Combinatorics based problem specific software architecture formulation using multi-objective genetic algorithm

In Software Development Process, the design of complex systems is an important phase where software architects have to deal with abstract artefacts, procedures and ideas to discover the most suitable underlying architecture. Due to uncontrolled modifications of the design and frequent change of requirements, many of the working systems do not have a proper architecture. Most of the approaches recover the architectural blocks at the end of the development process which are not appropriate to the system considered. In order to structure these systems software components compositions and interactions should be properly adjusted which is a tedious work. Search-based Software Engineering (SBSE) is an emerging area which can support the decision making process of formulating the software architecture from initial analysis models. Thus component-based architectures is articulated as a multiple optimisation problem using evolutionary algorithms. Totally different metrics is applied looking on the design needs and also the specific domain. Thus during this analysis work, an effort has been created to propose a multi objective evolutionary approach for the invention of the underlying software system architectures beside a versatile encoding structure, correct style metrics for the fitness operate to enhance the standard and accuracy of the software system design.