A hybrid defense mechanism for DDoS attacks using cluster analysis in MANET
P. Devi, A. Kannammal · 2012
Mobile Ad-hoc Networks (MANET) provides direct peer-to-peer communication between mobile nodes without any infrastructure. During data transfer, it is possible that an attacker may initiate Distributed Denial of Service (DDoS) that may interfere the entire communication. In this paper, we propose a Hybrid Defense Mechanism based on Cluster Analysis. The network traffic is investigated for the existence of DDoS using Cluster Analysis. The traffic thus investigated is subject to XOR marking which marks the non-legitimate traffic. This helps in identification of the node initiating the DDoS attack, thereby isolates the particular node from further communications. The experimentation is done with 2000 DARPA Intrusion Detection Scenario Specific Data Set to evaluate our method. From the experimental results, it is evident that our proposed scheme is effective in identification of DDoS attacks.