A Method for Data Interaction of Large-Scale Distributed Battle ...

multiple TCP streams between the data source and
destination. In this method, data can be transmitted in piece.
IV. APPLICATION CASE
Now, we give process of radar detection finding the
target in the surrounding environment in the battle simulation
to indicate the advantages of this method. In the tradition
technology system, various types of targets in the simulation
environment need to give their characteristics and movement
to radars. Then whether the targets are detected is decide by
the radar’s own performance and the detection information is
given. If there are several radars in the simulation, they all
simulate in this way. The interaction relationship between
nodes in the network is irregular mesh. If using the method
in this paper, various types of targets only need to give their
characteristics and movement to objective situation server.
Then the objective situation server coordinates and forms a
unified objective situation. The interaction relationship
between nodes in the network is the star model. Radars
cognize and judge this unified objective environment and
then send the cognizing results to the cognizing situation
servers. Increase or decrease in the number of radar, the
target environment, the increase or decrease of the radar’s
number does not affect the simulation system. It is no need to
change programs. Figure 4 shows simulation system of two
nodes. The node 1 is running a simulation model of aircraft 1,
ship1 and radar1 and node 2 is running a simulation program
of the aircraft 2, ship 2 and radar 2.The output data of
platform class entities in each node form an objective
situation layer on the server, the objective situation of layers
of two nodes are communicated through data sharing
mechanism, resulting in the objective situation layers of two
nodes in exactly the same. They provide the same cognizing
environment to the radar model running on these two nodes.
Figure 3. Ilustration for a Cross-domain Simulation System Constituted
by Two Nodes
V. CONCLUSION
The core idea of battle simulation interaction
organization is to divide the nodes into troop entity nodes,
communication organization nodes and command nodes. The
“Produced / Consumed” data of various types of nodes is
divided into objective situation layer data, cognizing
situation layer data and command situation layer data to
manage. By using “layer” as the interaction center, the layer
synchronization and separation between layers of simulation
situation data are achieved. In general simulation network,
the objective situation server, the cognizing situation layer
and command situation layer are added to maintain various
types of data.
Various types of entities in simulation have no direct data
exchange, but interact indirectly through the situation server.
It can solve the problem which the HLA/RTI technology
cannot solve in multiform nets and has the following
advantages:
1) It is coupled in low level and easy to expand. In
traditional simulation systems, a new simulation entity need
and may be related to occurrence of data dependencies to
contact all the nodes. Using the system architecture provided,
when a new entity joins simulation system, only need to
establish data producing and subscribing relationship within
the system node according to the types of data produced and
subscribed, and the entities that already exist within the
system does not need directly interaction。
2) It greatly reduced the scale of data interaction in the
large-scale distributed simulation. Various types of entities
don’t need data interaction any more, but indirectly through
various situation servers’ data interaction. Supposing there
are m combat platform entity nodes, n sensor entity nodes. In
traditional battle simulation system, the need to establish the
relationship need m× n data interaction, and using this
method, the data interaction reduced to m + n .
3) It is Easy to organize simulation model validation.
When conducting simulation entity model validation, the
input and output information of various entities can be
directly collected and analyzed on the server. The radar
model reliability can be validated through collecting
surrounding target information on the objective situation
server, radar cognizing information on cognizing situation
server. The radar detection model can be calibrated and
validated with the radar performance parametersAuthors and
Affiliations
REFERENCES
[1] WANG Xing-ren. “Development and Application of Modelling and
Simulation,” Science & Technology Review, vol.25, Feb. 2007,
pp.22-27.
[2] PENG Ying-wu, HU Xiao-feng, YANG Jing-yu, and REN Jun.
“Study on Warfare-Analysis Simulation System Based on HLA,”
Journal of System Simulation, vol.17, Nov. 2005, pp. 2796-2800.
[3] Kwak,S.D., Joint Synthetic Battlespace Technical Architecture(JSB-
TA), 1st ed., 2001.
[4] IEEE Std 1278.1-1995,IEEE Standard for Distributed Interactive
Simulaiton - Application Protocols,1995.
[5] WANG Yi-fu, CHEN Song-qiao. “Research and Emulation on Nonuniform
Information Dissemination for Grid Resource Discovery,”
Journal of Chinese Computer Systems, vol.28, Dec. 2007,pp.2203~
2207.
[6] Ling Yun-xiang. Model Searching Algorithm Based on Response
Order and Access Order in War-Game Simulation Grid, Berlin
Heidelberg:Springer-Verlag Press, 2006.
106