Title :
Prediction for Long Range Bursty Traffic
Author :
Wen, Yong ; Zhu, Guangxi ; Xie, Changsheng
Author_Institution :
Coll. of Comput. Sci., Huazhong Univ. of Sci. & Technol., Wuhan
Abstract :
Fractional AutoRegressive Integrated Moving Averag (FARIMA) processes with alpha-stable innovation can capture non-Gaussian, namely heavy tailness that is the key factor of the long range burstiness of self-similar traffic. A FARIMA process can be regarded as ARMA process driven by fractional differencing (FD) process. ARMA processes with infinite variance can be simulated with recurrent neural network (RNN) instead of conventional Least Squares methods. We adopt three intelligent methods to train the weights of RNN in order to minimize the dispersion. The final predicted values are combined previous three predicted values. Prediction experimental results for the actual traffic trace show that the three FARIMA predictors are efficient, the compound predictors are more accurate.
Keywords :
autoregressive moving average processes; local area networks; recurrent neural nets; telecommunication traffic; ARMA process; alpha-stable innovation; fractional autoregressive integrated moving average processes; fractional differencing process; heavy tailness; infinite variance; long range bursty traffic prediction; recurrent neural network; 1f noise; Computer science; Displays; Educational institutions; Predictive models; Recurrent neural networks; Software engineering; Technological innovation; Telecommunication traffic; Traffic control; long range bursty; prediction; traffic;
Conference_Titel :
Computer Science and Software Engineering, 2008 International Conference on
Conference_Location :
Wuhan, Hubei
Print_ISBN :
978-0-7695-3336-0
DOI :
10.1109/CSSE.2008.1456