A Methodology for Software Reliability Based on Statistitical Modeling

Abstract

Reliability is one of the quantifiable quality features of the software. Software Reliability Growth Models(SRGMs) are used to assess the reliability achieved at different test times based on statistical learning models. conventional time-based SRGMS may not be accurate enough in all situations , and such models cannot identify errors in small and medium-sized applications. Numerous traditional reliability measures are used to test software errors during application development and testing. In the software testing and maintenance phase, however, new errors are taken into account in real time in order to determine the reliability estimate. In this article, we suggest using the Weibull model as a computational approach to solving the problem of software reliability modeling. In the proposed model, a new distribution model is proposed to improve the reliability estimation method.  We evaluate the model developed and compare its performance with other popular growth models from the literature. Our test results show that the proposed Model is superior to S-shaped Yamada, Generalized Poisson, NHPP.