Abstract: |
The rapid diffusion of web services is changing the software engineering landscape. One of the most interesting features offered\nby service-oriented systems is the possibility to perform dynamic binding, i.e. choosing, among sets of semantically equivalent services, those which better contribute to meet some constraints (e.g., related to the cost or to any other Quality of Service\nattributes) and optimize some other criteria (e.g., the response time). Solving this problem is NP-hard, and approaches to tackle it\nusing Genetic Algorithms have been proposed.\n\nIn some cases, especially when it is not possible to find any solution to the aforementioned problem, it would be useful to relax\nconstraints, in order to find some alternative solutions that, while not meeting the initial constraints, at least offer a reasonable\nQuality of Service. This paper proposes the use of fuzzy logic to address the imprecision in specifying QoS constraints, estimating QoS\nvalues and expressing Service Level Agreements. |
