Adequacy-based Testing of Extra-Functional Properties of Embedded Systems (VR)

As software systems become increasingly complex, there is an imperious need to ensure their performance and correct resource management that includes memory usage, availability of bandwidth, CPU etc. Some of the resource-related bugs are unavailability of required resources, using the wrong resource, but also resource-caused performance bottlenecks.

Test data adequacy assessment aims to determine the extent to which a software under test has been sufficiently tested. Ideally, a highly adequate set of test data would imply that the software under test has been thoroughly tested and hence we can expect it to contain few remaining defects. In this project, the overall goal is to develop models and methods for adequacy-based testing of extra-functional properties in embedded systems.

Based on this research objective, the project is organised into four directions of research: (i) Formal Models for Performance and Resource-Usage Testing, focusing on using priced automata theory for establishing mappings from extra-functional requirements to test models with formal semantics, (ii) Model-Based Test Suite Generation, focusing on establishing methods for test suite generation, by means of model-checking of the above models, and derivation of test cases from the model checking traces, (iii) Compositional Testing, focusing on determining the extent to which test results, and extra-functional test adequacy in particular, of integrated systems can be inferred from the test adequacy of the individual components of the systems, and (iv) Algorithm Implementation and Experimentation, focusing on packaging project results in an accessible tool format, as well as on experimental evaluation of adequacy criteria precision, and costeffectiveness of the techniques proposed in the project.