Fast Simulation of Markov Fluid Models

In this paper we study continuous ow nite buuer systems with input rates modulated by Markov chains. Discrete event simulations are applied for estimating loss probabilities. The simulations are executed under a twisted version of the original probability measure (importance sampling). We present a simple rule for determining a new measure, then show that the new measure matches thèmost likely' empirical measure that we expect from large deviations arguments, and nally prove optimality of the new measure.