Research On Key Technology Of Adaptability In Dynamical Software Product Lines

Software system needs to dynamically adapt itself in response to the changing conditions and user new requirements at runtime.So it may be required the capability of dynamical adaptation and strong variability and flexibility.Currently,dynamic software product lines(DSPLs)produce software products capable of adapting to requirements that change at runtime and it is essential to study its dynamical adaptation.On one hand,the study on evolution of post-development from the perspective of macro-level software architecture of software systems,describes the dynamic change behavior of software systems very well,further more the research on the evolution model can better guide and make the model of software architecture dynamic evolution.On the other hand,the autonomic reconfiguration of pervasive system describes an approach that based on variability analysis and management,that approach transfers the knowledge from the software product line design to the online products and can make the system to autonomic adapt its behaviors spontaneously to achieve dynamic reconfiguration.Specifically,this paper mainly includes the following aspects:Firstly,for dynamic evolution of software architecture,adaptive evolution model and the invasive semantic-evolution model have been redescribed in detail.In regard to some defects in dynamic evolution of software architecture based on cellular automata model in recent research,including that specification is not detail and relationship between cellular is not clear,that it can't describe the constraints in process of dynamic evolution,this paper research refined the extend-cellular automata model of dynamic evolution of software architecture,which was based on the combination of evolutionary constraints applied to analyze the dynamic evolution of software architecture process,and SA dynamic evolution was correctly guided by applying both inter-cell behavior related constraints and control constraints.Also,the analysis on consistency and degree of evolution could more exactly guid SA dynamic evolution than ever on cellular automata model,which advanced the further study on dynamic evolution of software architecture.Finally,the case study shows that the proposed method is valuable and feasible,which can more fully applied to the dynamic evolution of software architecture.Secondly,according to the dynamic changing resource and user goal for systems,a graphical approach that displays variation scenarios was proposed based on the special analysis.The proposed approach has classify those scenarios into two categories——evolution and involution scenarios.Moreover,the comparison analysis between evolution and involution scenarios in different points may pave the way for the following.Finally,It researches the reconfiguration of systems and illustrates the proposed approach through ventilator case study.We have classified the approaches for developing CP using DSPLs in two categories from the product different perspective.By reusing reconfiguration knowledge and extend the DSPL and introduce an autonomic reconfigurator component,we demonstrate how to transfer the knowledge to the runtime configuration product.To know about the reusing knowledge in particular situation,the work applys model driven development and then performs some steps in detail.Futher more,we illustrate the reconfiguration applying the ventilator case analysis for the valuablity.