Research On Collaborative Service Composition Algorithm Based On Tuple Space

With the rapid expansion of the Internet service market,it is more and more important to study service composition.The traditional service composition algorithm has centralized engine,to understand all the information about each service,so that the service composition algorithm could plan the service portfolio from a global perspective,but in the real world there are a variety of services from different areas,it is difficult to have centralized engine to understand all the service in all areas.The centralized engine of the traditional service composition algorithm could combine a number of services from known service information to deliver to individual users,but individual users are difficult to obtain the dynamic context of external services accurately and in real time,which leads to the problem of inaccurate planning.In view of the problem,the paper uses the method of removing centralized engines of traditional service composition algorithm to provide service Agent for the external service providers.Then,external service providers push the dynamic context in real time to the corresponding service Agent.Third,perform multi-agent autonomous negotiation in person service ecological environment(PSE2),use tuple space as a place of negotiation(Blackboard Structure).Finally,support multiple external service Agent concurrent distributed collaboration,plan a set of candidate service combinations and evaluate their selections according to individual user preferences to propose a collaborative service composition algorithm based on tuple space.A model of service cooperative relationship autonomous planning was proposed,which provides model support for service composition algo rithm,and divides the model into two factors,service multi-agent mechanism and support the personal tuple space of service autonomous cooperative.And the structure design was proposed,multi-external service provider corresponding to the multi-external service agent in the personal tuple space collaboration process was given.The strategy of service agent decision-making to participate in the collaboration was proposed in this paper,which provides the theoretical support for the service composition algo rithm and gives the policy rules of whether each external service agent decision-making participates in cooperation and how to participate in the collaborative planning model.The decision process is divided into three steps to determine the integrity of t he service portfolio,the ability to participate in collaboration,the willingness to participate in collaboration,and the algorithm design was given.A combination algorithm based on service composition fragment tree was proposed in this paper.I n the service cooperative relationship autonomous planning model,each external service agent cooperates in the individual tuple space,and each external service agent evaluates itself according to the strategy of decision-making cooperation to evaluate whether it can be used as service group part of the participation of which,resulting in a number of service combinations of fragments,fragments of the relationship between the expansion of the composition of the composition of the composition of the tree.Multi-external service agent,personal tuple space,individual user based on the composition of the service group based on the composition of the composition of the tree combination tree algorithm to get closer to the user's desired service portfolio,and design pruning tree pruning strategy to ensure the efficiency of collaboration,and the algorithm flow was given.Through the analysis of the algorithm,the prototype system was designed and implemented to provide the prototype foundation for the practical application.