Home // International Journal On Advances in Networks and Services, volume 6, numbers 3 and 4, 2013 // View article
Authors:
Ioannis Moscholios
John Vardakas
Michael Logothetis
Michael Koukias
Keywords: Markov chain; quasi-random process; elastic-adaptive traffic; recursive formula; time-call congestion; bandwidth reservation
Abstract:
In this paper, we propose a multirate teletraffic loss model of a single link that accommodates elastic and adaptive services whose calls come from a finite traffic-source population. This call arrival process is known as a quasi-random process and is used in traffic modelling when the number of users who generate traffic is relatively small compared to the system capacity. In-service elastic and adaptive calls can tolerate bandwidth compression by extending their remaining servicetime (elastic calls) or not (adaptive calls). In this loss system, we study the effect of the bandwidth reservation policy on time congestion probabilities, call congestion probabilities and link utilization. The bandwidth reservation policy is considered when a certain quality of service for each service-class is required and is essential to be guaranteed. The proposed model does not have a product form solution, and therefore we propose approximate but recursive formulas for the efficient calculation of the above mentioned performance measures. The accuracy and consistency of the proposed model are verified by simulation and is found to be quite satisfactory. Finally, we generalize the proposed model to include calls from both finite and infinite number of traffic sources.
Pages: 163 to 174
Copyright: Copyright (c) to authors, 2013. Used with permission.
Publication date: December 31, 2013
Published in: journal
ISSN: 1942-2644