DISCLAIMER
Abstract
—Massively multiplayer online game servers are challenging to build and maintain; they require always-on availability,low-latency, and high predictability. In many ways, these systemscould benefit from pushing application logic to the connectedclients. However, historically peer-to-peer networks have notworked when applied to this domain. Pushing game logic tothe clients complicates development since client end hosts cannottrust their peers and they have unreliably persistent connections,increasing the possibility of cheating or network failure.Google’s Edge Network changes everything we have concludedabout peer-to-peer networks over the past decade. Having accessto thousands of servers all over the world solves problems ofavailability, data access, link saturation, security, and control.This new system allows for the inclusion of trusted peers inotherwise untrusted node clusters. Developers can explore peer-to-peer or decentralized algorithms while allowing completecontrol over their data, code, and players, in their developedvirtual worlds. They can offload application game logic to a low-latency scalable system, which provides a cost-effective solutionresiding closer to the connected clients.In this paper, we investigate the gains game developersmay obtain by exploring emerging edge cloud technology. Wedemonstrate how massively multiplayer online games benefitfrom this new system though simulation. We compare area-of-interest latency, which is the latency between client interactionswithin the virtual world for known World of Warcraft andGoogle data centers locations. We show the benefits to area-of-interest latency gained from using Google’s Edge Network,while minimally affecting client to server latency. Finally, wepresent a novel approach to maximize latency reductions forarea-of-interest latency by moving players to optimal peeringedge servers, thus reducing distance between clients.
Conclusion
We have shown improvements to latency between players ina virtual world using Google’s Edge Network. We presentedthe traditional model used currently, a model that connectsclients to their closest peering edge servers, and an optimizedsolution where clients move to different peering edge serversto reduce latency between players in the virtual world. Ofthese three, we conclude that clients connecting to their nearestpeering edge server results in the simplest implementationwhile maintaining low latency from the central server andother clients. The non-optimized edge network solution alsoprovides a simple path to numerous suggestions we havepresented for game architecture consideration.The optimized solution presents potential when client toserver latency is less important and we isolate game stateinformation to individual peering edge servers for extendedportions of gameplay. Although this system appears to be morecomplicated, game development decisions may vastly simplifythose complications further increasing the benefits gained fromthis system.We gain the benefits of peer-to-peer network topologieswhen using Google’s Edge Network, while maintaining thesecurity and control required by modern game developers.This work only focuses on the latency requirements due tothe importance of this metric [5] [6]. Future work needsto strengthen this model by creating the message systemneeded to allow gameplay. We assume that Google’s systemhandles data synchronization between peering edge serversand the central server [4], but further work could optimizethis synchronization and utilize the clients as members of thisnetwork topology.Overall, our improved results excite us to the possibilitiesthat peering edge servers provide for future work. The latencyreductions found within our work show that peering edgeservers have the possibility to significantly improve perfor-mance of massively multiplayer online games.
Abstract
—Massively multiplayer online game servers are challenging to build and maintain; they require always-on availability,low-latency, and high predictability. In many ways, these systemscould benefit from pushing application logic to the connectedclients. However, historically peer-to-peer networks have notworked when applied to this domain. Pushing game logic tothe clients complicates development since client end hosts cannottrust their peers and they have unreliably persistent connections,increasing the possibility of cheating or network failure.Google’s Edge Network changes everything we have concludedabout peer-to-peer networks over the past decade. Having accessto thousands of servers all over the world solves problems ofavailability, data access, link saturation, security, and control.This new system allows for the inclusion of trusted peers inotherwise untrusted node clusters. Developers can explore peer-to-peer or decentralized algorithms while allowing completecontrol over their data, code, and players, in their developedvirtual worlds. They can offload application game logic to a low-latency scalable system, which provides a cost-effective solutionresiding closer to the connected clients.In this paper, we investigate the gains game developersmay obtain by exploring emerging edge cloud technology. Wedemonstrate how massively multiplayer online games benefitfrom this new system though simulation. We compare area-of-interest latency, which is the latency between client interactionswithin the virtual world for known World of Warcraft andGoogle data centers locations. We show the benefits to area-of-interest latency gained from using Google’s Edge Network,while minimally affecting client to server latency. Finally, wepresent a novel approach to maximize latency reductions forarea-of-interest latency by moving players to optimal peeringedge servers, thus reducing distance between clients.
Conclusion
We have shown improvements to latency between players ina virtual world using Google’s Edge Network. We presentedthe traditional model used currently, a model that connectsclients to their closest peering edge servers, and an optimizedsolution where clients move to different peering edge serversto reduce latency between players in the virtual world. Ofthese three, we conclude that clients connecting to their nearestpeering edge server results in the simplest implementationwhile maintaining low latency from the central server andother clients. The non-optimized edge network solution alsoprovides a simple path to numerous suggestions we havepresented for game architecture consideration.The optimized solution presents potential when client toserver latency is less important and we isolate game stateinformation to individual peering edge servers for extendedportions of gameplay. Although this system appears to be morecomplicated, game development decisions may vastly simplifythose complications further increasing the benefits gained fromthis system.We gain the benefits of peer-to-peer network topologieswhen using Google’s Edge Network, while maintaining thesecurity and control required by modern game developers.This work only focuses on the latency requirements due tothe importance of this metric [5] [6]. Future work needsto strengthen this model by creating the message systemneeded to allow gameplay. We assume that Google’s systemhandles data synchronization between peering edge serversand the central server [4], but further work could optimizethis synchronization and utilize the clients as members of thisnetwork topology.Overall, our improved results excite us to the possibilitiesthat peering edge servers provide for future work. The latencyreductions found within our work show that peering edgeservers have the possibility to significantly improve perfor-mance of massively multiplayer online games.
Last edited: