Abstract
Current technologies for creating distributed simulations or virtual environments are too limited in terms of scalability and flexibility, particularly in the areas of network saturation, distribution of VR scenes, and co-ordination of large systems of active objects. This thesis proposes the use of mobile and distributed computing technology to alleviate some of these limitations. A study of contemporary technologies for distributed simulation and networked virtual environments has been made, examining the benefits and drawbacks of different techniques. The main theory that has been investigated is that processing of a global simulation space should be spread over a network of computers, the principle of locality cutting the network bandwidth required. Using a prototype language for distributed graph processing, which fully supports mobile programming, experimental systems have been developed to demonstrate the use of distributed processing in creating large-scale virtual environments. The working examples created show that the ideas proposed for distribution of interactive virtual environments are valid, and that mobile programming techniques provide a new direction of development for the field of simulation. A more detailed summary of the work is given in Appendix D. Five publications to date (shown overleaf) have resulted from my involvement in the work, and a number of others have resulted from the overall project.