|Title||Software Process Evaluation Using a Customizable Pattern-Based Simulator|
|Publication Type||Master's thesis|
|Year of Publication||2008|
|Publisher||Schulich School of Engineering, University of Calgary|
Software process analysis and improvement relies heavily on empirical research. However, controlled experiments and surveys as means of empirical research have two major drawbacks. First, whatever evidence is gained via empirical research is strongly context dependent. Second, they are costly. Software process simulation modeling supports empirical studies by both reducing the cost of experimentation, and facilitating the combination of isolated empirical evidence. The process simulation model GENSIM 2.0, developed as part of this work, addresses the above challenges. Compared to the existing process simulation models in the literature, the novelty of GENSIM 2.0 is twofold: (1) Model structure is customizable to organization-specific development processes. This is achieved by using a limited set of generic structures (macro-patterns). (2) Model parameters can be easily calibrated to available empirical data and expert knowledge. This is achieved by making the internal model structures explicit and by providing guidance on how to calibrate the model parameters. The main achievements of the work are: Generic structures of software development processes referred to as macro-patterns, GENSIM 2.0, a customizable software process simulator developed based on the proposed macro-patterns and the V-Model lifecycle and complete and detailed description of the calibration of the simulator. This work also presents examples of useful application scenarios of the simulator including finding the best combination of Verification and Validation (V&V) techniques with respect to specific time, quality and effort goals and analyzing the effect of the project staffing profile on its performance.