期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2015
卷号：4
期号：4
页码：1440-1447
出版社：Shri Pannalal Research Institute of Technolgy
摘要：In present years a major problem on today internet user is Denial of service (DOS) attack. In order to overcome from this attack, we proposed a technique called Multivariate Correlation Analysis (MCA). This can be used to extract the geometrical correlation between legitimate and the unknown attack of the network traffic features. Triangle Area method is used to enhance and speedup the process of Multivariate Correlation Analysis. This makes our solution capable of detecting the known and unknown attack. In many instance, Dos attack can be prevented, if the source IP address is traced back to its origin. Recently IP Trace Back mechanism called Packet Marking have been Proposed for achieving trace backs of Dos attack. In this paper, we show that probabilistic packet marking ¨C due to its efficiency it suffers under spoofing of marking field in the IP header by the attacker which can slow down the trace back by the victim. The proposed system can be evaluated using KDD Cup Data set.
关键词：Multivariate Correlation ; Analysis; Denial of Service; Triangle Area ; Method; Packet Marking; Probabilistic ; Packet Marking; KDD
