The Simulation Networking (SIMNET) program demonstrated the feasibility of conducting tactical training and combat development in a distributed simulation environment. As other programs build on the success of SIMNET, the number of types of vehicles simulated will continue to increase dramatically. In a distributed interactive simulation system, every vehicle is displayed by other manned vehicle simulators in several ways: visually, aurally, and with infrared, radar, sonar, and other sensors. Currently, every manned simulator must keep a set of representations for every other type of vehicle. Unfortunately, this means that whenever a new type of vehicle is added to the network, all other entities on the network must be updated to include representations of the new vehicle.

Alternatively, each simulator could bring with it a generic representation of itself which could be distributed to other entities on the network. Such an approach is referred to as a distributed representation system. This paper discusses a set of extensions to the Distributed Interactive Simulation (DIS) protocols that would be necessary to implement such a system.