5GTN performance measurements in Metaverse use case

Abstract

The Metaverse is rapidly evolving into a sophisticated, dynamic world that requires scalable infrastructure, low-latency communication, and consistent network reliability. This thesis investigates the potential of 5G technology to facilitate synchronized and scalable Metaverse applications through two experimental models. The Scalability Model assesses traffic behavior between two 5G-connected servers under varying user densities and user spawning intervals. The study demonstrates that augmented spawn rates and user concurrency result in non-linear escalation of uplink and downlink traffic by replicating authentic user interactions using scripted non-player character (NPC) behavior. The comparison of real-time performance between 5GTN and Wi-Fi in regulated VR-based interactions is advanced by the second model, the Advanced Avatars framework. The results highlight significant performance disparities, while Wi-Fi has delays due to contention, irregular avatar updates, and greater variability in traffic distribution, particularly during motion-intensive applications, 5GTN ensures stable traffic profiles with reduced jitter and enhanced synchronization. This model demonstrates that, even with robust backend infrastructure, user experience is significantly affected by the choice of access network.

Both models provide a comprehensive perspective on 5G performance in immersive environments, encompassing server-to-server communication and user-access-level analysis collectively. The results emphasize the impact of traffic distribution, access network characteristics, and spawning logic on Metaverse scalability. The performance evaluations carried out on both Metaverse research models, scalability testing and VR-controlled advanced avatars provide essential insights into system behavior under different user loads and network conditions. These findings enhance comprehension of the infrastructural and network requirements for immersive virtual environments. Thus, the findings provide a significant reference for researchers and developers in the design and optimization of future Metaverse platforms.