基于智能流量预测的入侵检测方法  

A method for intrusion detection based on intelligent prediction to network traffic

作　　者：于宝华[1] 巩林明[1] 邓洪涛[1] 李伟[1] 张振国[2] 

机构地区：[1]石河子大学,新疆石河子832000 [2]陕西科技大学,陕西西安710021

出　　处：《微型机与应用》2011年第1期50-52,56,共4页Microcomputer & Its Applications

摘　　要：为了提高入侵检测系统检测的实时性,提出了一种基于智能流量预测的入侵检测方法。该方法拟合了智能Agent的智能性、自主性和自适应性的优点以及灰色预测对不确定资源的科学预测的优点,用流量预测智能Agent预测得到的流量序列来代替未来一段时间段内的实际流量,并把这个预测序列作为检测对象集的一部分。然后用人工方法模拟了流量处理Agent与预测智能Agent的活动,并通过对实际的采集数据进行仿真,实验证明了预测智能Agent的预测活动的科学性。In order to improve the real-time detection of intrusion detection system, an intrusion detection method based on intelligent prediction to network traffic is proposed. The merits of intelligence, autonomy and adaptability to the intelligent Agent, and the benefits of science on grey prediction to the uncertain resource are fitted in this method. The method uses the prediction sequence of the intelligent Agent for forecasting to substitute the actual network traffic in a future time period, and view this prediction sequence as part of the detection set to the intrusion detection system. Then the activities of intelligent Agents for data processing and forecasting to network traffic are simulated in an artificial way. And the scientific nature of the prediction method employed by intelligent prediction Agent is proved through a simulation experiment to the actual acquisition data.

关 键 词：灰色预测 小波变换 数据处理智能Agent 预测智能Agent 

分 类 号：TP301[自动化与计算机技术—计算机系统结构] TP393.0[自动化与计算机技术—计算机科学与技术]