Heavy-traffic analysis through uniform acceleration of transitory queues with diminishing populations

Onderzoeksoutput: Boek/rapportRapportAcademic

36 Downloads (Pure)

Uittreksel

We consider the ¿(i)/GI/1 queue, in which the arrival times of a fixed population of n customers are sampled independently from an identical distribution. This model recently emerged as the canonical model for so-called transitory queues that are non-stationary, time-varying and might operate only over finite time. The model assumes a finite population of customers entering the queue only once. This paper presents a method for analyzing heavy-traffic behavior by using uniform acceleration, which simultaneously lets the population n and the service rate grow large, while the initial resource utilization approaches one. A key feature of the model is that, as time progresses, more customers have joined the queue, and fewer customers can potentially join. This {\it diminishing population} gives rise to a class of reflected stochastic processes that vanish over time, and hence do not have a stationary distribution. We establish that, by suitably rescaling space and time, the queue length process converges to a Brownian motion on a parabola, a stochastic-process limit that captures the effect of a diminishing population by a negative quadratic drift. The stochastic-process limit provides insight into the macroscopic behavior (for n large) of the transitory queueing process, and the different phenomena occurring at different space-time scales.
Originele taal-2Engels
Uitgeverijs.n.
Aantal pagina's43
StatusGepubliceerd - 2014

Publicatie series

NaamarXiv
Volume1412.5329 [math.PR]

Vingerafdruk

Heavy traffic
Queue
Stochastic processes
Join
Queueing
Time-varying
Resource utilization
Time scales
Rescaling
Stationary distribution
Brownian motion

Citeer dit

@book{06edf2049df64944a5059c81eec25a2d,
title = "Heavy-traffic analysis through uniform acceleration of transitory queues with diminishing populations",
abstract = "We consider the ¿(i)/GI/1 queue, in which the arrival times of a fixed population of n customers are sampled independently from an identical distribution. This model recently emerged as the canonical model for so-called transitory queues that are non-stationary, time-varying and might operate only over finite time. The model assumes a finite population of customers entering the queue only once. This paper presents a method for analyzing heavy-traffic behavior by using uniform acceleration, which simultaneously lets the population n and the service rate grow large, while the initial resource utilization approaches one. A key feature of the model is that, as time progresses, more customers have joined the queue, and fewer customers can potentially join. This {\it diminishing population} gives rise to a class of reflected stochastic processes that vanish over time, and hence do not have a stationary distribution. We establish that, by suitably rescaling space and time, the queue length process converges to a Brownian motion on a parabola, a stochastic-process limit that captures the effect of a diminishing population by a negative quadratic drift. The stochastic-process limit provides insight into the macroscopic behavior (for n large) of the transitory queueing process, and the different phenomena occurring at different space-time scales.",
author = "G. Bet and {Hofstad, van der}, R.W. and {Leeuwaarden, van}, J.S.H.",
year = "2014",
language = "English",
series = "arXiv",
publisher = "s.n.",

}

Heavy-traffic analysis through uniform acceleration of transitory queues with diminishing populations. / Bet, G.; Hofstad, van der, R.W.; Leeuwaarden, van, J.S.H.

s.n., 2014. 43 blz. (arXiv; Vol. 1412.5329 [math.PR]).

Onderzoeksoutput: Boek/rapportRapportAcademic

TY - BOOK

T1 - Heavy-traffic analysis through uniform acceleration of transitory queues with diminishing populations

AU - Bet, G.

AU - Hofstad, van der, R.W.

AU - Leeuwaarden, van, J.S.H.

PY - 2014

Y1 - 2014

N2 - We consider the ¿(i)/GI/1 queue, in which the arrival times of a fixed population of n customers are sampled independently from an identical distribution. This model recently emerged as the canonical model for so-called transitory queues that are non-stationary, time-varying and might operate only over finite time. The model assumes a finite population of customers entering the queue only once. This paper presents a method for analyzing heavy-traffic behavior by using uniform acceleration, which simultaneously lets the population n and the service rate grow large, while the initial resource utilization approaches one. A key feature of the model is that, as time progresses, more customers have joined the queue, and fewer customers can potentially join. This {\it diminishing population} gives rise to a class of reflected stochastic processes that vanish over time, and hence do not have a stationary distribution. We establish that, by suitably rescaling space and time, the queue length process converges to a Brownian motion on a parabola, a stochastic-process limit that captures the effect of a diminishing population by a negative quadratic drift. The stochastic-process limit provides insight into the macroscopic behavior (for n large) of the transitory queueing process, and the different phenomena occurring at different space-time scales.

AB - We consider the ¿(i)/GI/1 queue, in which the arrival times of a fixed population of n customers are sampled independently from an identical distribution. This model recently emerged as the canonical model for so-called transitory queues that are non-stationary, time-varying and might operate only over finite time. The model assumes a finite population of customers entering the queue only once. This paper presents a method for analyzing heavy-traffic behavior by using uniform acceleration, which simultaneously lets the population n and the service rate grow large, while the initial resource utilization approaches one. A key feature of the model is that, as time progresses, more customers have joined the queue, and fewer customers can potentially join. This {\it diminishing population} gives rise to a class of reflected stochastic processes that vanish over time, and hence do not have a stationary distribution. We establish that, by suitably rescaling space and time, the queue length process converges to a Brownian motion on a parabola, a stochastic-process limit that captures the effect of a diminishing population by a negative quadratic drift. The stochastic-process limit provides insight into the macroscopic behavior (for n large) of the transitory queueing process, and the different phenomena occurring at different space-time scales.

UR - http://arxiv.org/pdf/1412.5329v1

M3 - Report

T3 - arXiv

BT - Heavy-traffic analysis through uniform acceleration of transitory queues with diminishing populations

PB - s.n.

ER -