Modeling and analysis of reliability and optimal release policy of software with testing domain coverage efficiency

In order to verify whether the functions have been rightly implemented into the software, the developer is involved with the task of developing a set of test cases thereby influencing the set of functions and modules present in the software so that the developer can judge upon the defects laid into the implemented functions. The set of functions which these test cases influence comprises what in common terms is known as testing domain and that the rate of increase of the testing domain is vulnerable to the progress in testing process. The growth rate of the testing domain goes hand in hand with the fault content of the software, that is to say, as the testing domain spreads more number of existing faults are detected and subjected to removal, thereby causing the software fault content to decrease. Further the software developer is endowed with the task of determining the appropriate releasing time of the software into the market such that the cost is minimized and the reliability is maximized in the bargain. In order to counter effect the growing software fault content, this paper includes a testing domain dependent software reliability growth model (SRGM) incorporating the ideas of testing domain coverage efficiency as also the fault removal efficiency. The model so framed is based on Non-Homogenous Poisson Process (NHPP) and its validity has been verified by testing on fault data observed in earlier actual software development process. Along with illustration to the optimal time of software release the model has been subjected to goodness of fit comparison justifying that the developed model outperforms some of the existing ones in its capability of measuring the reliability.