Journal of Jilin University Science Edition

Previous Articles     Next Articles

Heterogeneous Parallel String Matching AlgorithmBased on Mobile Platform

LIU Lei1, LI Guangli1, XU Yue1, ZHANG Tongbo1, LV Shuai1,2,3   

  1. 1. College of Computer Science and Technology, Jilin University, Changchun 130012, China;2. College of Mathematics, Jilin University, Changchun 130012, China; 3. Key Laboratory of SymbolicComputation and Knowledge Engineering of Ministry of Education, Jilin University, Changchun 130012, China
  • Received:2016-05-13 Online:2017-01-26 Published:2017-02-02
  • Contact: LV Shuai E-mail:lus@jlu.edu.cn

Abstract: For common string matching problem in information processing, we analyzed the classical algorithms as Brute Force and KnuthMorrisPratt. According to the distribution characteristics of GPU heterogeneous parallel computing task, we designed a parallel scheme for data overlapping partition based on KnuthMorrisPratt algorithm, and proposed a heterogeneous parallel string matching algorithm KMP_MOP based on mobile platform. We tested the algorithm performance in PowerVR mobile platform by using strings with lengths of tens of millions. Also we compared execution of algorithm running on
other platforms, verifying the portability of the parallel algorithm. The experimental result indicates that the KMP_MOP algorithm can make full use of the GPU performance in mobile platform, and effectively improve the string matching efficiency of mobile platform devices with GPU.

Key words: mobile platform, string matching, heterogeneous parallel computing, overlapping partition

CLC Number: 

  • TP391