5 结束语
本文提出一种基于主成分分析的方法来划分子空间，分析和发现网络中的异常事件。本方法能够准确快速地标定异常发生的时间点，从而帮助网络安全应急响应部门及时发现宏观网络的流量异常状况，为迅速解决网络异常赢得时间。试验表明，我们采用的14个特征构成的分析矩阵具有较好的识别准确率和分析效率，我们接下来将会继续寻找更具有代表性的特征来构成数据矩阵，并研究更好的特征矩阵构造方法来进一步提高此方法的识别率，并将本方法推广到短时分析中。
6 参考文献
[1]XUK,ZHANG Z L, BHATTACHARYYA S. Profiling Internet backbone traffic: Behavior models and applications [C]// Proceedings of ACM SIGCOMM, Aug 22- 25, 2005, Philadelphia, PA, USA. New York, NY,USA:ACM,2005:169-180.
[2]HAWKINSDM, QQUI P, KANG C W. The change point model for statistical process control [J]. Journal of Quality Technology,2003, 35(4).
[3]THOTTANM,JI C. Anomaly detection in IP networks [J]. IEEE Transactions on Signal Processing, 2003, 51 )8):2191-2204.
[4]BARFORDP,KLINE J, PLONKA D, et al. A signal analysis of network traffic anomalies [C]//Proceedings of ACM SIGCOMM Intemet Measurement Workshop (IMW 2002), Nov 6-8, 2002, Marseilles, France. New York, NY,USA:ACM, 2002:71-82.
[5]LAKHINAA,CROVELLA M, DIOT C. Mining anomalies using traffic feature distributions [C]// Proceedings of SIGCOMM, Aug 22-25, 2005, Philadelphia, PA, USA. New York, NY,USA: ACM, 2005: 217-228.
[6]LAKHINAA,CROVELLA M, DIOT C. Diagnosing network-wide traffic anomalies [C]// Proceedings of ACM SIGCOMM, Aug 30 - Sep 3, 2004, Portland, OR, USA. New York, NY,USA: ACM, 2004: 219-230.
[7]SCHWELLERR,GUPTA A, PARSONS E, et al. Reversible sketches for efficient and accurate change detection over network data streams [C]//Proceedings of ACM SIGCOMM Internet Measurement Conference (IMC’04), Oct 25-27, 2004, Taormina, Sicily, Italy. New York, NY,USA: ACM, 2004:207-212.
[8]MAHONEYMV, CHAN P K. Learning rules for anomaly detection of hostile network traffic [C]// Proceedings of International Conference on Data Mining (ICDM’03), Nov 19-22, Melbourne, FL, USA . Los Alamitos, CA, USA: IEEE Computer Society, 2003:601-604.
作者简介：
王明华，国家计算机网络应急技术处理协调中心工程师，博士。先后参加4项国家“863”计划项目、国家信息安全专向研究项目和国家自然科学基金项目。已发表论文6篇。主要研究领域为互联网网络安全应急响应技术与宏观网络监测与预警。
