Internet Traffic Forecasting Model Using Self Organizing Map and Support Vector Regression Method

Authors

Enrico Laoh, Industrial Engineering Department, Faculty of Engineering, Universitas Indonesia, Depok 16424, IndonesiaFollow
Fakhrul Agustriwan, Industrial Engineering Department, Faculty of Engineering, Universitas Indonesia, Depok 16424, Indonesia
Chyntia Megawati, Department of Integerated System Engineering, College of Engineering, Ohio State University, Columbus, OH 43210, USA
Isti Surjandari, Industrial Engineering Department, Faculty of Engineering, Universitas Indonesia, Depok 16424, Indonesia

Abstract

Internet traffic forecasting is one of important aspect in order to fulfill the customer demand. So, the service quality of internet service provider (ISP) can be maintained at the good level. In this study self-organizing map (SOM) and support vector regression (SVR) algorithm are used as forecasting method. SOM is first used to decompose the whole historical data of traffic internet into clusters, while SVR is used to build a forecasting model in each cluster. This method is used to forecast ISPs traffic internet in Jakarta and surrounding areas. The result of this study shows that SOM-SVR method gives more accurate result with smaller error value compared to that of the SVR method.

Bahasa Abstract

Model Perkiraan Trafik Internet Menggunakan Metode Self Organizing Map dan Support Vector Regression. Peramalan trafik internet merupakan salah satu hal yang penting agar permintaan konsumen dapat dipenuhi, sehingga kualitas pelayanan dari penyedia layanan dapat terjamin dengan baik. Pada penelitian ini, digunakan metode peramalan berupa kombinasi algoritma self-organizing map (SOM) dan support vector regression (SVR). Metode SOM digunakan untuk membagi data historis trafik internet secara keseluruhan ke dalam beberapa klaster, sedangkan metode SVR digunakan untuk membentuk model peramalan pada setiap klaster yang terbentuk. Penelitian dilakukan dengan mengaplikasikan metode peramalan SOM-SVR pada data trafik penyedia jasa internet di Jakarta dan sekitarnya. Hasil peramalan pada penelitian ini menunjukkan bahwa model peramalan dengan metode SOM-SVR dapat memberikan prediksi yang lebih akurat terkait nilai error yang lebih kecil dibandingkan dengan metode SVR tunggal.

References

1. A. Hardiansyah, S.I. Lestarininganti, T.N. Nizar, Universitas Komputer Indonesia, Bandung, 2013. [In Indonesian]
2. 3G Americas, Traffic Management Technique for Mobile Broadband Network, 3G Americas, Washington DC, 2011.
3. H. Abrahamsson, Internet Traffic Management, Malardalen University Press, Vasteras, 2008.
4. G.R. Ash, Traffic Engineering and QoS Optimization of Integrated Voice & Data Networks, Morgan Kaufmann, San Fransisco, 2006.
5. K. Papagiannaki, N. Taft, Z.L. Zhang, C. Diot, IEEE Trans. Neural Netw. 16 (2005) 1110.
6. D. Prangchumpol, Int. Schol. Sci. Res. Innovation. 7/7 (2013) 1141.
7. M. Lippi, M. Bertini, P. Frasconi, IEEE Trans. Intell. Transp. Syst. 14 (2012) 871.
8. P. Bermolen, D. Rossi, Computer Network. 53 (2009) 191.
9. X. Liu, X. Fang, Z. Qin, C. Ye, M Xie, J. Netw. Syst. Manag. 19/4 (2011) 427.
10. C.W. Hsu, C.C. Chang, C.J. Lin. A Practical Guide to Support Vector Classification, Department of Computer Science, 2003. Available at: http://www. csie.ntu.edu.tw/~cjlin/papers/ guide/guide.pdf.
11. T. Kohonen, Self-organizing Maps, Springer, New York, 2011.
12. S.H. Hsu, J.P.A. Hsieh, T.C. Chih, K.C. Hsu, Expert Syst. Appl. 36 (2009) 7947.
13. C.L. Huang, C.Y. Tsai, Expert Syst. Appl. 36 (2009) 1529.
14. J. Nagi, K.S. Yap, S.K. Tiong, S.K. Ahmed, The 2nd International Power Engineering and Optimization Conference, Shah Alam, Selangor, Malaysia, 2008.
15. V. Vapnik, The Nature of Statistical Learning Theory, Springer-Verlag, New York, 1995.
16. R. Jacobs, M. Jordan, S. Nowlan, G. Hinton, Neural Comput. 3 (1991) 79.
17. E. Alpaydin, Introduction to Machine Learning, MIT Press, Cambridge, 2004.
18. R.F. de Brito, A.L. Olivera, IEEE World Congress on Computational Intelligence, Brisbane, QLD, Australia., 2012, pp.1-7.
19. H. Feng, Y. Shu, S. Wang, M. Ma, Proceedings of the IEEE International Conference on Communications, Istanbul, Turkey, 2006, pp. 597-602.
20. J. He, Rethinking Traffic Management: Design of Optimizable Networks, Princeton University, New Jersey, 2008.
21. J. He, J. Rexford, M. Chiang, Algorithms for Next Generation Networks, 2010, pp.18.
22. N.V. Kalyankar, Network Traffic Management, J. Comput. 1/1 (2009) 191.
23. R. Kohavi, F. Provost, Glossary of Terms: Machine Learning, Kluwer Academic, Boston, 1998.

Recommended Citation

Laoh, Enrico; Agustriwan, Fakhrul; Megawati, Chyntia; and Surjandari, Isti (2018) "Internet Traffic Forecasting Model Using Self Organizing Map and Support Vector Regression Method," Makara Journal of Technology: Vol. 22: Iss. 2, Article 1.
DOI: 10.7454/mst.v22i2.3351
Available at: https://scholarhub.ui.ac.id/mjt/vol22/iss2/1