Implementation of the optimizer of SOA system deployment architecture

Abstract

Optimization of business processes in SOA systems has been done using three separate types of methods: Resource Allocation, Service Scheduling and Service Composition. All three may influence each other, so the new method has been proposed to find an optimal combination of those three. It is based on a genetic algorithm that uses a simulator of the SOA system to evaluate solutions. The article describes a model for the optimization criteria for such solutions. Subsequently, some basic concepts used to implement the simulator and optimizer have been presented. Finally, the performance results of the optimizer have been described, including the conclusions on how they might be improved.