A pragmatic approach to traceability in model-driven development

Abstract

A common problem in model-driven software development (MDSD) processes is the tracing of requirements across different phases of the software development life cycle and multiple levels of abstraction down to the code level. Because debugging at the model level is not feasible yet, unwanted or unexpected behavior of the executable system needs to be analyzed at the code level at run-time and in a feedback loop must be traced back to and handled at the model level. Thus, traceability is a very important success factor and quality criterion in software engineering and maintenance and especially when developing high-quality model-driven infrastructures. In this paper, we present the conceptual design and prototypical implementation of a lightweight traceability approach which supports tracing requirements across different models and levels of abstraction. While providing support for representing different types of traceability links between design models and implementation details, our approach can easily be integrated into existing MDSD projects without increasing their complexity.