Variance Reductions Applied to Product - Form Multi - Class Queuing Networks : Antithetic Variates and Low Discrepancy Sequences

نویسنده

  • BRUNO TUFFIN
چکیده

Performances of product-form multi-class queuing networks can be determined from normalization constants. For large models, the evaluation of these performance metrics is not possible because of the involved amount of computer resources (either by using normaliza-tion constants or by using MVA approaches). Such large models can be evaluated with Monte Carlo methods. For a particular type of network, we propose a variance reduction technique based on antithetic variates. It leads to an improvement of Ross et al's algorithm which is designed to analyze the same family of models. For more general models, we use a mixture of Monte Carlo and quasi-Monte Carlo methods to improve the estimation with respect to Monte Carlo alone. (URA 227) Université de Rennes 1 – Insa de Rennes et en Automatique – unité de recherche de Rennes R eductions de la variance appliqu ees aux r eseaux de les d'attente multi-classes a forme produit : variables antagonistes et suites a discr epance faible R esum e : Les performances des r eseaux de les d'attente multi-classes a forme produit peuvent ^ etre d etermin ees a l'aide de constantes de normalisation. Pour de grands mod eles, l' evaluation de ces m etriques de performance est impossible car la quantit e de ressource infor-matique requise est trop importante (soit en utilisant des constantes de normalisation, ou en utilisant des approches de type MVA). De tels mod eles peuvent ^ etre evalu es par des m ethodes de Monte Carlo. Pour un type particulier de r eseau, nous proposons une technique de r eduction de la variance bas ee sur les variables antagonistes. Cela permet une am eliration de l'algorithme de Ross et al destin e a la m^ eme famille de mod eles. Pour des mod eles plus g en eraux, nous utilisons une combinaison des m ethodes de Monte Carlo et quasi-Monte carlo qui am eliore l'estimation par rapport a Monte Carlo seul. Mots-cl e : R eseaux a forme produit, Monte Carlo, r eduction de la variance, variables anta-gonistes, suites a discr epance faible, quasi-Monte Carlo

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Variance Reduction Technique for a Cellular System with Dynamic Resource Sharing

A new algorithm is given to improve the simulation of a cellular system with dynamic resource sharing. This technique is based on low discrepancy sequences, used in quasi-Monte Carlo methods and which, combined with Monte Carlo techniques using importance sampling and control variates (already developed for such a system), gives the best known results. INTRODUCTION We consider a cellular system...

متن کامل

Integrated Variance Reduction Strategies for Simulation

We develop strategies for integrated use of certain well-known variance reduction techniques to estimate a mean response in a finite-horizon simulation experiment. The building blocks for these integrated variance reduction strategies are the techniques of conditional expectation, correlation induction (including antithetic variates and Latin hypercube sampling), and control variates; and all p...

متن کامل

Generating Antithetic Random Variates in Simulation of a Replacement Process by Rejection Method

When the times between renewals in a renewal process are not exponentially distributed, simulation can become a viable method of analysis. The renewal function is estimated through simulation for a renewal process simulation for a renewal process with gamma distributed renewal times and the shape parameter a > 1. Gamma random deviates will be generated by means of the so called Acceptance Rejec...

متن کامل

Multi-process Parallel Antithetic Coupling For Backward and Forward Markov Chain Monte Carlo

Antithetic coupling is a general stratification strategy for reducing Monte Carlo variance without increasing the simulation size. The use of the antithetic principle in the Monte Carlo literature typically employs two strata via antithetic quantile coupling. We demonstrate here that further stratification, obtained by using k > 2 (e.g., k = 3 − 10) antithetically coupled variates, can offer su...

متن کامل

Investigating the effectiveness of Variance Reduction Techniques in Manufacturing, Call Center and Cross-docking Discrete Event Simulation Models

Variance reduction techniques have been shown by others in the past to be a useful tool to reduce variance in Simulation studies. However, their application and success in the past has been mainly domain specific, with relatively little guidelines as to their general applicability, in particular for novices in this area. To facilitate their use, this study aims to investigate the robustness of ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1996