A general simulation methodology for optical bar code evaluation

Abstract
A simulation methodology ofr estimating the probability of the occurrence of a given rare event in a complex probabilistic system is presented. The practical implementation of the critical value stratification variance reduction technique is given in detail to explore the problems that arise in the generation of sample values where one usually encounters dependent continuous distributions. The methodology is set in the context of evaluating optical bar code systems and the effectiveness of alternative decoding logic schemes. The objective in these applications is to estimate the probability of various types of decoding errors (1×10—4 or smaller). The methodology presented should prove a valuable guideline for developing specific procedures when one attempts to use Monte Carlo Simulation for estimating probabilities of rare events in complex systems.