一种基于改进流形学习方法的云计算入侵检测模型.pdfVIP

37 10 V ol. 37 N o. 10 2010 10 Computer Science Oct 20 10 ( 南京邮电大学计算机技术研究所 南京210003) ( 南京邮电大学计算机学院 南京210003) 基于互联网的超级计算模式云计算引起了人们极大的关注 也面临着越来越多的安全 胁主要构建能够 适应云计算环境的入侵检测系统框架将非线性流形学习算法引入本课题提出的模型 作为特征提取模块对云计算 环境下采集的网络数据进行预处理; 给出经典流形学习算法L LE 的改进研究 以提高后续分类性能实验表明 该算 法是可行和高效的 云计算 入侵检测 流形学习 T P393 A Novel Cloud Computing Intrusion Detection Model Based on Improved Manifold Learning CH E N Danw ei H OU Nan SU N Guozi ( Inst it ut e of Comput er T echnol ogy N anjing U ni v ersit y of P o st s and T elecommu nicati on s N an jing 2 10003 C hina) ( College of Comput er N an jing U niv er sit y of P o st s T el ecomm unicat ions N anjing 2 10003 China) Abstract T he clou d co mputing technolo gy w hich is a new Internetb ased sup ercomputing model h as arou sed g reat co ncer n f acing an incr easing number o f security thr eat s at the sam e tim e. T his pa per bu ilt a intru sio n detect ion system fram ew ork w hich is adapt to the clo ud comput ing envir onm ent pr oposed in this p aper . T he no nlinear m anif o ld lear ning alg or ithm w as intr oduced to the model proposed in t his thesis to be the co re alg or ithm of the data prepro cessing mod ule and the impr oved LL E algo rithm w as g iven in o rder to improv e the classif icatio n p erf o rm ance. T he simulation result sho w s th at the alg or ithm is eff ect ive and eff icient . Keywords Clo ud comput ing I ntr u sion detectio n M anif old learning [ 1]