深度包检测中一种正则表达式匹配算法的改进  

Improvement of regular expression matching algorithm for deep packet inspection

作　　者：张巍 陈佼[2] 赵美凯 

机构地区：[1]南方电网科学研究院有限责任公司,广东广州510080 [2]航天恒星科技有限公司,北京100086

出　　处：《现代电子技术》2015年第5期87-92,共6页Modern Electronics Technique

摘　　要：网络数据包内容检测技术已在网络安全、网络监视、HTTP负载均衡等方面得到广泛的应用,因此,对快速数据包内容的检测就变得异常重要。在数据包内容检测过程中,数据包的净载数据要通过一系列已经定义好的正则表达式模式进行数据匹配。在此,阐述目前数据包检测存在的问题,如传统数据包检测应用程序要求很大的内存空间去存储相应的正则表达式模式,提出一种大大降低对内存空间使用的改进算法。通过将该改进算法应用到以DFA为基础的包检测应用程序中,说明经过真实网络数据来检测算法的改进成果。结果表明了改进算法的有效性。Content inspection technology for network data packet has been widely used in network security, network monitoring, HTTP load balancing, etc. In content scanning of data packet, the payload data of the packet needs to be matched by a set of the specified regular expressions. The problem existing in the current data packet inspection is elaborated in this paper. That is, the traditional application program of data packet inspection needs a large memory space to store the corresponding regular expression pattern. A improved algorithm that effectively reduces memory space usage is proposed. The application of the improved algorithm in DFA-based packet inspection program indicates the improvement achievements, which were obtained by detection of the true network data.

关 键 词：正则表达式 深度包检测 DFA模式 内存使用 

分 类 号：TN915.083.4[电子电信—通信与信息系统] TP3[电子电信—信息与通信工程]