作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2013, Vol. 39 ›› Issue (7): 67-72. doi: 10.3969/j.issn.1000-3428.2013.07.015

• 先进计算与数据处理 • 上一篇    下一篇

一种基于动态组播树的可靠数据分发方法

邱 清,王意洁,马行空,李小勇   

  1. (国防科技大学计算机学院并行与分布处理国家重点实验室,长沙 410073)
  • 收稿日期:2012-07-04 出版日期:2013-07-15 发布日期:2013-07-12
  • 作者简介:邱 清(1988-),男,硕士研究生、CCF会员,主研方向:网络计算,数据分发技术;王意洁,教授、博士生导师; 马行空、李小勇,博士研究生
  • 基金资助:
    国家“973”计划基金资助项目(2011CB302601);国家自然科学基金资助项目(60873215);国家“863”计划基金资助项目(2011AA01A202); 高等学校博士学科点专项科研基金资助项目(200899980003);湖南省自然科学杰出青年基金资助项目(S2010J5050)

A Reliable Data Dissemination Method Based on Dynamic Multicast Tree

QIU Qing, WANG Yi-jie, MA Xing-kong, LI Xiao-yong   

  1. (National Key Laboratory for Parallel and Distributed Processing, School of Computer, National University of Defense Technology, Changsha 410073, China)
  • Received:2012-07-04 Online:2013-07-15 Published:2013-07-12

摘要: 已有的数据分发方法多侧重于某方面的优化,不能充分权衡数据分发的可靠性与高效性,难以满足应急场景下可靠快速分发数据的要求。为此,提出一种基于动态组播树的可靠数据分发方法。根据节点间延迟信息对节点进行层次化分簇和编号,以此构建层次化覆盖网,根据节点编号采用基于动态组播树的前缀匹配路由。理论分析和实验结果表明,与MEDYM方法相比,即使短时间内系统中产生大量消息,该方法仍能够可靠快速地分发数据,系统开销较小。

关键词: 数据分发, 应急场景, 动态组播树, 层次化分簇, 覆盖网, 前缀匹配

Abstract: Existing dissemination methods merely focus on single optimization objective and fail to ensure reliability and efficiency simultaneously, this makes them be inadaptable to meet the need of disseminating reliably and efficiently in the emergency environment. A reliable dissemination method based on dynamic multicast tree called RDBDMT is proposed. RDBDMT clusters and identify nodes in a hierarchical manner according to the delays between them, and based on which a hierarchical overlay is built. RDBDMT adopts prefix matching routing which is based on a dynamic multicast tree according to their identifications. Theoretical analysis and experimental results show that, RDBDMT is much more reliable and efficient than the existing methods, even in the emergency environment, where a large number of messages are published simultaneously in a very short time.

Key words: data dissemination, emergency environment, dynamic multicast tree, hierarchical clustering, overlay, prefix matching

中图分类号: