Adaptive IP traceback mechanism for detecting low rate DDoS attacks

Abstract

Distributed Denial of Service (DDoS) attacks is measured to be a vital threat to the net. Because of memory less feature of the net routing mechanism, it's extraordinarily onerous to trace back to the supply of those attacks. Ancient packet marking techniques are not any longer applied because of its high memory consumption and poor measurability. During this paper, we have a tendency to build use of entropy variation to observe the attack that differentiates between traditional and DDoS attack traffic. However it will solely be applied to observe the attack during which rate of the traffic flow is incredibly high. There are also many attacks during which the strength of the attack is also meager, however the results of those attacks might be still severe. So it's essential to observe such attacks that square measure in lower order in terms of information rate. We have a tendency to propose an Adaptive Attack Detection System (AADS) that is capable of detection DDoS attacks regardless of the attack strength.