4.6 Article

A three-phase approach to improve the functionality of t-way strategy

Journal

SOFT COMPUTING
Volume -, Issue -, Pages -

Publisher

SPRINGER
DOI: 10.1007/s00500-023-08199-5

Keywords

Combinatorial testing; T-way strategy; Model checking; Evolution strategy; Interaction strength

Ask authors/readers for more resources

Although the t-way strategy aims to generate a minimal test suite for error detection in software systems, it faces challenges regarding the quality of the generated test suite, the difficulty of manually preparing parameters and their values, and the slow generation speed and large size of the test suite. This paper proposes a three-phase approach called TPA to address these challenges, including injecting information about special errors into the test suite, automatically preparing parameters and their values, and improving the functionality of the t-way strategy using an adopted version of evolution strategy. Comparative evaluations confirm that TPA outperforms other evolutionary algorithms in terms of functionality.
Although t-way strategy tries to generate a minimum test suite (TS) for detecting errors in software systems, its functionality is affected by three important challenges. The first one, which relates to the quality of the generated TS, expresses that some complex errors (e.g., deadlocks in concurrent systems) may not be detected through the generated TS. The second one is that manually preparing parameters and their values in the modern software systems is difficult or even impossible, whereas the third one is the low generation speed and the large size of the generated test suite. In this paper, we propose a three-phase approach (so-called TPA) to handle these challenges. It seems that injecting some information about special errors into the test suite can raise its quality. For this purpose, TPA, in the first phase, uses an optimized version of model checking to extract such information from a model of the system under test. The extracted information is then injected into the test suite. In the second phase, TPA uses the generated state space in the first phase to automatically prepare parameters and their values. In the last phase, TPA applies an adopted version of evolution strategy to improve the functionality of t-way strategy in terms of generation speed and test suite size. Multiple and pairwise comparisons of results confirm that TPA has the best functionality in comparison with other evolutionary algorithms.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.6
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available