Network traffic anomaly detection based on self-similarity using FRFT

Since traditional abnormal detection methods have poor performance, and Hurst parameter estimation was affected by non-stationary traffic. An abnormal detection method based on Hurst parameter estimation using Fractional Fourier Transform (FRFT) was implemented. The experimental results show that FRFT estimation method was not affected by non-stationary time series and has better performance on Hurst estimation. We also verify the improvement in the network traffic anomaly detection.