The Research And Implementation Of Web Service Composition Simulation System Based On JADE Platform

With the development of the internet technologies and services, more and more Web services are distributed on the Internet, but the Web service with single function could not well satisfy the various needs. So an effective method must be found to reorganize these services as a powerful new service to provide a comprehensive and complex service for the needs. At present, it is lack of the dynamical and distributed simulation platform to simulating the combining process of Web Service.And it needs a large amount calculation in selecting the candidate services and the serial genetic arithmetic has not improved the system efficiency.Based on the above reasons, this paper proposed a theory to simulate different Web service using agents and stimulated Web service compositions with the interaction between agents on the multi-agent platform, and applied a parallel genetic algorithm to improve the calculation efficiency and then got the better service composition sequence more efficiently. According to the adapted methods this papers proposes a Web service composition stimulation platform. The main work is as follows:(1)This paper proposed the design principles to stimulate a Web service in the environment of the JADE platform. It was proved that the principle is validity by developing the multi-agent system. To facilitate subsequence developers extend the behaviors of agents this paper proposed a method of the Behavior of Warehouse in the design and implementation procedure therefore the system has a good scalability.(2) This paper proposed the layering design strategy. By configuring the association attributes of Agent, Behavior of the Agent and the association methods of Behavior, the design of agent was split into three layers, the presentation layer, the behavior layer and the method layer, which respectively were responsible for the upper agent design, for a behavior design and for the concrete function design in a behavior. The layering design strategy simplified the designing procedure of Agent based on the JADE platform.(3) First, the feasibility of the multi-Agent to simulate Web Service Combination was proved by simulating the experiment based on the QoS local optimum service combination. Second, the feasibility of parallel genetic algorithm in the system was proved by getting the global optimum Web service sequence with the parallel genetic algorithm and the simulation experiment for service combination. Lastly, this paper found a better deployment strategy most suitable for this platform by analyzing the genetic algorithm calculating result during the global optimum selection experiment.