A Prefiltering Approach to Regular Expression Matching for Network Security Systems

柳厅文   孙永  刘向阳  郭莉  方滨兴 



Abstract. Regular expression (RegEx) matching has been widely used in various networking and security applications. Despite much effort on this important problem, it remains a fundamentally difficult problem. DFA-based solutions can achieve high throughput, but require too much memory to be executed in high speed SRAM. NFA-based solutions require small memory, but are too slow. In this paper, we propose RegexFilter, a prefiltering approach. The basic idea is to generate the RegEx print of RegEx set and use it to prefilter out most unmatched items. There are two key technical challenges: the generation of RegEx print and the matching process of RegEx print. The generation of RegEx is tricky as we need to tradeoff between two conflicting goals: filtering effectiveness, which means that we want the RegEx print to filter out as many unmatched items as possible, and matching speed, which means that we want the matching speed of the RegEx print as high as possible. To address the first challenge, we propose some measurement tools for RegEx complexity and filtering effectiveness, and use it to guide the generation of RegEx print. To address the second challenge, we propose a fast RegEx print matching solution using Ternary Content Addressable Memory. We implemented our approach and conducted experiments on real world data sets. Our experimental results show that RegexFilter can speedup the potential throughput of RegEx matching by 21.5 times and 20.3 times for RegEx sets of Snort and L7-Filter systems, at the cost of
less than 0.2 Mb TCAM chip.
Keywords: regular expression, prefilter, RegEx print, TCAM.




首页
团队介绍
发展历史
组织结构
MESA大事记
新闻中心
通知
组内动态
科研成果
专利
论文
项目
获奖
软著
人才培养
MESA毕业生
MESA在读生
MESA员工
招贤纳士
走进MESA
学长分享
招聘通知
招生宣传
知识库
文章
地址:北京市朝阳区华严北里甲22号楼五层 | 邮编:100029
邮箱:nelist@iie.ac.cn
京ICP备15019404号-1