Dynamic Pervasive Service Composition Research Based On ASM

Along with the development of Internet technology, sensor network and mobile technology, pervasive computing gets more and more spotlights, more and more pervasive devices which provide versatile service functions have been brought forth in our working and living environment. A major challenge in pervasive computing is how to integrate all these provided services effectively and meaningfully to form novel service composition applications. This paper analyzes the problem of dynamic service composition in the environment of pervasive computing, and then proposes a formal model for dynamic service composition under pervasive computing. Finally verify the feasibility of the model by CoreASM.Firstly, based on the research in existing service composition technology, we propose a dynamic service composition framework in the environment of pervasive computing. First, we use the hierarchical structure service organization mode, which is to organize all the services in a layered-architecture, and then upon the hierarchical service model, we compose the services automatically based on Rule Tree. It first generates an optimal Destination-Path for each output that a user asks for, after that it composes all Destination-Path to construct the composite service that satisfies the user's request. This framework has the characteristics of high efficiency in re-composition and low complexity.Secondly, based on the framework above, we use ASM to build a ground model for dynamic service composition in pervasive computing, and then analyze the input part, output part and process part of every module in that ground model, and at the same time, refine the ground model, that is, for every part, we give an ASM rule to simulate its behavior.Finally, we choose the ASM engine CoreASM as our tool to verify the feasibility and correctness of ASM model we propose above. And the results of the experiments give us a conclusion that our ASM model can build the system of service composition under the environment of pervasive computing effectively, and it has a certain degree of practical significance for further application development under pervasive computing.