دانلود رایگان مقاله انگلیسی تحلیل کاستی و پیش بینی با به کار بردن مدل رشد قابلیت اطمینان نرم افزار چند مرحله ای - IEEE 2017

Abstract

In software development, defects are inevitable. To improve reliability, software reliability growth models are useful to analyze projects. Selecting an expedient model can also help with defect predictions, but the model must be well fitted to all the original data. A particular software reliability growth model may not fit all the data well. To overcome this issue, herein we use multistage modeling to fit defect data. In the multistage model, an evaluation is used to divide the data into several parts. Each part is fitted with its own growth model, and the separate models are recombined. As a case study, projects provided by a Japanese enterprise are analyzed by both traditional software reliability growth models and the multistage model. The multistage model has a better performance for data with a poor fit using a traditional software reliability growth model.

V. CONCLUSION

Both the software reliability growth model and the multistage model are successfully fitted to the data of two projects. Moreover, we developed an evaluation and prediction tool, allowing the company in the case study to appropriate allocate engineers for the debugging process, conserving resources and costs. For RQ1, we used the RSS value and the AIC value as standards to evaluate the growth models. For RQ2, AIC is used, demonstrating that the multistage model can provide a good solution when a project is ill fitted with a traditional growth model. Although the applicability of the multistage model is confirmed using a case study of a company, it is applicable to other database defects with stochastic events. This method provides a new point of view in software reliability growth models. In the future, we plan to expand the multistage model to consider other growth models (e.g., the average combination of different growth models). These new models should contribute to software reliability and may help resolve problems that are difficult to solve using traditional software reliability growth models.