Software component quality-finite mixture component model using Weibull and other mathematical distributions

Software component quality has a major influence in software development project performances such as lead-time, time to market and cost. It also affects the other projects within the organization, the people assigned into the projects and the organization in general. In this study a finite mixture of several mathematical distributions is used to describe the fault occurrence in the system based on individual software component contribution. Several examples are selected to demonstrate model fitting and comparison between the models. Four case studies are presented and evaluated for modeling software quality in very large development projects within the AXE platform, BICC as a call control protocol in the Ericsson Nikola Tesla R&D.