一种基于SSNFF的DDOS入侵监测模型  

A DDOS intrusion detection model based on SSNFF

作　　者：徐勇[1] 陶新民[1] 

机构地区：[1]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001

出　　处：《应用科技》2008年第9期14-17,共4页Applied Science and Technology

摘　　要：针对分布式拒绝服务攻击(DDOS)网络攻击,提出了一种新的基于小波分析的空间相关性选择正常流算法(SSNFF)的入侵检测模型.该模型利用SSNFF算法进行DDOS信号的提取,对提取的DDOS信号利用基于极大曲线长度阈值的去噪算法与阶越点判定算法进行入侵发生点的检测,克服了原有利用模极大值检测入侵点算法中不同幅值突变大小对检测性能的影响,进而给出一个使用中的判定模型,并对实际采集到的网络流量和仿真攻击流量的混合流做了计算机模拟验证,最后给出试验结果并进行了分析.A model for intrusion detection based on spatially selection normal flow filtration （SSNFF） is presented to deal with the network attacks such as distributed denial of service （DDOS） signals. It uses SSNFF algorithm to extract DDOS signal. Then a detection for intrusion node is performed using both denoising algorithm based on maximal curve length threshold and step node determination algorithm, thus eliminating the influence of amplitude variations on detection performance in the algorithm based on maximal modulus. Finally a model of DDOS intrusion detection is given. Computer simulations were made on the detection method for a real data set involving the normal web traffic collected from web server plus the DDOS attack flow. Some results are reported with relevant concluding remarks.

关 键 词：入侵检测 SSNFF DDOS 极大曲线长度阈值 模极大值 阶越点 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]