2009 Poster Sessions : Scalable Data Distribution in Virtual Worlds

Student Name : Tahir Azim
Advisor : Philip Levis
Research Areas: Computer Systems
Today's virtual worlds are characterized by centralized control and limited by the level of scalability and flexibility they can support. Our research focuses on improving the scalability of virtual worlds by controlling the bandwidth demands of their underlying distributed systems. The inherent geometric nature of a virtual world allows the system to concentrate its network resources on data pertaining to objects that are nearby. To exploit this property, we divide a virtual world into smaller contiguous sub-regions, called spaces, each of which is managed by a separate “space server”. We then employ a novel constant bit-rate approach, which upper bounds the outgoing data rate used by a space server, while giving priority to data destined for nearby spaces. We contend that this approach can make distributed systems for virtual worlds operate more efficiently and better able to handle millions of users.

Tahir Azim is a first-year PhD student at Stanford University's Computer Science Department. His research interests include routing in low-power wireless sensor networks and scalable networked systems for virtual worlds. Before starting his PhD, he received his MS degree in Computer Science from Stanford University and B.E. degree in software engineering from National University of Sciences and Technology, Pakistan.