Communication-Efficient Support for Spatial Filtering of State Updates in Distributed Virtual Environments

Spatial filtering is an interest management mechanism that eliminates unnecessary state updates transmitted in distributed virtual environments. This paper proposes a new protocol to reduce the communication overhead for supporting spatial filtering. The protocol maintains potential interest regions (PIRs) to assist in observing entities entering the areas-of-interest (AOIs) of other entities. PIRs enable the protocol to take advantage of the relative positions between entities and AOIs to save communication cost. We elaborate the design of the protocol. A number of concurrency issues that arise from the delay of message transmission are identified and resolved. We also propose an algorithm for computing PIRs to maximize the saving in communication cost. Experimental results show that the proposed protocol maintains highly accurate entity-in-AOI knowledge with substantially lower communication overhead compared to existing strategies.