Abstract
In this paper, we presented an analysis of VoIP traffic on both call and burst levels. The analysed traces were based on a real connection statistics and ON/OFF characteristics of VoIP coders equipped with voice activity detector (VAD). We stated that an exponential distribution is not valid for an approximation of connection duration times as well as for an approximation of the ON/OFF characteristics of voice coders regardless of their type. We also generated synthetic multiplexed VoIP trace and we computed its second order statistics - autocorrelation and Hurst parameter which showed that the trace reveals long-range dependence and self-similarity.
| Original language | English |
|---|---|
| Pages (from-to) | 103-108 |
| Number of pages | 6 |
| Journal | Systems Science |
| Volume | 32 |
| Issue number | 3 |
| Publication status | Published - 2006 |
Keywords
- Long-range dependence
- Self-similarity
- VoIP modelling
ASJC Scopus subject areas
- Control and Systems Engineering
- Modeling and Simulation
- Computer Science Applications
Fingerprint
Dive into the research topics of 'Analysis of VoIP traffic produced by coders with VAD'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver