on topological potential energy of content replica. From the point of selecting the optimal data source for the requester, using the topology information obtained by the node, combined with the cache policy and the topological potential announcement, the mobility of the requestor is supported. The performance analysis shows that the method reduces the delay of the content request. The cache hit ratio of nodes is improved effectively. 2. A dynamic data source management mechanism based on multi-agent is proposed. The content cache is carried out reasonably, the handoff delay caused by the moving of data source is reduced by the network characteristic, the performance of named data network is optimized, and the mobility of data source is supported. The simulation results show that the scheme can effectively reduce the handoff delay. This paper presents a probabilistic routing algorithm based on interest similarity in mobile NDN environment, which utilizes the characteristics of name routing. In the process of packet transfer, we select the node with higher interest similarity to forward, improve the efficiency of route discovery; in the process of packet return, combine the PIT table item information and probability routing to select the forwarding node. The content is cached at the requester, which can effectively reduce the round-trip delay, reduce the network overhead rate, improve the transmission efficiency and the total amount of network data transmission, and realize the application of NDN in mobile environment.
【学位授予单位】：解放军信息工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.02

【参考文献】