Object caching and prefetching in distributed virtual walkthrough

We investigate a virtual reality application where a user experiences a walkthrough of a large virtual environment in the Internet, modeling a physical museum, library or any place of interest. Such technology provides users with the ability to view and explore a remote environment, without physical presence. To deliver good performance for such applications, we need to address several issues in different research disciplines. First, we must be able to model virtual objects effectively. The recently developed techniques for multi-resolution object modeling in computer graphics are useful in simplifying object models and reducing their rendering time. Second, with the low bandwidth of the Internet, caching suitable object models of high affinity will reduce the amount of data requested over the network for faster response. Prefetching object models by predicting those likely to be used in the near future and downloading them in advance will lead to a similar improvement. Third, the Internet often suffers from disconnection. A caching mechanism allowing objects to be cached with at least their minimum resolutions can provide at least a coarse view of the objects for an improved visual representation. In this paper, we propose a multi-resolution caching mechanism and investigate its effectiveness in supporting virtual walkthrough applications in the Internet environment. The caching mechanism is further complemented with several object prefetching mechanisms for predicting future accessed objects. The performance and feasibility of our proposed mechanisms are quantified via simulation experiments.