Generalized-Gamma Software Reliability Model

摘要

In this paper, we consider a generalized software reliability model based on the non-homogeneous Poisson process (NHPP) under the assumption that the software fault detection time obeys a generalized-gamma distribution. The proposed model is regarded as a comprehensive model for several existing NHPP models, and is consistent to the probabilistic debugging theory based on the general order statistics concept. Further, we develop an efficient statistical estimation method for the model parameters based on the EM (Expectation-Maximization) algorithm and a heuristic approach.