R-tree Method of Matching Algorithm for Data Distribution Management

Data distribution management (DDM) is one of the services defined in the interface specification of high level architecture (HLA). Matching algorithm is the most important problem in the process of DDM implementation. The existing algorithms can only be used in certain condition, and the efficiency is not well enough for large-scale distributed simulations. The R-tree method is a new matching algorithm based on space index technologies. Rectangles of DDM regions are organized as an R-tree in the method, and operations related to matching can be done through the rectangle querying in the R-tree. To improve the efficiency of query, the method also uses a hashing index to point the leaf-node of the R-tree. The performance of the R-tree method is deeply studied in this paper. Experiments show that the R-tree method is better than other matching algorithms. It can efficiently save the cost of matching process of dynamic DDM, and can meet the need of real-time distributed interactive simulations. Changing the entry number of R-tree nodes can further optimize the matching algorithm.