The Research Of Software Reliability Based On Markov Analysis Method

In the software reliability domain, the research about the modeling of software reliability is one of the earliest aspects. There have been hundreds of software reliability models, but these models usually treat the software as a whole when they describe software reliability, however, these hypotheses incompletely conforms to the software system actual composition situation. Therefore, more or less limitations exist in these kinds of models; So People keep on exploring models to evaluate software reliability.In this paper, we apply Markov analysis theory to the software reliability prediction and discuss one kind software reliability model based on the module reliability and the module importance, in order to achieve the requirement of improving software quality. The proposed model utilizes the reliability data collected from software development projects and analyzes the module reliability and the module importance separately to study the whole software reliability of the system. Through application of this model to the actual project, we find that the new model has a better prediction precision and achieve ideal results. At last, we design a model system based on the new model to realize the automation of software reliability prediction.