Generalised framework for optimal pre and post release software testing in presence of warranty

Abstract

Significant advances in the field of software testing have been made by providing extended warranties, operational support and desired software reliability to the customers. However, the trade-off between minimising testing cost and maximising reliability to ensure well-timed market penetration has been a key concern for developers. Separate release and testing stop time strategy in order to optimise overall testing and market opportunity cost is proposed. We propose a generalised framework for developers to achieve multiple objectives of minimum testing and market opportunity cost; optimal warranty length; and optimal testing and release time. Software reliability growth models (SRGMs) are used to model an average number of bugs detected by testers (users) during the pre (post) release of software. Significant implications and quite intriguing facts related to separate release and testing stop time; early release ensuring higher reliability and; capitalising market opportunity to minimise overall testing cost are suggested based on proposed optimisation models. Numerical illustration is provided to validate the proposed framework.