Modeling and Distributed Simulation Techniques for Synthetic Training Environments

Synthetic Training Environments (STE) can be used to teach people to function within complex systems without the real-world limitations of safety, cost, training areas, or personnel. A cost-effective technique for creating these environments is to integrate a distributed system of computer simulations, virtual environments, and live participants. This paper characterizes the distributed simulation component of a synthetic training environment. It identifies salient issues for modeling and simulation for a distributed environment and discusses how this technology is currently used in synthetic training environments. Distributed simulations for the training environment require modeling of physical systems, environments, and human behaviors. In addition to the system models, a successful distributed simulation will have many specialized components, including user interfaces, graphics, networking, databases, and simulation management. Software design and implementation is complicated by issues such as complex interactions among models, real-time man-in-loop requirements, and transparently integrating real systems with virtual components so the STE presents a consistent and realistic virtual environment. One successful application of a simulation training tool is ModSAF, built by Loral Advanced Distributed Simulation. It is a Distributed Interactive Simulation (DIS) used for military training and combat doctrine development. ModSAF simulates the hierarchy of military units and their associated behaviors, combat vehicles, and weapons systems. The design provides an open, extensible software architecture, has proven to be scalable, provides fault tolerance, exhibits good transparency, and successfully allows for man-in-loop interactions to provide supervisory control of the situation being simulated. ModSAF provides a useful case study to examine key issues in using distributed simulation for training.