Enforcing Game Rules in Untrusted P2P-based MMVEs

Cheating is a major obstacle on the way to realize MMVEs based on peer-to-peer networks of untrusted nodes. Current approaches either rely on trusted nodes or use a pessimistic distributed simulation, with negative eect on the realtime responsiveness of the virtual world. We propose to use an optimistic simulation maintaining the realtime immersion of players. We show how dierent kinds of cheats can be prevented or detected using this approach based on a broadcast network. Most importantly, tampering with the world state by malicious players is prevented. Finally, we show how to make this approach scalable by using a representative subset of nodes to decide on the validity of player actions.