Component assessment and proactive model for support of dynamic integration in self adaptive system

Abstract

Component Based Software Engineering (CBSE) is a paradigm in use by most of the software developers. In a multi component system, each component is a probable point of malfunction. Typical work to make such system more vigorous and safe are both brittle and time intense. A model has been designed for self-adaptive system that automates the component integration process at runtime by accessing the equivalent component from diversified set of components that may be needed in future. The proposed general model is for proactive adaptation, which pre-fetch the component from the pre available repository. This model integrates the caching technique to reduce the amount of time that has been spent during search of best-fitted component to replace the required one, when a system fails to respond due to component failure. To pre-fetch the required component we perform the component assessment on the basis of numerical metadata for each component present in the repository. In the computation of numerical metadata we design an algorithm which uses the concept of Abstract Syntax Tree. To ensure the consistency in the system after the modification, we have used a technique which is used in GUI based component architecture model. To asses the component at run-time, we provide prototype in .Net technology using its attribute feature, which support run-time component evolution without its execution.