Improving Security in Multi-Agent Systems by Using a Novel Reputation Model

Abstract: Distributed Denial of Service (DoS) attacks are widely known attacks on networks that deny service over network. When such attacks use source address spoofing it is not easy to trace the source of attack. This is an open problem to be addressed. The existing traceback scheme employed either packet marking or packet logging approaches. However, for successful traceback, these schemes need large number of attack packets. To overcome this drawback Al-Duwairi and Manimaran presented hybrid traceback schemes which combine both approaches. The schemes include Distributed Link- List Traceback (DLLT) and Probabilistic Pipelined Packet Marking (PPPM). In this paper we implement the hybrid traceback schemes presented by them using a prototype application which demonstrates the proof of concept. The empirical results revealed that the proposed system traces back the source of attack effectively.

Keywords: IP Traceback, address spoofing, Denial of Service attacks