Software reliability growth modeling with dynamic faults and release time optimization using GA and MAUT

Abstract

Software testing is an essential part of software life cycle as during this period, an effort is made to improve software reliability and quality. In this phase, perfect debugging is not possible because of time lag in fault removal process or new faults may get introduced in fault removal and fault detection process. In this paper, we have studied software reliability growth model (SRGM) incorporating generalized modified Weibull (GMW) testing effort function in imperfect debugging environment with constant and time varying fault detection rates, respectively. The parameters involved in the models are estimated using maximum likelihood estimation (MLE) and non-linear least square estimation (NLLSE) methods. The performance of the proposed models is validated using mean square error (MSE), accuracy of estimation (AE), χ2 test, etc. Moreover, optimal release policy is discussed by keeping fault detection rate as a constant using both genetic algorithm (GA) and multi-attribute utility theory (MAUT). A comparison has been made with existing models reported in literature. From the empirical results, it is observed that our proposed models performed better. Further, the reliability measures are more factual in the case of time varying fault detection rate in comparison to constant fault detection rate model.