Converting UML Description of Software Architecture to Stochastic Process Algebra and Performance Evaluation
Rahmat Zolfaghari, Islamic Azad University, Hashtgerd Branch, Department of Computer Engineering, Tehran, Iran.
Manuscript received on May 23, 2013. | Revised Manuscript received on June 03, 2013. | Manuscript published on June 30, 2013. | PP: 227-232 | Volume-2, Issue-5, June 2013. | Retrieval Number: E1761062513/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Important qualitative parameters of the large software systems are determined by indicators of effectiveness of the software’s, such as response time, operating power and error rate. Procedure modeling is an approach for evaluating the effectiveness andvalidationof the systems and, as well as it predicts the requirements of qualitative and quantitative performance and provides a comparison between all kinds of designs with respect to performance indicators. Present study suggested a method for converting the UML description designing software to Stochastic Process Algebra (SPA)model, which provides the application of using the UML in designing software with high performance; in other words it putting the performance in designing software and a high quality software is designed.In order tomodeling the parts of system we use state chart and for the interactions between the parts we use the Collaboration chart with the performance profile (using the performance profile is the distinction between the suggested approach and the former ones.). An algorithm is provided for automatic production of the SPA performance model from the XML(Extensible Markup Language) documents and state and collaboration charts with performance profiles(stereotype, label and limitation), using the Export XMI software, Argo UML is provided, and then they gained SPA performance/operation model in the PEPA workbenchtool is loaded for performance analysis, so as the designer can test thefulfill of performance goals of his design according to type of different performance parameters and changing in value and chooses the best option in designing.
Keywords: Performance Evaluation, UML, Stochastic Process Algebra and performance profile.