Wanmo Kang
Abstract
We propose nonstandard simulation estimators of expected time averages over finite intervals [0, t], seeking to enhance estimation efficiency. We make three key assumptions: (i) the underlying stochastic process has regenerative structure, (ii) the time average approaches a known limit as time t increases and (iii) time 0 is a regeneration time. To exploit those properties, we propose a residual-cycle estimator, based on data from the regenerative cycle in progress at time t, using only the data after time t. We prove that the residual-cycle estimator is unbiased and more efficient than the standard estimator for all sufficiently large t. Since the relative efficiency increases in t, the method is ideally suited to use when applying simulation to study the rate of convergence to the known limit. We also consider two other simulation techniques to be used with the residual-cycle estimator. The first involves overlapping cycles, paralleling the technique of overlapping batch means in steady-state estimation; multiple observations are taken from each replication, starting a new observation each time the initial regenerative state is revisited. The other technique is splitting, which involves independent replications of the terminal period after time t, for each simulation up to time t. We demonstrate that these alternative estimators provide efficiency improvement by conducting simulations of queueing models.
Supplemental Material
Available for Download
Online appendix to designing mediation for context-aware applications. The appendix supports the information on article 8.
- Asmussen, S. 2003. Applied Probability and Queues, second ed. Wiley, New York.Google Scholar
- Billingsley, P. 1999. Convergence of Probability Measures, second ed. Wiley, New York.Google Scholar
- Bratley, P., Fox, B. L., and Schrage, L. E. 1987. A Guide to Simulation, Second Ed. Springer Verlag, New York. Google ScholarDigital Library
- Calvin, J. M. and Nakayama, M. K. 1998. Using permutations in regenerative simulations to reduce variance. ACM Trans. Model. Comput. Simul. 8, 153--193. Google ScholarDigital Library
- Calvin, J. M. and Nakayama, M. K. 2000. Central limit theorems for permuted regenerative estimators. Oper. Res. 48, 776--787. Google ScholarDigital Library
- Chow, Y. S., Hsiung, C. A., and Lai, T. L. 1979. Extended renewal theory and moment convergence in Anscombe's theorem. Ann. Probab. 7, 304--318.Google ScholarCross Ref
- Chung, K. L. 1974. A Course in Probability Theory, second ed. Academic Press.Google Scholar
- Crane, M. A. and Iglehart, D. L. 1975. Simulating stable stochastic systems III: regenerative processes and discrete event simulation. Oper. Res. 23, 33--45.Google ScholarDigital Library
- Duffield, N. G. and Whitt, W. 2000. Network design and control using on-off and multi-level source traffic models with heavy-tailed distributions. Chapter 17 in Self-Similar Network Traffic and Performance Evaluation, Kihong Park and Walter Willinger, eds. John Wiley and Sons, 421-445.Google Scholar
- Erdélyi, A. 1956. Asymptotic Expansions, Dover.Google Scholar
- Glynn, P. W. and Iglehart, D. L. 1986. Consequences of uniform integrability for simulation. Tech. rep. 40, Stanford University, Dept. of Operations Research.Google Scholar
- Glynn, P. W. and Whitt, W. 1992. The Asymptotic Efficiency of Simulation Estimators. Oper. Res., 40, 505--520. Google ScholarDigital Library
- Glynn, P. W. and Whitt, W. 1993. Limit theorems for cumulative processes, Stoch. Proc. Appl. 47, 299--314.Google ScholarCross Ref
- Glynn, P. W. and Whitt, W. 2002. Necessary conditions in limit theorems for cumulative processes, Stoch. Proc. Appl. 98, 199--209.Google ScholarCross Ref
- Glynn, P. W. and Wong E. W. 1996. Efficient simulation via coupling. Prob. Engr. Inf. Sci. 10, 165--186.Google ScholarCross Ref
- Gut, A. 1988. Stopped Random Walks, Springer.Google Scholar
- Hammersley, J. M. and Handscomb, D. C. 1964. Monte Carlo Methods, Methuen, London.Google Scholar
- Henderson, S. G. and Glynn, P. W. 2002. Approximating martingales for variance reduction in Markov process simulation. Math. Opes. Res. 27, 253--271. Google ScholarDigital Library
- Kang, W., Shahabuddin, P., and Whitt, W. 2007. Exploiting regenerative structure to estimate finite time averages via simulation: supplementary material available in the ACM Digital Library.Google Scholar
- Keilson, J. 1979. Markov Chain Models---Rarity and Exponentiality, Springer.Google Scholar
- Meketon, M. S. and Heidelberger, P. 1982. A renewal theoretic approach to bias reduction in regenerative simulations. Manag. Sci. 28, 173--181.Google ScholarDigital Library
- Meketon, M. S. and Schmeiser, B. 1984. Overlapping batch means; something for nothing? Proceedings of 1984 Winter Simulation Conference, 227--230. Google ScholarDigital Library
- Meyn, S. P. and Tweedie, R. L. 1993. Markov Chains and Stochastic Stability, Springer.Google Scholar
- Pawlikowski, K. 1990. Steady state simulation of queueing processes; a survey of problems and solutions. ACM Comput. Surv. 22, 123--170. Google ScholarDigital Library
- Ross, S. M. 1996. Stochastic Processes, second ed., John Wiley and Sons.Google Scholar
- Ross, S. M. 2003. Introduction to Probability Models, eighth ed. Academic Press, New York. Google ScholarDigital Library
- Smith, W. L. 1955. Regenerative stochastic processes. Proceedings of the Royal Society of London A232, 6--31.Google Scholar
- Whitt, W. 1972. Embedded renewal processes in the GI/G/s queue. J. Appl. Prob. 9, 650--658.Google ScholarCross Ref
- Whitt, W. 1992. Asymptotic formulas for Markov processes. Oper. Res. 40, 279--291. Google ScholarDigital Library
- Wong E. W., Glynn, P. W., and Iglehart, D. L. 1999. Transient simulation via empirically based coupling. Prob. Engr. Inf. Sci. 13, 147--167. Google ScholarDigital Library
Index Terms
- Exploiting regenerative structure to estimate finite time averages via simulation
Recommendations
The semi-regenerative method of simulation output analysis
We develop a class of techniques for analyzing the output of simulations of a semi-regenerative process. Called the semi-regenerative method, the approach is a generalization of the regenerative method, and it can increase efficiency. We consider the ...
Using permutations in regenerative simulations to reduce variance
Special issue on modeling and analysis of stochastic systemsWe propose a new estimator for a large class of performance measures obtained from a regenerative simulation of a system having two distinct sequences of regeneration times. To construct our new estimator, we first generate a sample path of a fixed ...
Estimation of steady-state central moments by the regenerative method of simulation
Let X be a possible recurrent regenerative process on state space S with steady-state distribution @p. Given a function @?: S -> R, we consider the problem of estimating the steady-state central moments @m"k(@?) = @?"s(@?(x)-r)^k@p(dx) where r is the ...
Comments