Accommodating performance limitations in distributed virtual reality systems

  • Authors:
  • J.R Ensor;G.U Carraro;J.T Edmark

  • Affiliations:
  • Bell Laboratories, 101 Crawfords Corner Road #4F607 Holmdel, NJ 07733, USA;Bell Laboratories, 101 Crawfords Corner Road #4F607 Holmdel, NJ 07733, USA;Bell Laboratories, 101 Crawfords Corner Road #4F607 Holmdel, NJ 07733, USA

  • Venue:
  • Computer Communications
  • Year:
  • 2000

Quantified Score

Hi-index 0.24

Visualization

Abstract

The paper briefly describes a sports simulator called Peloton. Simulation participants exercise or compete within virtual reality environments that represent the roadways along which they appear to bicycle, run, or walk. The visual display of each racecourse is a composition of multiple three-dimensional sub-spaces, which are called regions. Some regions of a virtual world are made up of graphical objects, and other regions are made up of still images or video streams. The paper also describes ways of dealing with limited computation and communication resources in distributed virtual reality systems. The discussion focuses on systems that represent their virtual worlds as multiple regions. It introduces means of altering the visual representations of three-dimensional spaces according to the computing and communication resources available during a simulation. The size, placement and graphical complexity of the regions of a virtual world may be tailored according to static as well as dynamic configurations of system components. Finally, the paper presents ways in which a given view of a virtual world-a scene-can be used for extended periods during a simulation. This technique is useful when scene updates are delayed or missing.