Research On QoS Based Composite Service Description And Validation

With the widespread of Web Services (WS), more deployed services are being registeredinto a services registry, consequently giving rise to a series of problems. A conspicuous one isthe large amount of functionally equivalent Web Services returned by a service registry. Aplausible solution is to involve QoS (Quality of Service) in the services discovery todifferentiate the functionally equivalent WS. However, it is not an easy task to distinguish allthe possible QoS metric of WS. W3C published a proposal called:" QoS for Web Services:Requirements and Possible Approaches " which aims to identify all the possible QoSrequirements such as response time and throughout for Web services. The first task of thepaper is to define a extendable Web service quality description model. The model is based onthe W3C proposal using a formalized method and it’s an advanced and general qualitydescription model. The following quality constraint model calculation is based on this. Thismodel is focus on two important issues: firstly it involving semantic technology to improvethe readability of the metric. Secondly, the model is flexible and expansible.QoS offering is a set of constraints on QoS metrics. For example, the response timeshould be less than one second. We hope we could choose the services from both function andperformance aspect. In order to express the QoS offering in a formal way, some models andlanguages like WSLA are developed. The paper presents a service quality constraint modelcould solve this problem. The model will reference the semantic based metric model when weformalized construct the constraints. A composite service is a set of services organizedaccording to certain rules to complete a specific task or business process together. It organizesthe Web resources to provide powerful and customized services. We use a compositesservices algorithm in the paper to calculate the composite services quality metric value.Before the services registry recommend composite services to consumers according to thefunctional requirement, we judge whether these services are performance qualified.In conclusion, this paper presents a novel and extensible ontology-based approach fordescription of the QoS constraints. Further, we propose a QoS calculation method for thecomposite service to judge whether the set of recommended services is qualified for the QoSconstraints.