Coping with hotspots: AOI adaption strategies for P2P networked virtual environments

Massively Multiplayer Online Games and Virtual Worlds are among the most popular applications on the Internet. As player numbers increase, the limits of the currently dominant client/server architecture are becoming obvious. To overcome those limits, the research community has developed protocols for these so-called Distributed Networked Virtual Environments (DVEs) based on per-to-peer technologies. One problem that has to be solved for practical peer-to-peer protocols for DVEs are hotspots: participants tend to form groups in the virtual environment. If these groups become very large, players have to send event messages to a large number of other players. This imposes a big load on the peer-to-peer protocol. To alleviate the problem, adaptive Area of Interest (AoI) sizes can be used: If the player density becomes too high, the players' AoI size will be decreased. In this paper we identify different adaption strategies and, using simulations, evaluate the influence of the respective strategies on the peer-to-peer protocol. As a result, we show that distributed strategies generally yield a better performance than purely local strategies.