Validation of Chaos Hypothesis in NADA and Improved DDoS Detection Algorithm

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