Software Reliability Growth Analysis And Dynamic Evaluation

As society continuously developing, computer science has been extensively advanced. Software products have infiltrated to various social fields so far, especially like the aerospace industry which is the crucial area of nation defense. Because the requirements in reliability and safety of software used in these areas are relatively higher than other fields, it has become an urgent task that ensuring and developing reliable software therefore. However, as the application scope expanding, the complexity and difficulty of software are increasing. People have to comprehend reducing development cost and enhancing software reliability. Software reliability growth has catch people's attention in this case. It is an importance way to improve and enhance the software reliability that implementing reliability growth, which can find and improve the software defects and drawbacks, raising its reliability and safety. Software reliability growth models play a very important role in researches, and the models of Non-homogeneous Poisson Process are the most widely used one.Combining with the engineering practice and current research achievements, theories and methods of software reliability growth were analyzed in this paper. The failure data collected by reliability growth test were sorted, analyzed and generalized according to their features. In the research of G-O model, Schneidewind model, Y-O model, M-O model, Duane model and other reliability growth model, this paper discussed many issues of these models, including the basic assumptions, analysis methods of formulas, how to select model and so on, which constructed the evaluation method for model parameters and reliability measures according to the failure data obtained in software test. As some examples verified, the dynamic evaluation of software reliability growth could be realized using test data and growth model together. Achievements in this paper can be applied as the technical support for researching and evaluating software reliability, which will play instructive role in enhancing and ensuring software reliability.