In this paper we study the automation of test derivation and execution in the area of conformance testing. The test scenarios are derived from multiple specification languages: LOTOS, PROMELA and SDL. A central theme of this study is the usability of batch-oriented and on-the-fly testing approaches. To facilitate the derivation from multiple formal description techniques and the different test execution approaches, an open, generic environment called ToRX is introduced. ToRX enables plugging in existing or dedicated tools. We have carried out several experiments in testing a conference protocol, resulting in requirements on automated testing and benchmarking criteria.

Abstract. We introduce JTorX, a tool for model-driven test derivation and execution, based on the ioco theory. This theory, originally presented in [12], has been refined in [13] with test-cases that are input-enabled. For models with underspecified traces [3] introduced uioco. JTorX improves over its predecessor TorX [14] by using uioco and this newer ioco theory. By being much easier to deploy, due to improved installation, configuration and usage. And by integrating additional functionality, next to testing: checking for (u)ioco between models [6]; checking for underspecified traces in a model; interactive or guided simulation of a model. This makes JTorX an excellent vehicle for educational purposes in courses on model-based testing, as experience has shown -and its usefulness is not limited to education, as experience has shown too.

Abstract. Effective risk management is a key to ensure that our nuclear power plants, medical equipment, and power grids are dependable; and it is often required by law. Fault Tree Analysis (FTA) is a widely used methodology here, computing important dependability measures like system reliability. This paper presents DFTCalc, a powerful tool for FTA, providing (1) efficient fault tree modelling via compact representations; (2) effective analysis, allowing a wide range of dependability properties to be analysed (3) efficient analysis, via state-of-the-art stochastic techniques; and (4) a flexible and extensible framework, where gates can easily be changed or added. Technically, DFTCalc is realised via stochastic model checking, an innovative technique offering a wide plethora of powerful analysis techniques, including aggressive compression techniques to keep the underlying state space small.

No abstract

No abstract