Scalable collision detection for massively multiplayer online games

We describe approaches for satisfying the real-time collision detection requirements of distributed virtual environments. We assume a distributed virtual environment is deployed using client/server architecture typical of commercial massive multiplayer online games. We exploit the scalability provided by the clustering of servers in the development of a real-time collision detection service that may scale to satisfy the requirements of virtual environments that are constructed of many thousands of objects. We present performance figures that show our approaches to be scalable in that an addition of servers to a cluster results in an increased number of objects that can be considered for collision detection in real-time.