An Architecture Model of Distributed Simulation System Based on Quotient Space

It is an important and difficult work to build a consistent and efficient architecture for large scale distributed simulation systems based on their strong and complex standard constraints. On the system design level, the information we have is more abstract, fuzzy and inaccurate. So a new architecture modeling method is needed. In this paper, we propose an architecture design model based on the quotient space theory. The quotient space theory is an important theoretical branch of granular computing and concentrates on the relationships and architecture of the different granules. In our model, system design information is integrated into different types of granules, and intends for the further optimal design. This model considers the simulation tasks, resources and services as a whole, and gives a two-step modeling process on system architecture design, including task-resource allocation model and resource-service allocation model. We illustrate our method with an information system example to show the modeling and solving process. The primary result shows that it has reduced the period of system design and increased system consistency and efficiency.