Data Colleting Service In Distribute Simulation Based On Mobile Agent

In the distributed interactive simulation, data collection is an important issue. The data collected can be used for simulation of the post hoc analysis, online monitoring and the simulation replay and etc. Data collection services in National Economy Mobilization Simulation and Training (NEMST) put new demands, the data collecting not only need to collect RTI's public data, it also need to collect the internal data of the simulation models. The data collection has many characteristics -- collection tasks with changing demand, dynamic creation of simulation objects, large amount of simulation data and real-time query and so on.The traditional collection methods are generally divided into centralized data collection and distributed data collection. Centralized data collection methods occupy big network bandwidth, has a great impact on the original simulation system, and a single collection federate will cause a single point of effect. Distributed data collection has online analysis and real-time query difficulties.This paper presents a mobile agent based method to remedy the defects of the original methods.It also show the basic design and implementation.Finally this module is used in a practical team trainning simulation system.This case show that the module has advantages in the collecting autonomously and little impact on the existing system.The module achieved satisfactory results in the availability and efficiency.