Resource-Cost Analysis for Massively Multiplayer Online Games

In this paper, an approach for modelling and analysis of massively multiplayer online games (MMOGs) is proposed. By using our models, based on Markov process and queuing theory, we can investigate the impact of various factors, such as game types, the number of players, the intensity of players or NCPs interaction, the number of regions in a game world, avatar region-transition rate, and network configuration, on the system performance. A number of resource-cost functions are defined, which map the system resource consumed to the cost during a game play. We choose the federated peer-to-peer architecture as a target for this study, since it is a typical and potential architecture proposed recently and attracting more attention.