Institutions | About Us | Help | Gaeilge
rian logo

Go Back
Dynamic Hybrid Strategy Models for Networked Mulitplayer Games
McCoy, Aaron; McLoone, Seamus; Ward, Tomas E.; Delaney, Declan
Two of the primary factors in the development of networked multiplayer computer games are network latency and network bandwidth. Reducing the effects of network latency helps maintain game-state fidelity, while reducing network bandwidth usage increases the scalability of the game to support more players. The current technique to address these issues is to have each player locally simulate remote objects (e.g. other players). This is known as dead reckoning. Provided the local simulations are accurate to within a given tolerance, dead reckoning reduces the amount of information required to be transmitted between players. This paper presents an extension to the recently proposed Hybrid Strategy Model (HSM) technique, known as the Dynamic Hybrid Strategy Model (DHSM). By dynamically switching between models of user behaviour, the DHSM attempts to improve the prediction capability of the local simulations, allowing them to stay within a given tolerance for a longer amount of time. This can lead to further reductions in the amount of information required to be transmitted. Presented results for the case of a simple first-person shooter (FPS) game demonstrate the validity of the DHSM approach over dead reckoning, leading to a reduction in the number of state update packets sent and indicating significant potential for network traffic reduction in various multiplayer games/simulations.
Keyword(s): Electronic Engineering; Distributed Interactive Applications; Hybrid Strategy Model; Networked Multiplayer Computer Games; Dead Reckoning; Packet-Reduction; User-Modelling
Publication Date:
Type: Conference item
Peer-Reviewed: Yes
Institution: Maynooth University
Citation(s): McCoy, Aaron and McLoone, Seamus and Ward, Tomas E. and Delaney, Declan (2005) Dynamic Hybrid Strategy Models for Networked Mulitplayer Games. In: 19th European Conference on Modelling and Simulation (ECMS 2005), 1-4th June 2005, Riga, Latvia.
File Format(s): application/pdf
Related Link(s):
First Indexed: 2020-01-31 06:27:02 Last Updated: 2020-04-02 07:47:35