随机模式匹配并行算法在工作站机群上的实现  

Implementation of parallel algorithms for randomized pattern matching on cluster of workstations

作　　者：薛淞文[1] 申卫昌[1] 剡公孝[1] 乔龙[1] 

机构地区：[1]西北大学信息科学与技术学院,西安710127

出　　处：《计算机工程与应用》2010年第21期129-131,共3页Computer Engineering and Applications

摘　　要：对随机模式匹配算法进行了改进,并根据MPICH并行编程环境中任务间通信的特点,设计了一种基于MPICH的改进的随机模式匹配并行算法。根据运行在COW(工作站机群)上的进程数目将文本串进行重叠划分,每个进程完成一个文本子串的模式匹配。实验结果表明,该改进的随机模式匹配并行算法有效地加快了模式匹配的速度,提高了工作站机群的资源利用率。A newly improved pattern matching algorithm is designed by analyzing randomized algorithm for pattern matching and an effective parallel algorithm for pattern matching based on MPICH is designed according to the features of task communication on MPICH parallel programming environment.Text string is divided into overlapped sub-string according to the quantity of process in COW（Cluster of Workstations）,and each process executes pattern matching parallelly.The experimental results show that the pattern matching speed is accelerated and the utilization of resources in COW is also improved.

关 键 词：模式匹配 消息传递编程标准的一种实现(MPICH) 并行算法 工作站机群 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]