Home // International Journal On Advances in Systems and Measurements, volume 6, numbers 3 and 4, 2013 // View article

Static Preprocessing for Automated Structural Testing of Simulink Models

Authors:
Benjamin Wilmes

Keywords: Search-Based Testing, Static Preprocessing, Automotive Industry, Simulink

Abstract:
A feasible automation of testing software models would be of great benefit to industry, given the advantages of early testing as part of an efficient quality assurance process. Despite search-based testing having been applied with promising results to automate structural test data generation for Simulink models, the approach lacks efficiency. This paper features three static-analysis-based preprocessing techniques which are carried out prior to an automated test data search, to mitigate this efficiency problem. The first technique identifies unsatisfiable coverage goals by analyzing the ranges of model internal signals and excludes them from the search. The second preprocessing technique aims at reducing the search space by analyzing which model inputs actually require stimulation in order to reach a certain model state. A third technique sequences the coverage-goal-related search processes in order to maximize collateral coverage and reduce the size of the generated test suite. These additional techniques are able to make the search-based approach considerably more efficient, as results of a case study with our search-based testing tool TASMO, applied to industrial Simulink models, reveal.

Pages: 310 to 323

Copyright: Copyright (c) to authors, 2013. Used with permission.

Publication date: December 31, 2013

Published in: journal

ISSN: 1942-261x