作者: Xinya Wu , Yonghong Chen
DOI: 10.1109/LCOMM.2013.102913.130932
关键词:
摘要: With the growth of Internet, one most common attacks comes in form a distributed denial service attack. In this paper, we validate that prediction error network traffic is chaotic which used NADA algorithm and improve DDoS detection based on NADA. Firstly, paper discusses use largest Lyapunov exponent to chaos hypothesis. Secondly, predict by using an exponential smoothing model instead forecasting method Then analyze theory Back Propagation Neural Network. The experimental results conducted discussed show our can detect up 98.04% accuracy.