摘要
通过对MapReduce模型执行过程的分析,针对单源最短路径算法难以随着云计算的产生和发展而应用及提高搜索效率的问题,本文设计和实现了一种基于MapReduce架构的并行单源最短路径算法。并基于Hadoop平台集群环境进行了研究与实验,结果表明,文中算法可以有效地找出整个图结构中的单源最短路径,且验证了算法性能的优越性。
Via the analysis to implementation process of mapreduce, aimming at the problem that single source shortest path algorithm is hard to be used with the appearance and development of cloud computing and the problem of searching efficiency,a parallel single source shortest path algorithm based on mapreduce framework is designed and implemented .research and experiment are done based on hadoop platform.As shown by the experimental results,the proposed algorithm can search the single source shortest path efficiently in the whole graphic structure,and its good performance is testified.
出处
《微计算机信息》
2011年第12期97-99,101,共4页
Control & Automation