A dynamic mapping method for reducing migrations in DVE systems

In a Distributed Virtual Environment (DVE) system, a multi-server infrastructure is often used to improve scalability and responsiveness. With this infrastructure, every client needs to be mapped to a proper server, and the mapping result is crucial because it affects the overall system performance and the users' interactive experiences. Existing mapping methods primarily consider the load balancing problem and the communication delay constraint between clients and servers but ignore the influence of client migrations at run-time. In this study, we propose a novel dynamic mapping algorithm which transfers cells between servers according to a cell selection scheme based on comprehensive factor computations. Experimental results show that our algorithm can decrease the probability of migrations and improve the users' degrees of satisfaction effectively in DVE systems.