2002
DOI: 10.1002/ett.4460130412
|View full text |Cite
|
Sign up to set email alerts
|

On steady‐state simulation of some weak regenerative networks

Abstract: Abstract. We discuss the possibilities of the so-called weak regeneration in the steady-state analysis of multiserver queues and some Jackson-type and multiclass queueing networks when classical regenerative simulation is ineffective. A few constructions of weak regeneration for open Jackson-type and multiclass queueing networks are given in an explicit form. Conditions are found when efficiency of weak regeneration (expressed in the simulation time needed for confidence estimation with a given precision) stay… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
5
0

Year Published

2003
2003
2016
2016

Publication Types

Select...
3
2
1

Relationship

0
6

Authors

Journals

citations
Cited by 6 publications
(5 citation statements)
references
References 41 publications
0
5
0
Order By: Relevance
“…Of course, these results do not rule out the possibility of identifying nonrandomized regeneration times such as those generated by visits to an atom in a Markov chain. For nontrivial examples of such regeneration times in queueing systems, see [15,13,14,18].…”
Section: Discussionmentioning
confidence: 99%
See 1 more Smart Citation
“…Of course, these results do not rule out the possibility of identifying nonrandomized regeneration times such as those generated by visits to an atom in a Markov chain. For nontrivial examples of such regeneration times in queueing systems, see [15,13,14,18].…”
Section: Discussionmentioning
confidence: 99%
“…We would like to thank Evsey Morozov and Irina Aminova for sharing working versions of their paper [14] with us. The work of the first author was supported by National Science Foundation Grant DMI-0085165.…”
Section: Acknowledgmentsmentioning
confidence: 99%
“…All cases considered in the paper satisfy this condition because the number of dependent cycles is upper bounded by the finite constant D := R 0 R 1 · · · R M , and thus at most D successive cycles are dependent. (For the one-dependent cycles the confidence interval includes an estimate of the covariance between two adjacent cycles and is given in [14]. )…”
Section: Regenerative Simulationmentioning
confidence: 99%
“…Before presenting the simulation results, we mention that the workload process in a GI/G/1 system (with renewal input and the i.i.d service times) is still a Markov chain and satisfies Lindley's recursion (14). Of course in general the exact expression for a high load probability for such a system is unknown.…”
Section: Stationary Workload In M /G/1 Queuementioning
confidence: 99%
“…To use estimator (12), we need to construct regenerations {β n }. For a general acyclic network it can done as in [10]. Note, however, that it is a rather complicated construction.…”
Section: Quasi-regenerationmentioning
confidence: 99%