Dependence Based EFSM Self Adaptive Model Mapping Technology

In recent years, the study of information and computer science shows a booming development trend, while software systems has been applied in almost every aspect of human life. With the increasing requirements of software intelligence, the study about self-adaptive system was beckoned.With the increasing size of software systems, the structure of software systems has been more complex. Model based software engineering has obtained an increasing number of attentions of researchers. Models, as the high abstract of software system, have obvious advantages of descripting structures and functions of software system, which provided a support for software development, test and many other areas. Extended Finite State Machines (EFSM), as a commonly used standard description, has been applied to descript dynamic system in many fields, such as internet and telecoms.Models can be applied in different environments and the environments may be changed, which is showed as restrict environment. The existing model mapping technology is seen as a model slicing, named as model projection. By this method, the reachability analysis is applied to reduce the environment influence and offer an analysis about the EFSM model behaviors in the restricted environment. Model projection has been applied in model checking and verification.In this paper, we proposed a new method, dependence based self-adaptive EFSM model mapping, to preserve the model functions in the restricted environment to some extent, which is different with the existing methods. Combing the information about the EFSM model which is obtained by dependence analysis, the self-adaptive model mapping rules defined by this paper is applied to obtain some elements in new environment. In addition, this method can be used to increase the efficiency of software development and improve the quality of software. The execution of this method can be seen as the process of identifying the "critical events", which is defined as the events that affect the model functions directly and cannot be repaired easily. By the method provided in this paper, it helps to give a reasonable allocation of resources both hardware and software.Compared with the existing method, the experience results show that, the models achieved by our method show a better performance about self-adaptive abilities.