Although there is no consensus on an exact definition yet, the metaverse is generally understood as an open virtual world in which users can interact, contribute content, and engage in monetized transactions (both over user-created content and out-of-metaverse services and goods). Interactions are usually understood as taking place in a three-dimensional world (possibly experienced through a virtual reality device), but earlier online experiments (including textual MMORPGs of the 80s) have shown that other modes of interactions are possible.
As these virtual experiences progressively combine gaming elements, monetary transactions, and monetized content (either provided by the game’s developer but also increasingly by other users and actors acting as content creators), the reliability of such platforms is becoming increasingly important.
Unfortunately, the robustness and fault tolerance that ensures the consistency, replication, and persistence of a metaverse’s state over multiple machines have been little studied. Distributed 3D and VR games today are typically built on industrial engines (two leading examples are the Unity and Unreal engines, which power some of the most visible games in the industry). These engines typically embark some distributed logic (known as a netcode) to distribute a game’s state across multiple machines. Netcodes either follow a client-server (the most common) or P2P logic and rely on heuristics and optimization (such as delays and rollbacks) to ensure an acceptable level of quality of experience (e.g. high reactivity and low latency), even in a degraded network.
In the past, weaknesses in the security of netcodes have been exploited to obtain an unfair advantage over other players, using a variety of attacks. The problem, however, has been little studied by the research community and is poised to grow in importance with the growth of decentralized metaverse applications beyond simple gaming. In particular, we anticipate that metaverse operators will have to reconcile the need to replicate content (for reliability, availability, and performance reasons), possibly reusing earlier research on P2P games, and the need to ensure some consistency between the copies of the world-state in spite of potentially rogue participants. Such a context directly chimes in with research Byzantine-Fault-Tolerant (BFT) algorithms that are able to withstand the corruption of a proportion of the devices they run on. BFT replication protocols are however not usually designed with the needs of a metaverse world in mind. In particular, BFT replication protocols are usually very costly, do not scale well, and induce high delays. These limits directly conflict with the soft real-time constraints put on on metaverse virtual world to support seamless and high-quality user interactions.
To address these challenges, this PhD aims to develop novel Byzantine-tolerant replication mechanisms that espouse the unique requirements of a metaverse’s virtual world, notably in terms of responsiveness and latency. In particular, the PhD will explore how Byzantine-Fault-Tolerant techniques can be accelerated using two central strategies: (i) by using adaptive levels of protection for different parts of a metaverse world, and (ii) by allowing for optimistic local decisions coupled with micro-rollbacks in case malicious discrepancies are later discovered.
[1] E. Williams, B. Eyo, and S. Akpan. "Linden Labs Second Life: Understanding the Business Model and Sources of Commercial and Social Success or Decline of Second Life". In: Computer and Information Science 4.2 (2011), p. 96.
[2] B. Francis. "Epic and Time Magazine debut interactive MLK Jr. exhibit in Fortnite". In: Game Developer (Aug. 2021). Accessed January 23, 2022.
[3] A. Lioret, L. Diler, S. Dalil, and M. Mota. "Hybrid Prediction for Games' Rollback Netcode". In: SIGGRAPH Posters. ACM, 2022, 3:1-3:2.
[4] M. Ahmed, S. Reno, M. R. Rahman, and S. H. Rifat. "Analysis of Netcode, Latency, and Packet- loss in Online Multiplayer Games". In: 2022 International Conference on Augmented Intelligence and Sustainable Systems (ICAISS). 2022, pp. 1198-1202.
[5] T. Walker, B. Gilhuly, A. Sadeghi, M. Delbosc, and S. L. Smith. "Predictive Dead Reckoning for Online Peer-to-Peer Games". In: IEEE Transactions on Games (2023).
[6] K.-T. Chen, Y.-C. Chang, H.-J. Hsu, D.-Y. Chen, C.-Y. Huang, and C.-H. Hsu. "On the quality of service of cloud gaming systems". In: IEEE Transactions on Multimedia 16.2 (2013), pp. 480-495.
[7] P. Quax, P. Monsieurs, W. Lamotte, D. De Vleeschauwer, and N. Degrande. "Objective and subjective evaluation of the influence of small amounts of delay and jitter on a recent first person shooter game". In: Proceedings of 3rd ACM SIGCOMM workshop on Network and system support for games. 2004, pp. 152-156.
[8] A. Yahyavi, K. Huguenin, J. Gascon-Samson, J. Kienzle, and B. Kemme. "Watchmen: Scalable Cheat-Resistant Support for Distributed Multi-player Online Games". In: IEEE 33rd International Conference on Distributed Computing Systems, ICDCS 2013, 8-11 July, 2013, Philadelphia, Pennsylvania, USA. IEEE Computer Society, 2013, pp. 134-144.
[9] R. Diaconu, J. Keller, and M. Valero. "Manycraft: Scaling Minecraft to Millions". In: Annual Workshop on Network and Systems Support for Games, NetGames '13, Denver, CO, USA, December 9-10, 2013. IEEE/ACM, 2013, 1:1-1:6.
[10] J. Gascon-Samson, B. Kemme, and J. Kienzle. "Lamoth: A Message Dissemination Middleware for MMOGs in the Cloud". In: Annual Workshop on Network and Systems Support for Games, NetGames '13, Denver, CO, USA, December 9-10, 2013. IEEE/ACM, 2013, 9:1-9:2.
[11] L. Lamport, R. E. Shostak, and M. C. Pease. "The Byzantine Generals Problem". In: ACM Transactions on Programming Languages and Systems 4.3 (1982), pp. 382-401.
[12] A. Miller, Y. Xia, K. Croman, E. Shi, and D. Song. "The Honey Badger of BFT Protocols". In: Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security. CCS '16. Vienna, Austria: ACM, 2016, pp. 31-42. isbn: 978-1-4503-4139-4.
[13] I. Abraham, D. Malkhi, K. Nayak, L. Ren, and M. Yin. "Sync HotStuff: Simple and Practical Synchronous State Machine Replication". In: IEEE Symposium on Security and Privacy (S&P). 2020, pp. 106-118.
[14] T. Albouy, D. Frey, M. Raynal, and F. Taïani. "Good-Case Early-Stopping Latency of Synchronous Byzantine Reliable Broadcast: The Deterministic Case". In: DISC. Vol. 246. LIPIcs. Schloss Dagstuhl - Leibniz-Zentrum fu ̈r Informatik, 2022, 4:1-4:22.
[15] D. Frey, L. Guillou, M. Raynal, and F. Taïani. "Consensus-Free Ledgers When Operations of Distinct Processes are Commutative". In: PaCT. Vol. 12942. Lecture Notes in Computer Science. Springer, 2021, pp. 359-370.
[16] T. Albouy, D. Frey, M. Raynal, and F. Taïani. "Byzantine-Tolerant Reliable Broadcast in the Presence of Silent Churn". In: SSS. Vol. 13046. Lecture Notes in Computer Science. Springer, 2021, pp. 21-33.
[17] A. Auvolat, D.Frey, M. Raynal, and F. Taïani."Byzantine-tolerant causal broadcast". In: Theor. Comput. Sci. 885 (2021), pp. 55-68.