• DocumentCode
    3373460
  • Title

    A novel method of network traffic anomaly detection

  • Author

    Lianmin Hu

  • Author_Institution
    Dept. of Phys. & Electr. Eng., Leshan Teachers Coll., Leshan, China
  • Volume
    9
  • fYear
    2011
  • fDate
    12-14 Aug. 2011
  • Firstpage
    4757
  • Lastpage
    4759
  • Abstract
    The purpose of this paper is to propose a new algorithm which is based on combining the linear model and the method about smooth exponential. It is used to simulate the AR model of the Single sliding window sequence of observations and gets the average value of the square of slide window observed values´ residual noise and achieve the previous statistics on the latter statistic forecast with exponential smoothing method so as to decide whether the network traffic is normal or not. This algorithm is more efficiently comparing with GLR method and more reliable comparing with smooth exponential method, and has been proved is effectively in the detection for network traffic anomalies.
  • Keywords
    security of data; smoothing methods; AR model; GLR method; exponential smoothing method; linear model; network traffic anomaly detection; observation single sliding window sequence; residual noise; Correlation; Mathematical model; Noise; Prediction algorithms; Predictive models; Routing protocols; Smoothing methods; anomaly detection; network traffic; residual noise;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Electronic and Mechanical Engineering and Information Technology (EMEIT), 2011 International Conference on
  • Conference_Location
    Harbin, Heilongjiang
  • Print_ISBN
    978-1-61284-087-1
  • Type

    conf

  • DOI
    10.1109/EMEIT.2011.6024099
  • Filename
    6024099