命名数据网络中一种基于节点分类的数据存储策略

发布时间：2018-11-02 08:09

【摘要】：缓存是命名数据网络(named data networking,NDN)有别于传统网络最突出的特性之一,NDN中默认所有节点都具有缓存所有经过数据的功能.这种"处处缓存"策略导致网内大量冗余数据的产生,使网内缓存被严重浪费.针对上述问题,首次提出了一种基于节点分类(based on node classification,BNC)的数据存储策略.基于节点位置的不同,将数据返回客户端所经过的节点分为"边缘"类节点与"核心"类节点.当数据经过"核心"类节点时,通过权衡该类节点的位置与数据在不同节点的流行度分布,将数据存储在对其他节点最有利的节点中;当数据经过"边缘"类节点时,通过该数据流行度来选择最有利于客户端的位置.仿真结果表明,提出的策略将有效提高数据命中率,减少数据请求时延和距离.
[Abstract]:Cache is one of the most outstanding features of named data network (named data networking,NDN), which is different from traditional network. By default, all nodes in NDN have the function of caching all passing data. This "everywhere cache" strategy leads to a large amount of redundant data in the network, which is a serious waste of cache in the network. To solve the above problems, a data storage strategy based on node classification (based on node classification,BNC (Node Classification) is proposed for the first time. Based on the difference of node location, the nodes passed by the data return client are divided into "edge" class node and "core" class node. When the data passes through the "core" class node, the data is stored in the node that is most favorable to other nodes by weighing the position of the node and the distribution of the data in different nodes. When the data passes through the Edge class node, the data popularity is used to select the location that is most beneficial to the client. Simulation results show that the proposed strategy can effectively improve the data hit rate and reduce the delay and distance of data request.
【作者单位】：重庆邮电大学光纤通信技术重点实验室;
【基金】：国家自然科学基金项目(61371096,61275077) 国家“九七三”重点基础研究发展计划基金项目(2012CB315803) 重庆市自然科学基金项目(cstc2013jcyjA40052) 重庆市教委科学技术研究项目(KJ130515)~~
【分类号】：TP393.02

【相似文献】