A Compact Semantic Model for Characterization of Stochastic Temporal Properties of Concurrent Systems

In this paper, a new semantic model is proposed for characterizing the performance properties of stochastic concurrent systems, called Maximality-based Labeled Stochastic Transition System (MLSTS). A Stochastic Process Algebra, called S-LOTOS, is associated with these MLSTS models as a specification language to describe the stochastic temporal aspects of concurrent systems under the assumption of generally distributed durations of actions. The MLSTS models can be automatically generated from S-LOTOS specifications according to the (true concurrency) maximality semantics. In addition, we bring out the main advantage of the MLSTS as it is shown on practical examples in reducing the number of states and transitions w.r.t. standard ST-semantic models, which are frequently used in specification and modeling of stochastic systems with non-Markovian Process Algebra. Using our MoVeS Tool, we bring out the main advantage of our approach which is a drastic reduction obtained in the number of states and transitions, in comparison to standard ST-semantic models.