An Approach For QoS-aware Service Composition Based On Ant Colony Algoirthms

In recent years, more and more Web services are shared on the Internet. But most of the functions provided by these services are simple, which cannot meet all the requirements of user’s. Therefore studies have been carried out to select and assemble simple Web services based on the needs of users, which in turn generate more powerful Web service composition and achieve the purpose of Web service’s reuse and appreciation.This paper puts forward an approach for QoS-aware Web service composition based on an improved ant colony algorithm. The main contribution of this paper is puts forward an integration technology from existing Web services to the final Web service composition. To accomplish the goal, firstly, according to the input/output parameters description in WSDL, Web services were matched. Secondly, abstract service composition scheme were extracted from the matching degree. Then a best service was selected from a set of candidate services using the local optimization strategy of ant colony algorithm. Finally, the best Web service composition was worked out based on ant colony algorithm hybrid optimization strategy. Besides, a dynamic strategy was proposed in this paper to accommodate to the dynamic property of web service. The mathematical model of Web service composition in this paper is suitable for solving Web service composition problems. Besides it can be adapted to different demand of multiple service composition, this save time and avoid the waste of resources. The final experimental results shows that the improved algorithm has a satisfactory convergence and a higher success rate compared with the traditional algorithm, but the classification of the web services may take much time in the case of excessive web services.