Research And Application On Dynamic Combination Of Semantic Component For Adaptive Middleware

With the proposal of ubiquitous computing environments, as well as the rise of component-based software development, the research on component-based adaptive middleware has attracted a wide spread attention. However, traditional components lack description of behavior and interaction, leading to unpredictable combination of components and difficult in meeting complex services. It's imprecise and lack of standard while analyzing equivalence and compatibility of component behaviors. Meanwhile, it takes no analysis, comparison and optimized selections among different adaptive actions, reducing the overall system reliability, stability and efficiency greatly.Aiming at the problems above, this paper introducing semantic component into adaptive middleware, allowing the system with self-adjusting ability facing the changes of external environment. Meanwhile, we analyze behavior equivalence of components based onπcalculus before dynamic adjustment of components, and ensure the compatibility of new component in original system. Also we make comparison and optimized selection among different adaptive actions, and ensure the reliability, stability and efficiency of whole system to meet users'requirement of service quality. The main work in this paper is as follows:Firstly, this thesis proposes the formalized modeling of component dynamic behavior based onπcalculus, and puts emphasis on analyzing and verifing component behavior equivalence and compatibility. When making research on component behavior equivalence, we first put forward some theories and give concrete examples to analyze behavioral equivalence, then the Mobility Workbench is used to model and verify the component behavior equivalence. When studing on component behavior compatibility, we first give some methods for behavioral compatibility analyzing then give a case study. The behavior equivalence ensures component interface matching and the consistency of component functional behavior. The behavior compatibility guarantees the correctness of internal process structure and inexistence of deadlock or inaccessible state, ensuring the normality, stability and dependability of the whole system.Secondly, based on the researchment of anterior several chapters, we put forward the dynamic adaptive combinatorial approach of components. Thereinto, adaptive factors, used to measure adaptive ability of combinatorial operations, are introduced into component combination. And this paper classifies and calculates the adaptive factors later. Besides, the component dynamic combinatorial flow is given, allowing comparison and optimized selection among different adaptive actions according to the value of adaptive factor. Finally, this paper designs and develops a mobile music system, in which we use the dynamic adaptive combinatorial approach based on semantic component. We first give a brief introduction and a specific design of the system, then present the method of component development and implementation. In the end, we demonstrate that the system has the merits of dynamic adaption and excellent performance quality.