Title :
QoS guarantee in a batched poisson multirate loss model supporting elastic and adaptive traffic
Author :
Moscholios, I.D. ; Vardakas, John S. ; Logothetis, M.D. ; Boucouvalas, A.C.
Author_Institution :
Dept. of Telecommun. Sci. & Technol., Univ. of Peloponnese, Tripolis, Greece
Abstract :
In this paper, we consider a single link that supports both elastic and adaptive traffic of Batch Poisson arriving calls, under the Bandwidth Reservation (BR) policy, whereby we can achieve specific QoS per service-class. Arriving batches have a generally distributed batch size, and can be serviced either as a whole or in part (partial batch blocking discipline), depending on the available link bandwidth. Blocked calls are lost. Accepted calls of a batch can compress or expand their bandwidth; elastic calls expand or compress their service time accordingly, while adaptive calls do not alter their service time. This system does not have a Product Form Solution. For the efficient calculation of time and call congestion probabilities as well as link utilization, we derive approximate but recursive formulas. The accuracy of the model is completely satisfactory and is verified together with the model´s consistency, through simulation. Comparison of the new model with existing models reveals its necessity.
Keywords :
Poisson distribution; quality of service; queueing theory; recursive estimation; telecommunication congestion control; telecommunication traffic; BR policy; QoS; accepted call compression; adaptive traffic; bandwidth reservation; batch Poisson arriving call; batch size distribution; batched Poisson multirate loss model; call congestion probability; elastic calls compression; elastic calls expand; elastic traffic; link utilization; model consistency; recursive formula; Adaptation models; Bandwidth; Markov processes; Numerical models; Probability; Quality of service; Simulation; Batched Poisson Process; Elastic-Adaptive Traffic; Recursive Formula; Reservation; Time-Call Congestion;
Conference_Titel :
Communications (ICC), 2012 IEEE International Conference on
Conference_Location :
Ottawa, ON
Print_ISBN :
978-1-4577-2052-9
Electronic_ISBN :
1550-3607
DOI :
10.1109/ICC.2012.6363934