Improving collision detection in distributed virtual environments by adaptive collision prediction tracking

Collision detection for dynamic objects in distributed virtual environments is still an open research topic. The problems of network latency and available network bandwidth prevent exact common solutions. The consistency-throughput tradeoff states that a distributed virtual environment cannot be consistent and highly dynamic at the same time. Remote object visualization is used to extrapolate and predict the movement of remote objects reducing the bandwidth required for good approximations of the remote objects. Few update messages aggravate the effect of network latency for collision detection. In this paper, new approach extending remote object visualization techniques is demonstrated to improve the results of collision detection in distributed virtual environments. We showed how this can significantly reduce the approximation errors caused by remote object visualization techniques. This is done by predicting collisions between remote objects and adaptively changing the parameters of these techniques.