NGS: An application layer network game simulator

Abstract

In the last five years the popularity of Massively Multiplayer Online Games (MMOGs) has exploded. Unfortunately, the demand has far outweighed the resources developers can provide. Many MMOGs are suffering from scalability issues, resulting in sharding, down time, and server crashes. To solve these problems, the research community is investigating peer-to-peer (P2P) overlay networks to support MMOGs, as P2P networks are theoretically and practically scalable. The majority of analysis of P2P gaming architectures has been qualitative, making it difficult to understand the strengths and weaknesses of each system. This is partially due to the lack of appropriate simulation tools. To address this problem we have developed an application layer network game simulator - NGS - for modelling network game architectures. NGS includes mechanisms to collect quantitative metrics, which may then be used to perform comparisons with other architectures. NGS is flexible enough to model Client/Server, Region based, Neighbour based, and hybrid architectures. It is extensible and modular, and will enable the research community to evaluate the benefits and weaknesses of existing and new network gaming architectures. Results demonstrating the extensibility and performance of NGS, and comparisons of the performance of several different architectures are included.