Towards a proof method for Paradigm

The paper describes two perspectives on a verification approach for Paradigm, a coordination modeling language specifying an architecture in terms of components and their collaborations. One perspective concentrates on a single collaboration: per collaboration, properties can be derived through a small set of proof rules. The other perspective concentrates on dynamic dependencies between collaborations: guided by the architecture and driven by shared components behavioral properties of the complete model can be established. Two Paradigm models, a parallel assignment and a linear pipeline of workers and buffers, illustrate the approach.

• View at Publisher