Scalable Incremental Test-case Generation from Large Behavior Models
- Cite this paper as:
- Aichernig B.K., Ničković D., Tiran S. (2015) Scalable Incremental Test-case Generation from Large Behavior Models. In: Blanchette J., Kosmatov N. (eds) Tests and Proofs. TAP 2015. Lecture Notes in Computer Science, vol 9154. Springer, Cham
Model-based testing is a popular black-box testing technology that enables automation of test generation and execution, while achieving a given coverage. The application of this technology to large and complex systems is still a challenging problem, due to the state-space explosion resulting from the size of specification models.
In this paper, we evaluate a test-case generation approach that tackles this complexity along two axes. Firstly, our approach relies on a synchronous specification language for test models, thus avoiding the problem of interleaving actions. Secondly, our specification language enables incremental test-case generation by providing support for compositional modeling, in which each requirement or view of the system is expressed as a separate partial model. The individual requirement models are then naturally combined by conjunction, which is incrementally computed during the generation of tests.
We apply our test-case generation technique to two large industrial case studies: (1) an electronic control unit (ECU) of an agricultural device; and (2) a railway interlocking system. We demonstrate the scalability of our approach by creating a series of test models with increasing complexity and report on the experimental results.
Unable to display preview. Download preview PDF.