Fundamentals of Queueing Theory
Fundamentals of Queueing Theory
| Hi-index | 0.00 |
The paper deals with an end-to-end delay model in IP network and the proposed model is focused on time sensitive applications based on Real Time Protocol, especially IP telephony. In order to predict overall delay of datagrams in IP network we need to express a contribution of all individual components influencing the delivery of signal on the transmission path. Our work is focused on the relation between datagrams going through a two priority queue' serving system and consisting of sources with Poisson's probability distribution what is the case of voice traffic. While the first queue operates with higher priority, the datagrams in the second queue are served with the lower priority. We express this relation in proposed model and we have determined its validity in experimental testbed. The proposed model is suitable for the approximation of the traffic based on Real Time Protocol which consists of sources with the Poisson's probability distribution up to the 70 % of line load. If the traffic load goes over 70% of the line load, presented by interface in router, we were not able to reproduce tests reliably due to the unpredictable behaviour of call processing and loss. The results can contribute to the speech quality computational methods because the delay is a key parameter in speech quality assessment by a non-intrusive approach.