Abstract
Many software reliability growth models based upon an NHPP (nonhomogeneous Poisson process) have been proposed to measure and assess the reliability of a software system quantitatively. Generally, the error-detection rate per residual error during software testing is considered to be dependent on the elapsed time of testing. In this paper, to describe such a software reliability growth aspect realistically, we classify the software errors detected by the testing into two classes: some are easy to be detected and corrected and the other difficult to be detected. We describe the error-detection phenomena of two classes of errors as the dissimilar NHPP's. Then, we propose exponential-S-shaped software reliability growth models by superposing the reliability growth processes for the two error-detection phenomena. Finally, numerical illustrations of software reliability assessment are shown by applying the actual error data, and the model proposed here is compared among the existing software reliability growth models in terms of goodness-of-fit.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
