Verification of probabilistic systems - Methods and tools
Résumé
Historically, functional verification and performance evaluation have been two distinct stages in the development of applications. Each one had its own models and methods. For 15 years, numerous works covered both areas. These works are now referred to as probabilistic verification or, more accurately, by verification of probabilistic systems. This direction of research is prompted by the new needs of the modelers. They wish, for instance, to compute the probability that a property expressed as some logical formula is satisfied. They also wish to analyze a system including both non-deterministic and probabilistic features. The goal of this chapter is to introduce three important topics related to this research: - the definition of high level stochastic models; - the verification of Markov chains (MC); -the verification of Markov decision processes (MDP). We will always follow the same organization for these topics: - the detailed presentation of one approach, - an overview of the other approaches, - the description of an analysis tool related to this topic.