计算机工程

• 先进计算与数据处理 • 上一篇    下一篇

递归流分类算法研究与改进

余虎,黄宇   

  1. (合肥工业大学计算机与信息学院,合肥 230009)
  • 收稿日期:2014-12-01 出版日期:2015-12-15 发布日期:2015-12-15
  • 作者简介:余虎(1990-),男,硕士研究生,主研方向:流分类技术,网络安全;黄宇,硕士研究生。
  • 基金项目:
    安徽省自然科学基金资助项目“具有隐私保护的网络海量信息可信性度量若干关键技术研究”(11040606M138)。

Research and Improvement of Recursive Flow Classification Algorithm

YU Hu,HUANG Yu   

  1. (School of Computer and Information,Hefei University of Technology,Hefei 230009,China)
  • Received:2014-12-01 Online:2015-12-15 Published:2015-12-15

摘要: 流分类算法的性能直接影响防火墙、路由器等设备的处理速度。递归流分类(RFC)算法具有分类速度快的优点,但随着规则数目的增大,存储开销也随之增加。为此,通过对RFC算法进行分析,提出一种改进算法Optimize_RFC,对块的位数进行异或运算,压缩等价类表,减少内存 消耗。实验结果表明,Optimize_RFC算法在保持相对较快分类速度的同时,可降低预处理阶段的内存占用。

关键词: 流分类, 递归流分类算法, 等价类表, 异或运算, 内存优化

Abstract: Data flow classification algorithm performance directly affects the processing speed in firewalls,routers or other equipments.Recursive Flow Classification(RFC) algorithm has the advantage as faster packet classification,but also exists the disadvantage as huge memory consumption with the number of rules increases.By analyzing the shortcomings of RFC algorithm,this paper proposes an improved algorithm called Optimize_RFC optimizing the equivalence class table with bit XOR operation,reducing memory consumption.Experimental results show that Optimize_RFC algorithm can greatly reduce the memory consumption while maintaining a relatively high speed of classification in the preprocessing stage.

Key words: flow classification, Recursive Flow Classification(RFC) algorithm, equivalence class table, XOR operation, memory optimization

中图分类号: