Truncated Non-homogeneous Poisson Process Models — Properties and Performance

Abstract

All non-homogeneous Poisson process (NHPP) software reliability growth models of the finite failures category share the property that every time to failure distribution is defective. The reason for this phenomenon is the fact that according to these models it is always possible that no fault is left in the software. Excluding this possibility leads to a class of infinite failures category NHPP models which we call truncated NHPP models. In this paper, we derive the truncated models connected with three well-known finite failures NHPP models and discuss some of their properties. We then compare the performance of these new models with that of their counterparts.