Simulation-based interval estimation of software reliability measures

摘要

The software intensity function is one of the important reliability measures in non-homogenous Poisson process (NHPP) based software reliability model. In previous study, we proposed interval estimation method to derive the confidence interval of software intensity function to take uncertainty into account by introducing simulation-based bootstrap, where thinning algorithm was used to simulate NHPP. In this report, toward this end of investigating the effect of different NHPP simulation methods to the result of confidence interval of software intensity function, we use four kinds of NHPP simulation methods to generate bootstrap samples, based on which the confidence interval of the software reliability measure will be constructed.