Approximate model checking of stochastic COWS

Abstract

Given the description of a model and a probabilistic formula, approximate model checking is a verification technique based on statistical reasoning that allows answering whether or not the model satisfies the formula. Only a subset of the properties that can be analyzed by exact model checking can be attacked by approximate methods. These latest methods, though, being based on simulation and sampling have the advantage of not requiring the generation of the complete state-space of the model. Here we describe an efficient tool for the approximate model checking of services written in a stochastic variant of COWS, a process calculus for the orchestration of services. © 2010 Springer-Verlag Berlin Heidelberg.