网络流量识别中的DFA优化技术
首发时间:2015-04-07
摘要:网络流量检测中的深度包检测(Deep Packet Inspection , DPI)采用正则表达式匹配算法,正则表达式匹配算法是一种多模式匹配特征匹配算法,采用确定性有限自动机(Deterministic Finite Automaton,DFA)表示一组正则表达式特征,实现一次内容扫描可匹配多个特征。基于硬件的正则表达式匹配算法面临存储空间需求大等挑战。近年来,Smith等人提出了一种基于扩展有限自动机(extended Finite Automaton,XFA)的正则表达式匹配算法,即在状态上增加辅助变量和简单操作指令,消除了DFA状态空间爆炸问题,从状态方面减少存储空间需求。为了进一步减少XFA存储空间需求,本文提出一种新的基于XFA的匹配算法以提高匹配性能;同时对特征进行分级存储以确保匹配的效率。
For information in English, please click here
The optimization techniques of the DFA with the network traffic classification
Abstract:Network traffic classification of Deep Packet Inspection (Deep Packet Inspection, DPI) using a regular expression matching algorithm, a regular expression matching algorithm is a kind of multiple pattern matching feature matching algorithm, using Deterministic Finite Automaton (Deterministic Finite Automaton, the DFA) represents a set of regular expression characteristics, and implement a content scan can match multiple characteristics.The regular expression matching algorithm which is based on hardware faced storage space challenges .In recent years, Smith and others put forward a kind of based on extended Finite Automaton (extended Finite Automaton, XFA) of a regular expression matching algorithm, namely the increasing on the state of auxiliary variables and simple instructions, eliminating the DFA state space explosion problem, from the aspects of state to reduce the storage space requirements.In order to further reduce the XFA storage space requirements, this paper presents a new matching algorithm based on XFA to improve matching performance.At the same time the characteristics of hierarchical storage in order to ensure the efficiency of matching.
Keywords: Network traffic classification Regular expression DFA
基金:
论文图表:
引用
No.4634479104091814****
同行评议
共计0人参与
勘误表
网络流量识别中的DFA优化技术
评论
全部评论0/1000