A New Algorithm for Performance Evaluation of Homogeneous Architectural Styles

Software architecture is considered one of the most important indices ofsoftware engineering today. Software Architecture is a technical description of asystem indicating its component structures and their relationships, and is theprinciples and rules governing designing. The success of the software depends onwhether the system can satisfy the quality attributes. One of the most critical aspectsof the quality attributes of a software system is its performance. Performanceanalysis can be useful for assessing whether a proposed architecture can meet thedesired performance specifications and whether it can help in making keyarchitectural decisions. Architecture style is a set of principles which an architectuses in designing software architecture. Since software architectural styles havefrequently been used by architects, these styles have a specific effect on qualityattributes. If this effect is measurable for each existing style, it will enable thearchitect to evaluate and make architectural decisions more easily and precisely. Inthis paper an effort has been made to introduce a new algorithm for investigatingperformance in homogeneous architectural styles based on Markov chains.Moreover, How to use this Markov chains model for performance evaluation isshown and simulation and the implications of this transformation are describedcompletely. Finally, to represent the usage of our proposed algorithm, a case studyis presented as an example.Keywords: Software architecture,

کلیدواژه ها:

Software architecture ، Markov model ، Architectural styles ، Performance attribute ، homogeneous styles