TY - JOUR
T1 - Online IPA gradient estimators in stochastic continuous fluid models
AU - Wardi, Y.
AU - Melamed, B.
AU - Cassandras, C. G.
AU - Panayiòtou, C. G.
N1 - Funding Information:
1The first two authors were supported in part by the National Science Foundation under Grant DMI-0085659 and by DARPA under Contract F30602-00-2-0556. 2The last two authors were supported in part by the National Science Foundation under Grants EEC-95-27422 and ACI-98-73339, by AFOSR under Grants F49620-98-1-0387 and F49620-01-0056, by the Air Force Research Laboratory under Contract F30602-99-C-0057, by EPRI/ ARO under Contract WO8333-03, and by the NOKIA Research Center (Boston). 3Professor, School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, Georgia. 4Professor, Department of Management Science and Information Systems, Rutgers University, Piscataway, New Jersey. 5Professor, Department of Manufacturing Engineering, Boston University, Boston, Massachusetts. 6Graduate Research Assistant, Department of Manufacturing Engineering, Boston University, Boston, Massachusetts.
PY - 2002/11
Y1 - 2002/11
N2 - This paper applies infinitesimal perturbation analysis (IPA) to loss-related and workload-related metrics in a class of stochastic flow models (SFM). It derives closed-form formulas for the gradient estimators of these metrics with respect to various parameters of interest, such as buffer size, service rate, and inflow rate. The IPA estimators derived are simple and fast to compute, and are further shown to be unbiased and nonparametric, in the sense that they can be computed directly from the observed data without any knowledge of the underlying probability law. These properties hold out the promise of utilizing IPA gradient estimates as ingredients of online management and control of telecommunications networks. While this paper considers single-node SFMs, the analysis method developed is amenable to extensions to networks of SFM nodes with more general topologies.
AB - This paper applies infinitesimal perturbation analysis (IPA) to loss-related and workload-related metrics in a class of stochastic flow models (SFM). It derives closed-form formulas for the gradient estimators of these metrics with respect to various parameters of interest, such as buffer size, service rate, and inflow rate. The IPA estimators derived are simple and fast to compute, and are further shown to be unbiased and nonparametric, in the sense that they can be computed directly from the observed data without any knowledge of the underlying probability law. These properties hold out the promise of utilizing IPA gradient estimates as ingredients of online management and control of telecommunications networks. While this paper considers single-node SFMs, the analysis method developed is amenable to extensions to networks of SFM nodes with more general topologies.
KW - Stochastic fluid models
KW - infinitesimal perturbation analysis
KW - network management and control
UR - http://www.scopus.com/inward/record.url?scp=0036409469&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0036409469&partnerID=8YFLogxK
U2 - 10.1023/A:1020892306506
DO - 10.1023/A:1020892306506
M3 - Article
AN - SCOPUS:0036409469
VL - 115
SP - 369
EP - 405
JO - Journal of Optimization Theory and Applications
JF - Journal of Optimization Theory and Applications
SN - 0022-3239
IS - 2
ER -