Victim Based Statistical Filtering : A New Deterrent Against Spoofed DoS Traffic

Suhail Qadir Mir1 and S.M.K. Quadri2 

1 Post Graduate Department of computer sciences, University of Kashmir, India 

2Department of Computer Science, Jamia Millia Islamia, India 

ABSTRACT 

The extensive use of Internet and network based information resources on a global scale has led to the rise in a wide range of security incidents. One such attack is a TCP-SYN DoS attack, which makes use of IPSpoofing for its effectiveness. This paper presents a robust scheme for filtering spoofed DoS IP Packets in the Internet. We have proposed a robust filtering algorithm namely, Victim Based Statistical Filtering in this paper. The algorithm is inspired from the Hop-Count Filtering (HCF) method, which uses a corelation between IP addresses and their respective hop-counts to the destination server, to filter out the spoofed IP packets from the legitimate ones. The variation that we have proposed is adding the monitoring information of the usage levels of port numbers of the destination machine (victim), in the HCF Algorithm. The proposed VBSF algorithm was empirically evaluated and it was found to exhibit better performance than its predecessor.

KEYWORDS 

Hop Count, Flooding, TCP-SYN, DoS Attack, IP Spoofing, IP Filtering.

For more details..