Thesis
Component-based software engineering: Qualification of components during design
Washington State University
Master of Science (MS), Washington State University
2004
Handle:
https://hdl.handle.net/2376/235
Abstract
In the software industry component-based software development is becoming an increasingly popular and sought-after approach. Integrating components into a software system requires some component-based software engineering practices. Gap fulfillment is a commonly used approach to finding the right component for a software system. This approach has two possible outcomes: usable components are found or no relevant component is found. Component-based software engineering practices require interaction standards, composition standards, and component models. This thesis will focus on a systematic process for selecting and qualifying components during the design phase. Components will be selected and qualified based on interaction and composition information that is found in both static and dynamic UML diagrams. The approach discovers architectural mismatches based on coverage, fit, and coupling measures found in an integrated UML model. The integrated UML model combines Class and Sequence Diagrams into Object Method Directed Acyclic Graphs (OMDAG). The coverage and fit qualification measures rely on extraction of information from Constrained Class Tuples (CCTS), which are derived from Class Diagrams in the Integrated UML Model. The Class and Sequence Diagrams are combined into the Constrained Object Method Directed Acyclic Graph (COMDAG) so that coupling measures can be calculated. Coupling metrics are derived by executing Use Cases, from operational profiles, through the COMDAG. In our approach our selection of candidate components is based solely on available black-box information found in requirements and UML documents. An initial Off-The-Shelf-Option (OTSO) is performed to determine candidate components. For each of the candidate components, we use Set Theory to qualify them against our designed components. The combination of the coverage, fit, and coupling measures provide the information for a decision about a candidate component’s fitness for a particular software design.
Metrics
10 File views/ downloads
50 Record Views
Details
- Title
- Component-based software engineering
- Creators
- Andrew Steven O'Fallon
- Contributors
- Anneliese Andrews (Degree Supervisor)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; Pullman, Wash. :
- Identifiers
- 99900525285501842
- Language
- English
- Resource Type
- Thesis