Content area
Full Text
This research article aims to identify and introduce cost-saving measures for increasing the return on investment during the Software Development Life Cycle (SDLC) through selected quantitative analyses employing both the Monte Carlo Simulation and Discrete Event Simulation approaches. Through the use of modeling and simulation, the authors develop quantitative analysis for discovering financial cost and impact when meeting future demands of an organization's SDLC management process associated with error rates. Though this sounds like an easy and open practice, it is uncommon for most competitors to provide empirical data outlining their error rates associated with each of the SDLC phases nor do they normally disclose the impact of such error rates on the overall development effort. The approach presented in this article is more plausible and scientific than the conventional wait-and-see, whatever-fate-may-bring approach with its accompanying unpleasant surprises, often resulting in wasted resources and time.
Keywords: discrete event simulation (DES), Monte Carlo simulation (MCS), error or defect rate, return on investment (ROI), software development life cycle (SDLC)
The science behind software development in metric terms of return on investment (ROI) is well known and taught by many. Much work has been accomplished in this area albeit lacking details of execution on a real-life problem (Ferreira, Collofello, Shunk, & Mackulak, 2009; Zhang, Kitchenham, & Pfahl, 2008; Zhang, Kitchenham, & Pfahl, 2010). The art of software development is a learned behavior and not one with which everyone becomes comfortable due to its intricacies and learning cycle. The same may be said with respect to software development life cycle (SDLC) management and distribution as depicted in Table 1, where the different phases of an SDLC process, when applied, provide specific inputs and expected outputs.
Life Cycle Phase (Process) Flow
As with many processes, there is a beginning point and a delivery epoch. SDLC methodology is no different. It enables standardization for planning and organizing, and facilitates cost estimation. Though there are several different models available, many are tweaked to best fit the current process or a sequence of activities in a software development project. The life cycle used in this article (Table 1) has nine phases beginning with the requirement review and ending with the deployment decision. As one begins with the first phase (i.e., requirement...