A Model Based Approach on Multi-Agent System and Genetic Algorithm to Improve the Process Management in Service Oriented Architecture

Abstract

Service oriented architecture is based on the provision of services. To enhance the performance of the systems by providing a better combination of services, it is necessary to extract more information compared to the one in the service registry. In this regard, the accomplished works have been focusing on the basic concepts of service-oriented architecture. The service composition is based on the information in service registry, provided by the service provider. Further, centralized combination with insufficient information does not meet the system performance requirements. This solution helps to facilitate resource distribution and reduces tasks of the central unit. In this paper, efforts have been made to use the agents in the proposed model to enhance the processes of a system. In the proposed model, agents have been used in order to extract and manage the essential information in service registry table. This information forms the basis of monitoring and selection of services. Agents are used for selecting and making efficient composite services. Finally, different communications and configuration mechanisms are implemented using multi-agent systems that can perform service-oriented architecture. Moreover, genetic algorithm is used to enhance architectural processes. In the proposed model, the genetic algorithm and multi-agent system has enhanced productivity of the system and its important quality attributes. The system runs in the international conference of computer society of Iran (ICCSI). Implementation of this model in the real environment and its comparison with its implementation on the prototype could be helpful to justify better efficiency and accuracy for future applications.