Network Traffic Prediction and Applications Based on Time Series Model

Network traffic prediction is a very complex and difficult issue in the network management and design. This paper shows a model with a new algorithm (MLSL), and the model parameters can be modified by the new algorithm, which improves the adaptive ability of the model and makes the model adaptive function. Simulation and actual network traffic data experiment has proved that this algorithm has the advantage of high prediction accuracy and fast convergence, and its computing complexity is lower than other related algorithms.

[1]  Jun Lv,et al.  Network traffic prediction and fault detection based on adaptive linear model , 2004, 2004 IEEE International Conference on Industrial Technology, 2004. IEEE ICIT '04..

[2]  Walter Willinger,et al.  On the self-similar nature of Ethernet traffic , 1993, SIGCOMM '93.

[3]  Sally Floyd,et al.  Wide-Area Traffic: The Failure of Poisson Modeling , 1994, SIGCOMM.

[4]  Birsen Yazici,et al.  Kalman filtering for self-similar processes , 2006, Signal Process..

[5]  Giuseppe Martinelli,et al.  Efficient adaptive linear predictor , 1982 .

[6]  Oliver W. W. Yang,et al.  Prediction-based admission control using FARIMA models , 2000, 2000 IEEE International Conference on Communications. ICC 2000. Global Convergence Through Communications. Conference Record.

[7]  S. Haykin,et al.  Adaptive Filter Theory , 1986 .

[8]  Vern Paxson,et al.  Fast, approximate synthesis of fractional Gaussian noise for generating self-similar network traffic , 1997, CCRV.