Data Distribution Management for Distributed Supply Chain Simulation

Interest in Distributed (Interactive) Simulation has over the last two decades been mostly dominated by the military and the Department of Defense. More recently, non-military, commercial applications are starting to appear in this area. Examples of these applications can be found in traffic control management, supply chain simulation and logistics. The simulation of a Distributed Supply Chain typically involves the combination of several stochastic discrete event models, where each model is contained within a simulation package such as Extend, Simul8, Witness, Arena. Several key issues must be addressed in the realization of a Distributed Supply Chain Simulation, examples of which are model registry, model linking, data distribution management, time management. This paper discusses the issue of Data Distribution Management (DDM) in Distributed Supply Chain Simulation. In the context of Supply Chain Simulation, DDM will involve deciding how the output of one model can be used as input to other models. GRIDS (Generic Runtime Infrastructure for Distributed Simulation), an extensible, service-based RTI developed to investigate interoperability issues in Distributed Simulation will be used to effect interoperability among the supply chain simulations. The GRIDS extensible service architecture is realized by thin agents. These agents may be used to support the simulation by providing tasks such as optimizations and assistance. This paper discusses how data distribution management can be incorporated into the Thin Agents to assist in the connection of the models in a Distributed Supply Chain simulation.