VoIP Source Model based on the Hyperexponential Distribution

In this paper we present a statistical analysis of Voice over IP (VoIP) packet streams produced by the G.711 voice coder with voice activity detection (VAD). During telephone conversation, depending whether the interlocutor speaks (ON) or remains silent (OFF), packets are produced or not by a voice coder. As index of dispersion for both ON and OFF times distribution was greater than one, we used hyperexponential distribution for approximation of streams duration. For each stage of the hyperexponential distribution, we tested goodness of our fits using graphical methods, we calculated estimation errors, and performed Kolmogorov-Smirnov test. Obtained results showed that the precise VoIP source model can be based on the five-state Markov process.

Authors:

• Arkadiusz Biernacki

Keywords:

• VoIP source modelling
• distribution approximation
• hyperexponential distribution.

References:

[1] K. Mitchell, J. Place, and A. Liefvoort, "Analytic Modeling with Matrix
Exponential Distributions," A Simulation Councils Proceedings Series,
vol. 28, 1996.
[2] P. T. Brady, "A technique for investigating on-off patterns of speech,"
Bell System Technical Journal, vol. 44, pp. 1-22, 1965.
[3] P. T. Brady, "A statistical analysis of on-off patterns in 16
conversations," Bell System Technical Journal, vol. 47, pp. 73-91, 1968.
[4] P. T. Brady, "A model for generating ON-OFF speech patterns in twoway
conversation," Bell System Technical Journal, vol. 48, pp. 2445-
2472, 1969.
[5] J. Gruber, "A Comparison of Measured and Calculated Speech Temporal
Parameters Relevant to Speech Activity Detection," Communications,
IEEE Transactions on [legacy, pre - 1988], vol. 30, pp. 728-738, 1982.
[6] F. Barcelo, "Statistical properties of silence gap in public mobile
telephony channels with application to data transmission," presented at
IEEE International Conference on Communications, (ICC 2001),
Helsinki, 2001.
[7] J. Wenyu and H. Schulzrinne, "Analysis of on-off patterns in VoIP and
their effect on voice traffic aggregation," 2000.
[8] H. Schulzrinne, "Voice Communication Across The Internet: A Network
Voice Terminal," Dept. of Computer Science, University of
Massachusetts, Amherst TR 92-50, 1992.
[9] E. Casilari, H. Montes, and l. F. Sandova, "Modelling of Voice Traffic
Over IP Networks," presented at 3rd Internacional Symposium on
Communication Systems, Networks and Digital Signal Processing
(CSNDPS 2002), Stafford 2002.
[10] B. Bellalte, M. Oliver, and D. Rincon, "Capacity and traffic analysis of
voice services over GPRS mobile networks," Technical University of
Catalonia, University Pompeu Fabra 2002.
[11] T. D. Dang, B. Sonkoly, and S. Molnár, "Fractal Analysis and Modelling
of VoIP Traffic," presented at NETWORKS 2004, Vienna, Austria,
2004.
[12] A. Biernacki, "Statistical analysis of VoIP streams (available in Polish
only)," presented at 7th Conference ÔÇ×Internet - Wroclaw 2005",
Wroclaw, 2005.
[13] N. Blefari-Melazzi, J. Daigle, and N. M. Femminella, "Stateless
Admission Control for QoS Provisioning for VoIP in a DiffServ
Domain," presented at ITC 18th, Berlin, 2003.
[14] M. Narbutt and L. Murphy, "VoIP Playout Buffer Adjustment using
Adaptive Estimation of Network Delays," presented at 18th International
Teletraffic Congress (ITC-18), 2003.
[15] Vox-Gratia, "OpenH323," 2005.
[16] E. S. Inc., "Ethereal, A Network Protocol Analyzer," 2005.
[17] S. Asmussen, O. Nerman, and M. Olsson, "Fitting phase-type
distributions via the EM algorithm," Scandinavian Journal of Statistics,
pp. 419-441, 1996.
[18] O. Haggstrom, S. Asmussen, and O. Nerman, "EMPHT - a program for
fitting phase-type distributions," Technical report, Department of
Mathematics, Chalmers University of Technology, Goteborg 1992.
[19] R. B. D'Agostino and M. A. Stephens, Goodness-of-Fit Techniques:
Marcel Dekker, 1986.