http://swrc.ontoware.org/ontology#Article
Software Reliability Measurement with Prior - Information on Initial Fault Content
en
論文
Department of Industrial and Systems Engineering Faculty of Engineering Hiroshima University
Department of Industrial and Systems Engineering Faculty of Engineering Hiroshima University
Department of Applied Mathematics and Physics Faculty of Engineering Kyoto University
Department of Industrial and Systems Engineering Faculty of Engineering Hiroshima University
Mitsuhiro Kimura
Shigeru Yamada
Hiroaki Tanaka
Shunji Osaki
Most existing software reliability growth models have been restricted in use with respect to the software reliability assessments during the later stages of integration or system testing. By introducing the prior probability distribution on the initial fault content in a software system software reliability growth modeling as a binomial model is discussed here; allowing us to assess software reliability even during the earlier stage of testing. The model is described by a nonhomogeneous Markov process based on the assuinption that the fault-detection rate is proportional to the number of remaining faults in the system. In particular assuming the prior distribution to be a Poisson and a binomial distributions we discuss the software reliability measurement. Several assessment measures for software reliability and the maximum-likelihood estimates for the required model parameters are derived. The application of this model is demonstrated by analyzing a set of actual fault-detection data observed during integration testing.
Most existing software reliability growth models, have been restricted in use with respect to the software reliability assessments during the later stages of integration or system testing. By introducing the prior probability distribution on the initial fault content in a software system, software reliability growth modeling as a binomial model is discussed here; allowing us to assess software reliability even during the earlier stage of testing. The model is described by a nonhomogeneous Markov process based on the assuinption that the fault-detection rate is proportional to the number of remaining faults in the system. In particular, assuming the prior distribution to be a Poisson and a binomial distributions, we discuss the software reliability measurement. Several assessment measures for software reliability and the maximum-likelihood estimates for the required model parameters are derived. The application of this model is demonstrated by analyzing a set of actual fault-detection data observed during integration testing.
AN00116647
情報処理学会論文誌
34
7
1601-1609
1993-07-15
1882-7764