摘要
提出了一种新型网络入侵检测模型,在该模型中,首先将截获的数据包结合历史数据包数据库进行协议分析,找出可能存在的入侵行为的相关数据包,然后采用前馈多层感知器神经网络对这些相关的数据包进行回归分析,最终获得检测结果。该模型与传统采用单数据包检测方式的网络入侵检测系统(NIDS)模型相比,具有更低的漏检率。
A new network intrusion detection model was proposed. Based on the model, the currently data packet was integraed with historical data packet to process a protocol analysis, then the data packet that correlated with possiblely intrusion affair could be found out. The multilayer forward neural network was used to process a regression analysis to such data packets and got the results of intrusion detection. With the new model based on multiple data packet, the missed rate of traditional network intrusion detection system was decreased.
出处
《计算机应用》
CSCD
北大核心
2006年第4期806-808,共3页
journal of Computer Applications
基金
湖南省自然科学基金委资助项目(03JJY3104)
关键词
网络入侵检测系统
数据挖掘
前馈多层感知器
协议分析
Network Intrusion Detection System(NIDS)
data mining
Multiple-Level Perception(MLP)
protocol analysis