Software reliability and cost models with warranty and life cycle

Abstract

With increasing number of tech-savvy users, the demand for software with advanced features is also growing exponentially. To meet these expectations, software firms are coming up with innovative strategies to attract customers to buy their products and remain loyal. One of the key strategies is to provide product assurance is post-sales support. Software firms provide a warranty to fix the failures encountered during the operational phase for a pre-defined time period. Software reliability literature consists of various modeling framework to decide release time under pre-defined warranty period under the consideration of no post-release testing. Very limited work has been done in the direction of post-release testing with warranty. In this work, we propose a framework to obtain optimal release and warranty time of software to minimize the overall cost under reliability constraint. We have provided the algorithm to obtain the analytical solution of release and warranty time. To make our proposed work more realistic we have considered the change point phenomenon and imperfect debugging of fault in our modeling framework. We have also conducted a sensitivity analysis on the model parameters to check their impact of the firms scheduling strategy. The proposed work is validated on a real-life data set.