Hyper-Erlang Based Model for Network Traffic Approximation
The long-tailed distribution characterizes many properties of Internet traffic. The property is often modeled by Lognormal distribution, Weibull or Pareto distribution theoretically. However, it hinders us in traffic analysis and evaluation studies directly from these models due to their complex representations and theoretical properties. This paper proposes a Hyper-Erlang Model (Mixed Erlang distribution) for such long-tailed network traffic approximation. It fits network traffic with long-tailed characteristic into a mixed Erlang distribution directly to facilitate our further analysis. Compared with the well-known hyperexponential based method, the mixed Erlang model is more accurate in fitting the tail behavior and also computationally efficient.
KeywordsCumulative Distribution Function Expectation Maximization Expectation Maximization Algorithm Tail Behavior Erlang Distribution
Unable to display preview. Download preview PDF.
- 4.Arlitt, M., Jin, T.: A workload characterization study of the 1998 World Cup Web site. IEEE Network (Special Issue on Web Performance) 14(3), 30–37 (2000)Google Scholar
- 5.Mahanti, A., Williamson, C., Eager, D.: Traffic Analysis of a Web Proxy Caching Hierarchy. IEEE Network (Special Issue on Web Performance) 14(3), 16–23 (2000)Google Scholar
- 6.Kalyanaraman, S., Vandalore, B., Jain, R., et al.: Performance of TCP over ABR with Long-Range Dependent VBR Background Traffic over Terrestrial and Satellite ATM Networks. In: Proceedings of 23rd Annual Conference on Local Computer Networks (LCN 1998), Lowell, MA, October 1998, pp. 70–78 (1998)Google Scholar
- 7.Ata, S., Murata, M., Miyahara, H.: Analysis of Network Traffic and Its Application to Design of High-speed Routers. IEICE Transactions on Information and Systems E83-D(5), 988–995 (2000)Google Scholar
- 8.Asaka, T., Ori, K., Yamamoto, H.: Method of Estimating Flow Duration Distribution Using Active Measurements. IEICE Transactions on Communications E86-B(10), 3030–3037 (2003)Google Scholar
- 9.Downey, A.B.: Evidence for Long-tailed Distributions in the Internet. In: Proceedings of ACM SIGCOMM Internet Measurement Workshop 2001, San Diego, CA, USA (November 2001)Google Scholar
- 10.Greiner, M., Jobmann, M., Lipsky, L.: The Importance of Power-tail Distributions for Modeling Queueing Systems. Operations Research 47(2) (March-April 1999)Google Scholar
- 11.Shortle, J.F., Fischer, M.J., Gross, D., et al.: Using the Transform Approximation Method to Analyzed Queues with Heavy-Tailed Service. Journal of Probability and Statistical Science 1(1), 15–27 (2003)Google Scholar
- 16.Bilmes, J.A.: A Gentle Tutorial of the EM Algorithm and its Application to Parameter Estimation for Gaussian Mixutre and Hidden Markov Models. Technical Report, TR-97-021, International Computer Science Institue, Berkeley CA (April 1998)Google Scholar
- 18.Klemm, A., Lindemann, C., Lohmann, M.: Modeling IP Traffic Using the Batch Markovian Arrival ProcessGoogle Scholar
- 19.NASA HTTP traces (May 2004), http://ita.ee.lbl.gov/html/contrib/NASA-HTTP.html