No abstract

No abstract

We present two formalisations of the Business Process Modelling Notation (BPMN). In particular, we introduce a semantic model for BPMN in the process algebra CSP; we then study an augmentation of this model in which we introduce relative timing information, allowing one to specify timing constraints on concurrent activities. By exploiting CSP refinement, we are able to show some relationships between the timed and the untimed models. We then describe a novel empirical studies model, and the transformation to BPMN, allowing one to apply our formal semantics for analysing different kind of workflows. To provide a better facility for describing behaviour specification about a BPMN diagram, we also present a pattern-based approach using which a workflow designer could specify properties which could otherwise be difficult to express. Our approach is specifically designed to allow behavioural properties of BPMN diagrams to be mechanically verified via automatic model-checking as provided by the FDR tool. We use two examples to illustrate our approach.

Abstract. We present a framework for statically detecting deadlocks in a concurrent object language with asynchronous invocations and operations for getting values and releasing the control. Our approach is based on the integration of two static analysis techniques: (i) an inference algorithm to extract abstract descriptions of methods in the form of behavioral types, called contracts, and (ii) an evaluator that computes a fixpoint semantics returning a finite state model of contracts. A potential deadlock is detected when a circular dependency is found in some state of the model. We discuss the theory and the prototype implementation of our framework. Our tool is validated on an industrial case study based on the Fredhopper Access Server (FAS) developed by SDL Fredhoppper. In particular we verify one of the core concurrent components of FAS to be deadlock-free.

Abstract. We present case studies which show how the paradigm of learning-based testing (LBT) can be successfully applied to black-box requirements testing of industrial reactive systems. For this, we apply a new testing tool LBTest, which combines algorithms for incremental black-box learning of Kripke structures with model checking technology. We show how test requirements can be modeled in propositional linear temporal logic extended by finite data types. We then provide benchmark performance results for LBTest applied to three industrial case studies.