Improved Detection Approach for Distributed Denial of Service Attack Based on SVM
Xiang Xu, Ding Wei, Yuelei Zhang · 2011
The intrusion detection rate is greatly influenced by the parameters of the support vector machine (SVM) model. In order to overcome the parameter limits to improve the identify accuracy of Distributed Denial of Service (DDoS) attack, this paper presents a new detection method based on Kernel Principle Component Analysis (KPCA) and Particle Swarm Optimization (PSO)-Support Vector Machine (SVM). The KPCA was used to obtain the important characteristics of the intrusion data to eliminate the redundant features. Then the PSO was used to optimize the SVM parameters. Experimental results show the proposed approach can enhance the detection rate, and performs better than the PCA based methods.