Variance-based sampling for simulating cycle time - Throughput curves using simulation-based estimates

John Fowler, S. E. Leach, G. T. Mackulak, B. L. Nelson

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Simulation at several different traffic intensities is required when generating a cycle time-throughput (CT-TH) curve. Previous work has shown that a variance-based allocation using an asymptotic variance approximation results in more precise confidence intervals than those achieved through naïve sampling in which an equal amount of effort is allocated to each traffic intensity being investigated. Many systems for which these CT-TH curves are desired are too complex for an asymptotic variance approximation to be easily determined. This paper presents a fixed-budget variance-based sampling allocation procedure for the simulation of CT-TH curves using variance estimates of the sample mean calculated from pilot simulation runs. The proposed allocation procedure significantly improved the range of precision over the naïve allocation.

Original languageEnglish (US)
Pages (from-to)69-80
Number of pages12
JournalJournal of Simulation
Volume2
Issue number2
DOIs
StatePublished - Jul 2008

Fingerprint

Throughput
Sampling
Curve
Asymptotic Variance
Estimate
Simulation
Traffic
Sample mean
Approximation
Confidence interval
Range of data

ASJC Scopus subject areas

  • Software
  • Modeling and Simulation

Cite this

Variance-based sampling for simulating cycle time - Throughput curves using simulation-based estimates. / Fowler, John; Leach, S. E.; Mackulak, G. T.; Nelson, B. L.

In: Journal of Simulation, Vol. 2, No. 2, 07.2008, p. 69-80.

Research output: Contribution to journalArticle

Fowler, John ; Leach, S. E. ; Mackulak, G. T. ; Nelson, B. L. / Variance-based sampling for simulating cycle time - Throughput curves using simulation-based estimates. In: Journal of Simulation. 2008 ; Vol. 2, No. 2. pp. 69-80.
@article{bc95be0954924d5cb361fe15f45fbff3,
title = "Variance-based sampling for simulating cycle time - Throughput curves using simulation-based estimates",
abstract = "Simulation at several different traffic intensities is required when generating a cycle time-throughput (CT-TH) curve. Previous work has shown that a variance-based allocation using an asymptotic variance approximation results in more precise confidence intervals than those achieved through na{\"i}ve sampling in which an equal amount of effort is allocated to each traffic intensity being investigated. Many systems for which these CT-TH curves are desired are too complex for an asymptotic variance approximation to be easily determined. This paper presents a fixed-budget variance-based sampling allocation procedure for the simulation of CT-TH curves using variance estimates of the sample mean calculated from pilot simulation runs. The proposed allocation procedure significantly improved the range of precision over the na{\"i}ve allocation.",
author = "John Fowler and Leach, {S. E.} and Mackulak, {G. T.} and Nelson, {B. L.}",
year = "2008",
month = "7",
doi = "10.1057/palgrave.jos.4250033",
language = "English (US)",
volume = "2",
pages = "69--80",
journal = "Journal of Simulation",
issn = "1747-7778",
publisher = "Palgrave Macmillan Ltd.",
number = "2",

}

TY - JOUR

T1 - Variance-based sampling for simulating cycle time - Throughput curves using simulation-based estimates

AU - Fowler, John

AU - Leach, S. E.

AU - Mackulak, G. T.

AU - Nelson, B. L.

PY - 2008/7

Y1 - 2008/7

N2 - Simulation at several different traffic intensities is required when generating a cycle time-throughput (CT-TH) curve. Previous work has shown that a variance-based allocation using an asymptotic variance approximation results in more precise confidence intervals than those achieved through naïve sampling in which an equal amount of effort is allocated to each traffic intensity being investigated. Many systems for which these CT-TH curves are desired are too complex for an asymptotic variance approximation to be easily determined. This paper presents a fixed-budget variance-based sampling allocation procedure for the simulation of CT-TH curves using variance estimates of the sample mean calculated from pilot simulation runs. The proposed allocation procedure significantly improved the range of precision over the naïve allocation.

AB - Simulation at several different traffic intensities is required when generating a cycle time-throughput (CT-TH) curve. Previous work has shown that a variance-based allocation using an asymptotic variance approximation results in more precise confidence intervals than those achieved through naïve sampling in which an equal amount of effort is allocated to each traffic intensity being investigated. Many systems for which these CT-TH curves are desired are too complex for an asymptotic variance approximation to be easily determined. This paper presents a fixed-budget variance-based sampling allocation procedure for the simulation of CT-TH curves using variance estimates of the sample mean calculated from pilot simulation runs. The proposed allocation procedure significantly improved the range of precision over the naïve allocation.

UR - http://www.scopus.com/inward/record.url?scp=44649124885&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=44649124885&partnerID=8YFLogxK

U2 - 10.1057/palgrave.jos.4250033

DO - 10.1057/palgrave.jos.4250033

M3 - Article

AN - SCOPUS:44649124885

VL - 2

SP - 69

EP - 80

JO - Journal of Simulation

JF - Journal of Simulation

SN - 1747-7778

IS - 2

ER -