Research And Implementation Of Data Distribution Management Algorithms In HLA

High Level Architecture is a framework of distributed simulation method,which is used to solve the problems of complex distributed system simulation.It is a new computer simulation system.The simulation framework separates the simulation process from the underlying communication through a special supporting structure and becomes an object oriented framework.The data distribution management service which is provided by high level architecture is the key to realize the large-scale simulations technology,its purpose is to send the data sent by the phase control,reduce the useless data transmission simulation system,reduce simulation delay,reduce the number of useless simulation nodes according to the processing overhead,enhance the simulation systems of the extensibilities.The contents and results of paper include:(1)This paper discusses the generation,development,advantages and research status at home and abroad of high level architecture,the development of applications in the life of the domestic and foreign research results of data distribution management algorithms;discusses several classical data distribution management algorithms of region matching algorithm and theirs implementation principles.(2)This paper analyzes the concept and implementation principle of classical sort-based data distribution management algorithm.In the original sort-based algorithm,each dimension of interactive information needs a corresponding matrix to store,that is,the need for a large number of storage space;In the execution of algorithm,we need to distinguish all kinds of boundary values.Based on these shortcomings,this paper proposes an improved sort-based DDM algorithm.The improved algorithm intersects information stored in a multidimensional matrix in all dimensions,greatly reduced the algorithm of storage space,but also reduces the matrix phase of "and";the improved algorithm release order boundary region and stored separately,reducing the decision boundary of a large number of improved algorithm the efficiency.And through the analysis of the experimental results,the improved algorithm has advantages in the time and efficiency of the execution of the system.Moreover,due to store the data in the stage of regional publishing and subscribe boundary values stored separately,and sorted in the information data is very large,under the condition of boundary value sequence will occupy a relatively large time overhead,so,in the improved algorithm,using the lower time complexity of algorithm will also help to reduce the complexity of execution time.Because in the initial stage of different types of storage areas are separately sorted,compared to the original algorithm,the improved sort-based algorithm select sorted method appropriate can help improve the efficiency of the algorithm,and through the analysis of experimental results of the different sorting algorithm in different environments,the effect of improved algorithm stability(3)This paper analyzes how to choose different data distribution management algorithms adaptively in different simulation environments.The experimental data show that each data distribution management algorithm will fluctuate in the execution efficiency in different simulation environments.In this paper,by setting the parameters that may influence the results,each of the three classic data distribution management algorithms execution time and accuracy of the matching results of experiments.This paper also analysis application in different complex environments by using quantitative results,data selection basis of data distribution management algorithm,improve the efficiency of simulation system.