Research And Implementation On Data Filtering Algorithm In RTI

HLA (High Level Architecture) is an open and object-oriented architecture which has defined the application programming interface (API) for simulation. RTI (Run Time Infrastructure) is a federation run support system provides these API, and its function is similar with a distributed operation system. DDM (Data Distribution Management) is one of six management services of RTI. DDM realizes the data filtering through subscription/update mechanism to limit the production and receival of redundant data.This article discusses the concept and structure of HLA framework, the function and relationship between the various parts of RTI, studies the content and realization process of DDM, and pointed out the key issues in the process of system running. On this basis the following two key aspects are researched:(1) We make a detailed analysis about grid filtering, point out the strengths and weaknesses of grid filtering, discuss the optimal process of grid cell and propose a method of double-level data filtering based on dynamic grid. Apply the filtering template on the grid of subscription/update region overlapped, filter the rough data.This method not only improve the accuracy of the region matching, but also reduce the amount of data transmission in network and consumption of the multicast address.(2) In the complex simulation systems, with the advance of simulation time the changes of partial simulation entity's subscription/update region will affect filtering efficiency of systems. Based on the analysis of the relationship between grid cell-size and update time, the rate of data filtering, receiving time and queuing time, propose an adaptive dynamic grid-based method. This method dynamically adjusts the grid size to improve the efficiency of DDM according to the changes of simulation system.