GTRI

Case Study

Creating a New Networking Standard for Defense Simulations

Published: December 2, 2000


Click for article gallery (1 image).

Soldiers and pilots who can train together on networked simulators not only learn how to handle military situations, they develop the teamwork that's essential for a successful mission.

But F-16 pilots operating a simulated mission in an anechoic chamber in Maryland might not be network compatible for a simultaneous mission simulation with an enemy air defense system in Houston and a test control group in New Mexico. Likewise, NATO member nations might have difficulty with allied forces training using various nations' warfighting systems simulations.

The solution is a common, high-level network architecture for an array of simulators widely differing in age, purpose and technological complexity. The U.S. Defense Department began work on this new standard in the mid-1990s under the auspices of the Defense Modeling and Simulation Office and a consortium of government, industry and academic researchers.

Among those involved are Georgia Tech Research Institute (GTRI) engineers, who began supporting the development of the new High Level Architecture standard four years ago. It is challenging work, in part because the development of High Level Architecture (HLA) — a general purpose architecture for simulation reuse and interoperability — is dependent upon a host of ever-changing technologies, says Senior Research Engineer Margaret Loper. She is one of 12 members of the Distributed Simulation Systems (DSS) research team, a part of the Information Technology and Telecommunications Laboratory at GTRI.

Distributed simulation is a technology area that has grown increasingly reliable and realistic because of the evolution of all the technologies that go into it, Loper explains.

"You can run very powerful, realistic-looking simulations on a PC, whereas 10 years ago you had to have a very high-end Silicon Graphics machine to be able to do it," Loper says. "So the networking simulations are just one part of the picture. You also have to deal with all the advances in graphics and visual systems, hardware platforms, communication networks, artificial intelligence and behavior modeling. The list goes on and on."

GTRI's responsibilities in the development of technologies that support HLA embrace several major initiatives.

High Level Architecture Testing and Support

GTRI's original role in the HLA project was development of the test process, procedures and tools for determining whether a particular simulation complies with HLA. A professor from the College of Computing was also assigned the task of designing and documenting time-management services for the HLA interface.

Federation Verification Tool

Another testing project, this effort focuses on testing a group of networked simulations called a federation.

"Once you bring your simulations together and you've got them talking, sending data back and forth using the HLA, how do you verify that all the simulations are really doing what they're supposed to be doing?" says Loper, in framing the task before the GTRI researchers.

Typically, a network of simulators is employed to model a large, complex scenario. The scenario's accuracy depends upon the proper interaction among the individual simulators in the network.

"Each simulation has its own set of rules and responsibilities about what it's supposed to do — whether it's modeling aircraft or tanks, where and when it's supposed to move, what it's supposed to fire at and under what conditions," Loper says. "The federation verification tool runs while the simulations are interacting. It knows what each simulation is responsible for, and it verifies that each one is doing what it is supposed to do. If there's a problem, the tool can pinpoint the simulation that may be causing others problems in the network."

Advanced Simulation Technology Thrust

When it comes to getting simulations to work together, timing is everything. The trick is to develop time-management specifications that eventually will allow synchronization of simulations varying greatly in internal architecture, function and purpose. The specifications would even address the need to track time — some simulations are more time-management reliant than others.

Dr. Richard M. Fujimoto of the Georgia Tech College of Computing accomplished the first part of the task by devising architecture specifications that allow simulations with different needs to share a time-management reference. The project's next phase involves developing the next generation of time-management algorithms that will allow disparate simulations to effectively work together.

High Level Architecture Simulation Interface

With the advent of the HLA, existing simulations can be called legacy systems because they were not designed to operate in the new HLA environment. So GTRI is now developing interfaces, tools and procedures to allow legacy systems to migrate or connect to HLA.

This project has resulted in the development of a software tool, the Distributed Simulation Interface Framework (DSIF), which is being used to migrate simulations to HLA.

MIMI HLA Model Migration

In another project, GTRI researchers have drawn upon their work in HLA to bring an operational legacy simulation into an HLA environment.

MIMI, which stands for the Mobilization and Deployment Capability Assurance Project (MADCAP) Integration Management Initiative, is a model used by the U.S. Army Forces Command to plan the mobilization of soldiers for military contingencies. GTRI used the DSIF tool to build a software interface between MIMI and the HLA.

"Now that MIMI supports the HLA standards, the model, which was previously used only in stand-alone mode, can be linked with other DoD simulations for training, analysis or acquisition," says DSS researcher David Roberts.

Human Behavior Modeling and Model-Based Design

One of the new research areas for the DSS group combines its HLA experience with research conducted in Georgia Tech's School of Industrial and Systems Engineering by Dr. Christine Mitchell. DSS researchers are building a model of the decision-making process used by a military squad leader operating in an urban environment. This model will help evaluate the information requirements needed to design computer displays helpful to the soldier. Using HLA, this model of the soldier can be integrated with other simulations to evaluate requirements in a realistic simulated environment in real-world military exercises.

"This project is a good example of how the HLA can make a difference," Roberts explains. "Simulations that were developed for different reasons can now be brought together into one environment for training and new concept development."