Structural infinitesimal perturbation analysis (SIPA) for derivative estimation of discrete-event dynamic systems

In this paper we present an algorithm for derivative estimation of discrete-event dynamic systems (DED's). This algorithm is applicable to derivative estimation with respect to most structural parameters in DED's. It extends the applicability of existing perturbation analysis techniques (including infinitesimal perturbation analysis) on structural parameters. Furthermore, it offers an alternative approach to derivative estimation in DED's which sheds additional light on discrete-event simulation. Finally, unifying relationships with existing techniques are discussed. >

[1]  Yu-Chi Ho,et al.  A gradient technique for general buffer storage design in a production line , 1979 .

[2]  Christos G. Cassandras,et al.  Infinitesimal and finite perturbation analysis for queueing networks , 1982, 1982 21st IEEE Conference on Decision and Control.

[3]  Xi-Ren Cao,et al.  The phantom customer and marked customer methods for optimization of closed queueing networks with blocking and general service times , 1983, SIGMETRICS '83.

[4]  X. Cao,et al.  Realization probability in closed Jackson queueing networks and its application , 1987, Advances in Applied Probability.

[5]  Y. Ho,et al.  Smoothed (conditional) perturbation analysis of discrete event dynamical systems , 1987 .

[6]  Wei-Bo Gong.  Smoothed perturbation analysis algorithm for a G/G/1 routing problem , 1988, WSC '88.

[7]  Y. Ho,et al.  Extensions of infinitesimal perturbation analysis , 1988 .

[8]  C. Cassandras,et al.  On-line sensitivity analysis of Markov chains , 1989 .

[9]  Reuven Y. Rubinstein,et al.  Sensitivity Analysis and Performance Extrapolation for Computer Simulation Models , 1989, Oper. Res..

[10]  Alan Weiss,et al.  Sensitivity Analysis for Simulations via Likelihood Ratios , 1989, Oper. Res..

[11]  P. Vakili.  Using Uniformization for Derivative Estimation in Simulation , 1990 .

[12]  Y. Ho,et al.  An infinitesimal perturbation analysis algorithm for a multiclass G/G/1 queue , 1990 .

[13]  Weibo Gong,et al.  A new class of gradient estimators for queueing systems with real-time constraints , 1990, 29th IEEE Conference on Decision and Control.

[14]  Christos G. Cassandras,et al.  Robustness properties of perturbation analysis estimators for discrete event systems with unknown distributions , 1990, 29th IEEE Conference on Decision and Control.

[15]  P. L’Ecuyer,et al.  A Unified View of the IPA, SF, and LR Gradient Estimation Techniques , 1990 .

[16]  Peter W. Glynn,et al.  Likelihood ratio gradient estimation for stochastic systems , 1990, CACM.

[17]  Paul Glasserman,et al.  Gradient Estimation Via Perturbation Analysis , 1990 .

[18]  Xi-Ren Cao,et al.  Perturbation analysis of discrete event dynamic systems , 1991 .

[19]  M. Fu,et al.  Consistency of infinitesimal perturbation analysis for the GI/G/m queue , 1991 .

[20]  David D. Yao,et al.  Algebraic structure of some stochastic discrete event systems, with applications , 1991, Discret. Event Dyn. Syst..

[21]  Pierre L'Ecuyer,et al.  Comparing alternative methods for derivative estimation when IPA does not apply directly , 1991, 1991 Winter Simulation Conference Proceedings..

[22]  Yu-Chi Ho,et al.  Perturbation analysis: concepts and algorithms , 1992, WSC '92.

[23]  Pierre Brémaud,et al.  On the pathwise computation of derivatives with respect to the rate of a point process: The phantom RPA method , 1992, Queueing Syst. Theory Appl..

[24]  Pierre Brémaud,et al.  Maximal coupling and rare perturbation sensitivity analysis , 1992, Queueing Syst. Theory Appl..

[25]  Leyuan Shi.  Variance property of discontinuous perturbation analysis , 1996, Winter Simulation Conference.