The limiting value of derivative estimators based on perturbation analysis

Infinitesimal perturbation analysis is a method of obtaining estimates of performance sensitivity through simulation of a stochastic system. Expressions are derived for the limiting value of a broad class of such estimators associated with queueing networks, in terms of the unique solution to a set of linear equations. The approach used is to augment the underlying queueing process with information about which servers have been “perturbed” and by how much. The augmented process can then be studied as an absorbing Markov chain in which the absorbing sets consist of states in which all servers are equally perturbed. An application of these results is a simple proof of the consistency of an estimator previously investigated experimentally.

[1]  Kai Lai Chung,et al.  Markov Chains with Stationary Transition Probabilities , 1961 .

[2]  J. Kemeny,et al.  Denumerable Markov chains , 1969 .

[3]  Michael A. Crane,et al.  Simulating Stable Stochastic Systems, II: Markov Chains , 1974, JACM.

[4]  D. Iglehart,et al.  Discrete time methods for simulating continuous time Markov chains , 1976, Advances in Applied Probability.

[5]  Xi-Ren Cao,et al.  Perturbation analysis and optimization of queueing networks , 1983 .

[6]  Y C Ho,et al.  Equivalent network, load dependent servers, and perturbation analysis—an experimental study , 1986 .

[7]  P. Glynn,et al.  Discrete-time conversion for simulating semi-Markov processes , 1986 .

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

[9]  Yu-Chi Ho,et al.  Performance evaluation and perturbation analysis of discrete event dynamic systems , 1987 .

[10]  Rajan Suri,et al.  Infinitesimal perturbation analysis for general discrete event systems , 1987, JACM.

[11]  P. Glasserman.  Infinitesimal perturbation analysis of a birth and death process , 1988 .

[12]  Xi-Ren Cao,et al.  Convergence properties of infinitesimal perturbation analysis , 1988 .

[13]  Xi-Ren Cao,et al.  A Sample Performance Function of Closed Jackson Queueing Networks , 1988, Oper. Res..

[14]  R. Suri,et al.  Perturbation analysis gives strongly consistent sensitivity estimates for the M/G/ 1 queue , 1988 .

[15]  X.-R. Cao,et al.  Realization factors and perturbation analysis of open queueing networks , 1989, Proceedings of the 28th IEEE Conference on Decision and Control,.

[16]  R. Suri,et al.  Perturbation analysis: the state of the art and research issues explained via the GI/G/1 queue , 1989, Proc. IEEE.

[17]  P. Glynn,et al.  Discrete-time conversion for simulating finite-horizon Markov processes , 1990 .

[18]  Xi-Ren Cao.  Realization factors and sensitivity analysis of queueing networks with state-dependent service rates , 1990, Advances in Applied Probability.

[19]  Paul Glasserman,et al.  Structural Conditions for Perturbation Analysis Derivative Estimation: Finite-Time Performance Indices , 1991, Oper. Res..