Fuzzy rule-based quantitative framework for system testability measurement
Date
2024
Authors
Lee, S.
Efatmaneshnik, M.
James, A.
Mayer, W.
Smith, J.
Grabert, T.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
2024 IEEE International Symposium on Systems Engineering (ISSE), 2024, pp.1-4
Statement of Responsibility
Conference Name
10th IEEE International Symposium on Systems Engineering, ISSE 2024 (16 Oct 2024 - 19 Oct 2024 : Perugia, Italy)
Abstract
System testability refers to the degree to which a system supports and facilitates testing to determine whether the test criteria have been met. Ensuring that a system is testable within the constraints of a project is crucial, as it directly impacts the ability to identify defects and verify functionality. A significant challenge in estimating system testability is the dependency on numerous variables, many of which cannot be adequately captured using precise, crisp values. For example, the expertise of the tester is inherently a fuzzy variable that influences the testability of a system. In this study, we present a fuzzy rule-based quantitative Framework for measuring system testability. The objective here is to establish a foundation for developing measures and guidelines that enhance system testability to then facilitate optimal test planning. We analyze multiple input parameters from diverse perspectives affecting system testability using Fuzzy Logic (FL).
This approach helps to identify optimal strategies for enhancing system testability by considering the inherent uncertainty and variability in these factors. Determining the appropriate level of system testability involves mapping these fuzzy input parameters to corresponding outputs, providing a more nuanced and accurate assessment. Twenty-one factors affecting the testability of complex systems were identified through a Delphi method. These factors were then clustered into six main classes of factors. Importance weights were also assigned to each factor and class of factors through the same Delphi method. A Two-layered fuzzy estimation framework not only aids in better test planning but also contributes to the overall reliability and robustness of the system being tested. The approach can also assist with optimal test resource allocation to maximize testing success and system testability.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
Copyright 2024 IEEE