首页|层次化解析系统中两级节点协同的副本策略

层次化解析系统中两级节点协同的副本策略

Two-level Node Collaborative Replica Strategy in Hierarchical Resolution System

扫码查看
名字解析系统负责存储维护信息中心网络中名字和地址之间的映射关系,并提供名字解析服务.层次化的解析系统因可扩展性良好而备受关注,但解析请求一旦无法就近完成则需要沿其层次化结构进行长距离转发.当只有少数节点存储热门名字映射记录时,长解析时延的问题将更严重.副本策略是提高解析服务性能的重要机制,通常解析节点会向其他节点推送自身存储的映射关系从而扩散副本,但这种策略往往不能精准推送用户最感兴趣的名字,也难以控制副本数目.本文提出一种两级节点协同的副本策略,管理域中的高层级解析节点通过信息收集窗口获知服务范围内用户最感兴趣的名字后,将主动从其他节点请求副本.为底层节点分发副本时,以时延优化的边际收益衡量是否应为每个名字再多放置一个副本,从而自适应地决定不同名字的副本数目和位置.实验结果表明,本文的策略准确捕捉了用户的解析需求偏好,有效地提高了管理域内的解析命中率,解析时延相较于无副本时降低34.1%.
The name resolution system is responsible for storing the mapping between names and addresses,and providing name resolu-tion services in ICN.Hierarchical resolution systems are of interest because of good scalability,but requests need to be forwarded along the hierarchical structure over long distances.The problem of long resolution latency is exacerbated when only a few nodes store popular records.Replica strategy is an important mechanism to improve name resolution performance.Typically,a resolution node pushes its records to other nodes,but this strategy often fails to accurately push the names that users are most interested in,and it is also difficult to control the number of replicas.In this paper,we propose a two-level node collaborative replica strategy.After obtaining the names that are most interesting to users within the scope of the service,the high-level nodes will actively request replicas from other nodes and distribute them to the low-level nodes on the marginal benefits of placing each replica.Ultimately,the policy adaptively decides the number and location of replicas for different names.Experimental results show that our policy accurately captures user's preference,effectively improves the resolution hit rate in the management domain,and reduces resolution time by 34.1%.

name resolution systemICNreplica strategylatencynode collaboration

廉文瀚、王劲林、尤佳莉

展开 >

中国科学院声学研究所 国家网络新媒体工程技术研究中心 北京 100190

中国科学院大学 北京 100049

鹏城实验室 深圳 518055

名字解析系统 信息中心网络 副本策略 解析时延 节点协同

国家重点研发计划宽带通信与新型网络应用示范课题

2020YFB1806402

2024

网络新媒体技术
中国科学院声学研究所

网络新媒体技术

CSTPCD
影响因子:0.208
ISSN:2095-347X
年,卷(期):2024.13(3)
  • 2