Dynamic grid-based vs. region-based data distribution management strategies for large-scale distributed systems

In large-scale distributed simulation systems, the primary goals of the data distribution management are to manage the distribution of state updates and interaction information, as well as to limit and control the volume of data exchanged during a simulation. We focus upon two DDM implementations: dynamic grid-based and region-based. Using these two DDM methods, we run numerous simulations to compare the time performance, message overhead and multicast group usage of each method. On a cluster of workstations, with a mini-RTI kit, we simulate several ground-based radar entities, which attempt to detect simulated squadron planes. Our goal is to determine which of the two DDM schemes is more efficient for our ground-based radar vs. squadron planes scenario. The paper presents background information on HLA/RTI and the two DDM schemes we compare, explains the scenario, metrics and parameters we use in our experiments, presents the analysis of our results and our conclusions, based on those results.