Optimization of software architecture selection for the system under design and reengineering

The problem of optimization of software system architecture choice is considered taking into account changes of requirements in iterative technologies of design and reengineering. The solution is offered on the base of correction of alternatives' characteristics which were accounted on previous iteration or modification of existing software system during reengineering. This method prevents recalculations for evaluation and selection process. The method of pairwise substitution is applied for alternatives' characteristics correction. Its concept lies in compensation on supremacy of criterion change. Multicriteria optimization of substitution is carried out using nonlinear scalar convolution what improves validity of selected decision.