Research On Service Composition In Service Oriented Architecture

SOA (Service Oriented Architecture) is an important method to achieve business agility. SOA possesses the advantages such as agility, loose coupling, and reusability. Through the collaboration and integration of services, different platforms can be organized to handle flexible business processes. Using SOA architecture can improve the agility of information systems, improve the corporate capacity for dealing with complex business environments and improve the competitiveness of enterprisesA basic method of organizing services in SOA is service composition. Building business processes through service composition can achieve the business agility and meet the complex business needs of enterprises.Service composition process is an effective way to get reusable and value-added services; it can achieve business needs through the integration of different functional services. So service composition is a hot topic in SOA.The paper starts from the needs of business-oriented service composition, analyzes the characteristics of service and the structure features of business process. By combining the QoS factors of business service, the paper then researches the issues of service-oriented business process decomposition, service selection and service composition. The paper's main work includes:The paper presents a QoS based service-oriented process decomposition method:This method first applies Ontology Web Language for Services (OWL-S) in business process modeling, then transforms complex business process into several typical service combination structures by modules aggregation and combination, and then obtains an optimal service-oriented decomposition with QoS constraints calculation. A vehicle loan grant process is analyzed as an example to verify the effectiveness of this method.The paper improves the selection method based on semantic service. The selection process starts from three areas including functionality, input and output and QoS constraints. The research focuses on the function similarity calculation and the influence of input/output dependencies in selection.Service composition based on genetic algorithm: composition process focused on solving two problems: service composition can be smoothly completed user needs; in the service composition process to consider the quality of service (QoS) constraints. Application of genetic algorithm, through the use of dynamic encoding and improved fitness function and adding process inspection, improve the efficiency of the algorithm.