Imperfect debugging models with two kinds of software hazard rate and their Bayesian formulation
Shigeru Yamada, Kouichi Sera · Electronics and Communications in Japan (Part III Fundamental Electronic Science) · 2000
Ensuring the reliability of systems and their backbone, the software system has become increasingly important with the development of advanced information technology. Most of the models previously proposed as software evaluation techniques described the phenomena generating software failures or the events detecting faults. However, most of these models assumed a perfect debugging environment in which all of the faults detected during the testing phase or the operation stage are corrected or removed without introducing new faults. In practice, however, assuming an imperfect debugging environment which allows new faults to be introduced during correction is believed to be more realistic. Therefore, in this paper, we consider two kinds of software failures caused by faults introduced before testing began and new faults introduced during corrections to develop a software reliability model in an imperfect debugging environment and to examine the goodness-of-fit and validity of the model. We also introduce a Bayes theory for the constructed model and examine its applicability and utility to real data. © 2000 Scripta Technica, Electron Comm Jpn Pt 3, 84(3): 12–20, 2001