信息中心网络路由算法的设计与仿真实现

发布时间：2018-03-05 00:21

本文选题：信息中心网络　切入点：朋友关系　出处：《东北大学》2014年硕士论文　论文类型：学位论文

【摘要】：近年来,随着用户的不断增长、网络应用的日趋丰富和技术水平的持续提高,互联网的终端、承载网络和应用已经发生了翻天覆地的变化,越来越多的应用关注以内容为中心的数据检索和服务访问,而当前网络是建立在端到端的主机通信上的,已经不能满足具有高扩展性和高效性的内容分发日益增长的需求。为了从根本上解决当前互联网面临的问题,国内外研究学者致力于信息中心网络的研究。NDN作为信息中心网络中的一种重要体系结构,其基于内容名称进行路由,关注"what"而非‘"where",可以有效地减少网络中数据冗余,提高内容检索速率。本文以NDN体系结构为基础,提出了信息中心网络路由算法。针对当前NDN中全转发路由策略易产生过多数据冗余的问题,本文提出了基于社交关系的路由算法和基于内容流行度的QoS感知的蚁群路由算法。将社会关系中的邻居关系和朋友关系应用到NDN路由中,在对FIB表进行最长前缀匹配之前执行邻居查找和朋友查找,提高路由的效率与成功率。依据历史经验信息建立朋友关系,考虑请求成功率、节点之间距离等因素计算朋友亲密度,指导兴趣包向更亲密的朋友节点进行路由请求。在数据包返回请求节点过程中,考虑内容流行度、节点自身及其邻居的缓存情况选择性地进行缓存,在缓存不足时优先替换流行度较低的内容条目。为了使系统具备智能群体的自组织性,能够利用局部信息,通过与其他节点之进行交互而完成全局寻路,本文借鉴蚂蚁觅食的寻路思想,通过发送兴趣蚂蚁更新请求节点到数据源节点之间路径上的信息素值,建立从请求节点到数据源的最优路径。本文考虑带宽、延迟、延迟抖动等QoS参数计算自适应信息素更新和挥发公式,避免路由算法陷入局部最优。依据节点中内容的访问频度计算内容流行度,使网络中更流行的内容具有更频繁的路径更新速率,从而使更多的路由请求从最优的数据源节点获取内容。考虑节点中的缓存容量、兴趣包的访问频率等设计缓存策略和基于效用值的缓存替换策略。为了验证本文提出的信息中心网络路由算法的可行性和有效性,本文基于Visual Studio平台,在cernet2拓扑上进行了仿真实现和性能分析。仿真结果表明,本文设计的基于社交关系的路由算法和基于内容流行度的QoS感知的蚁群路由算法在提高网络成功率、减少路由时间、提高缓存命中率、减少网络数据冗余方面都具有较好的性能。
[Abstract]:In recent years, with the continuous growth of users, the increasingly rich network applications and the continuous improvement of technology, the terminals, bearer networks and applications of the Internet have undergone earth-shaking changes. More and more applications focus on content-centric data retrieval and service access, and the current network is based on end-to-end host communications. Can no longer meet the growing demand for highly scalable and efficient content distribution. As an important architecture in the information center network, researchers at home and abroad devote themselves to the research of the information center network. Its routing based on the content name and paying attention to "what" instead of "where" can effectively reduce the data redundancy in the network. This paper proposes an information center network routing algorithm based on NDN architecture. Aiming at the problem that the full forwarding routing strategy in current NDN is prone to generate too much data redundancy, This paper proposes a routing algorithm based on social relations and an ant colony routing algorithm based on QoS awareness based on content popularity. The neighbors and friends in social relations are applied to NDN routing. In order to improve the efficiency and success rate of routing, the FIB table performs neighbor lookup and friend lookup before matching the longest prefix. Based on the historical information, the friend relationship is established, and the request success rate is considered. Some factors such as distance between nodes calculate the affinity of friends and instruct the interest packet to route the request to the closer friend node. In the process of returning the packet to the requesting node, the popularity of the content is considered. In order to make the system have the self-organization of intelligent group, it can make use of local information in order to selectively cache itself and its neighbors, and replace the content items with low popularity when the cache is insufficient. By interacting with other nodes to complete the global path finding, this paper uses the idea of searching for food from ants to send interested ants to update the pheromone values on the path between the request node and the data source node. The optimal path from the request node to the data source is established. In this paper, the adaptive pheromone update and volatilization formulas are calculated by considering the QoS parameters such as bandwidth, delay, delay jitter, etc. The content popularity is calculated according to the access frequency of the content in the node, so that the more popular content in the network has a more frequent path update rate. Thus allowing more routing requests to get content from the optimal data source node. Consider the cache capacity in the node, In order to verify the feasibility and effectiveness of the proposed routing algorithm of information center network, this paper is based on Visual Studio platform. The simulation results show that the proposed routing algorithm based on social relations and the ant colony routing algorithm based on QoS awareness based on content popularity can improve the network success rate and reduce the routing time. Improving cache hit ratio and reducing network data redundancy have better performance.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.08;TP18

【相似文献】