A XML Tree Based Leveled Filtering Method and Its Application

摘要

The system's scalability is an important problem in the research of distributed virtual environment. By combining XML technology with active leveled interest management, a new XML tree based leveled filtering method -LevelFilter is presented in this paper. LevelFilter evaluates the level of interests among objects via the similarity matching of XML trees, and processes leveled packet filtering on active routers. Finally, by applying LevelFilter to the design of motion LOD (Level of Detail) in virtual human simulation system, it can manipulate the content detail of virtual human by the DOCBF (Detail of Content Based Filtering) to reduce traffic load. Therefore, LevelFilter can highly improve the system's scalability.