Visibility Path-finding in relation to Hybrid Strategy-based Models in Distributed Interactive Applications
Madden, D and Delaney, Declan and McLoone, Seamus and Ward, Tomas (2004) Visibility Path-finding in relation to Hybrid Strategy-based Models in Distributed Interactive Applications. In: 8th IEEE International Symposium on Distributed Simulation and Real Time Applications, October 2004, Budapest, Hungary.
The hybrid strategy-based modeling approach is a method for reducing the number of network packets that need to be transmitted to maintain global consistency in Distributed Interactive Applications. It combines a short-term model such as dead reckoning with a long-term strategy model. A key aspect of this approach is to determine strategies that users adopt in navigating the simulated environment to satisfy some objective or goal. Computer-generated artificial entities called BOTS, navigate by employing an Artificial Intelligence technique called path finding. This paper proposes using the A* path finding algorithm to automatically compute strategies that human users might take through the simulated environment. Since the A* algorithm operates on a graph representation of the environment and because of the real-time constraints imposed on Distributed Interactive Applications, the paper also carries out a comparative analysis of two extreme graph representations of the environment - a standard regular grid and a minimal grid representation. The comparison shows that the minimal grid leads to an order of magnitude reduction in real-time computation compared to the regular grid. In addition the paths computed using the minimal grid and the A* algorithm are used to determine strategy models as part of the hybrid strategy-based modeling approach. It is shown that this reduces the network traffic required to maintain global consistency of entity dynamics in two simulated environments.
Repository Staff Only: item control page