Source-linked AI summary
6G-enabled Edge AI for Metaverse: Challenges, Methods, and Future Research Directions
Luyi Chang, Zhe Zhang, Pei Li, Shan Xi, Wei Guo, Yukang Shen, Zehui Xiong, Jiawen Kang, Dusit Niyato, Xiuquan Qiao, Yi Wu
TL;DR
The survey addresses Metaverse demands for high bandwidth, low latency, reliability, intelligence, and large-scale data processing. It synthesizes 6G-enabled edge AI architectures, challenges, existing methods, and future research directions for integrating virtual and real worlds.
Problem
The Metaverse requires edge AI architectures that can support high bandwidth, low latency, reliability, intelligence, and massive data processing for immersive virtual–real interaction.
Method
The survey defines and analyzes 6G-enabled edge AI and Metaverse, investigates three edge-Metaverse architectures and their solutions, and reviews challenges, methods, and future research directions.
Results
The survey investigates three 6G edge-intelligence Metaverse systems, presents methods for core challenges, and identifies open topics and future research directions.
Takeaways & Limitations
The survey provides researchers and practitioners with insights and guidance for expanding Metaverse research through 6G-enabled edge intelligence.
Takeaways & Limitations
The survey notes that some discussed methods suffer from insufficient trained-model performance.
Abstract
from arXiv · showhide
6G-enabled edge intelligence opens up a new era of Internet of Everything and makes it possible to interconnect people-devices-cloud anytime, anywhere. More and more next-generation wireless network smart service applications are changing our way of life and improving our quality of life. As the hottest new form of next-generation Internet applications, Metaverse is striving to connect billions of users and create a shared world where virtual and reality merge. However, limited by resources, computing power, and sensory devices, Metaverse is still far from realizing its full vision of immersion, materialization, and interoperability. To this end, this survey aims to realize this vision through the organic integration of 6G-enabled edge AI and Metaverse. Specifically, we first introduce three new types of edge-Metaverse architectures that use 6G-enabled edge AI to solve resource and computing constraints in Metaverse. Then we summarize technical challenges that these architectures face in Metaverse and the existing solutions. Furthermore, we explore how the edge-Metaverse architecture technology helps Metaverse to interact and share digital data. Finally, we discuss future research directions to realize the true vision of Metaverse with 6G-enabled edge AI.
1 INTRODUCTION
The survey frames 6G-enabled edge AI as a response to latency, stability, and security constraints in edge AI, motivated by the Metaverse’s demanding requirements for bandwidth, latency, reliability, and intelligence. It surveys architectures, challenges, methods, and future research directions connecting these technologies.
- Edge AI enables real-time data processing and model inference on devices, but its applications still require optimization.
- High latency, fragile stability, and low security constrain edge AI deployment across remote devices and cloud-connected networks.Some haptic AI applications require at least 1 Mbit/s transfer rates and no more than 1 ms latency.
- 6G-enabled edge AI integrates AI with 6G communication to provide lower latency, more stable connectivity, and a more secure network architecture.The passage describes a THz-band peak rate of 1Tb/s and network delay below 1 ms.
- The Metaverse demands ultra-high bandwidth, ultra-low latency, ultra-high reliability, and intelligent services because it merges complex virtual and physical infrastructures.
- The survey distinguishes its focus from prior work by studying Edge Cloud-Metaverse, Mobile Edge Cloud-Metaverse, and Decentralized-Metaverse architectures.It also summarizes related challenges and future research directions for 6G-enabled edge AI in the Metaverse.
- The paper contributes definitions and a tutorial, architecture and solution analysis, challenge synthesis, and future research directions for 6G-enabled edge AI in the Metaverse.Its organization covers 6G edge intelligence, Metaverse concepts, architectures and methods, future questions, and conclusions.
2 6G & EDGE INTELLIGENCE
The section presents 6G as an integrated communication and computing technology whose performance and infrastructure support edge intelligence. It describes edge-intelligence benefits and centralized, decentralized, and hybrid model-training architectures.
- 6G: 6G integrates sensing, storage, communication, control, and computing, with high performance, global coverage, real-time processing, reliability, and energy efficiency.
- 6G: Compared with 5G, 6G is described as improving peak transmission rate, reliability, traffic density, positioning accuracy, and connection density by more than 100 times.
- 6G applications and infrastructure: 6G supports new media, services, and infrastructure through near-zero-latency sensory interconnection, precise service models, and integrated ground, UAV, and satellite networks.
- Edge intelligence: Edge AI deploys machine-learning algorithms on edge nodes, while increasing edge-node computing power supports future 6G applications.
- Edge intelligence benefits: 6G-oriented edge intelligence is presented as providing low latency, computing offload, and high performance, with balanced storage, efficient transmission, and high reliability.The described reliability mechanism uses spatial multiplexing with communication-interruption probability below one in a million.
- Training architectures: Centralized architecture assigns data collection, model training, and inference to the cloud server after edge devices upload training data.
- Training architectures: Decentralized architecture has edge nodes train models locally and exchange updates to obtain a shared global model, whereas hybrid architecture combines edge-server coordination with local training and cloud training.
3.1 Metaverse Features
The Metaverse is a persistent, social, interoperable virtual space that interacts with the physical world through avatars, immersive technologies, and digital assets. Its technical foundation supports high-fidelity interaction and multiple connected application scenarios.
- The Metaverse is an online shared virtual-reality world where customized 3D avatars support activities such as shopping, telecommuting, and video conferencing.
- Its physical-virtual fusion feeds avatar stimuli back to users through XR and intelligent wearables, enabling higher-fidelity interaction beyond screens and mobile devices.
- Digital twins, XR, blockchain, computer vision, and related technologies support physical-entity mapping, immersion, and the Metaverse’s economic system.
- Interoperability allows services and virtual goods purchased in one Metaverse scenario to be used across multiple scenarios.
- Sociality is essential because the Metaverse supports higher-level social behaviors beyond the physical world, including virtual office and entertainment.
- The Metaverse has longevity because avatars, behaviors, personal information, and digital assets can remain stored even after real-world organizations or individuals disappear.
3.2 Metaverse Architectures
Metaverse architectures combine physical, virtual, and technical layers to support real-time interaction, immersive services, and data exchange between users and digital worlds.
- The Metaverse architecture comprises physical, virtual, and technical layers that support real-time interaction between physical and virtual worlds.
- Users access virtual applications through VR, HMD, AR, and gesture-sensing devices, while avatars enable customized identities and social interaction.
- Smart wearables upload real-world data and return avatar feedback, serving as an interaction medium between physical and virtual worlds.
- Physical service providers maintain communication and computing resources, including resource allocation to reduce physical–virtual interaction delay.
- The virtual layer contains environments, services, currencies, and avatar behaviors, supported by high-definition rendering and simulation.
- Technical Layer: Digital twins map physical entities into real-time virtual models, while blockchain provides distributed, tamper-proof storage that mitigates resource and privacy problems.
3.3 Metaverse Applications
Metaverse applications span education, telecommuting, smart cities, product testing, and industrial production by combining immersive virtual environments with real-world modeling and simulation.
- Metaverse applications include entertainment, games, education, smart cities, healthcare, industrial simulation, and digital tourism.
- Immersive Education: Digital twins support immersive education by placing learning content in virtual scenarios, including language-learning platforms with characters and conversation settings.
- Telecommuting: Metaverse telecommuting creates virtual offices and conference rooms, while spatial audio addresses monotony and limited spatiality in videoconferencing.
- Smart City: Digital twins map roads, buildings, and vehicles into virtual cities, enabling simulation experiments for facility planning and resource allocation.
- Product Testing: Products can be tested simultaneously in physical and virtual spaces, improving the efficiency of testing and certification while reducing costs.
- Industrial Applications: Digital twins model factory structures, production lines, and processes so workers can adjust capacity, equipment, and staffing in virtual environments.
3.4 Metaverse History
The Metaverse evolved from a fictional always-online virtual world toward increasingly immersive and interactive systems through Internet, digital-twin, VR, and decentralization advances.
- The term “Metaverse” originated in Neal Stephenson’s 1992 novel Snow Crash, which described a persistent virtual world paralleling reality.
- Internet expansion and digital-twin development made Metaverse platforms feasible by 2003, exemplified by Linden Laboratory’s Second Life.
- VR development in 2014 shifted the Metaverse from planar, passive, one-way interaction toward three-dimensional, active, interactive experiences.
- Since 2016, VR devices such as Oculus and HTC Vive have enhanced immersion by translating avatar-generated visual, auditory, and tactile stimuli into physical experience.
4.1 Edge Intelligence-Based Metaverse Architectures
Edge intelligence architectures address Metaverse bandwidth, latency, scalability, privacy, and model-performance constraints through cloud-edge collaboration, federated learning, MEC, and blockchain.
- Edge Cloud-Metaverse Architecture: Server-centric Metaverse networks face variable cloud-terminal bandwidth, video frame drops, and high latency, motivating cloud-edge architectures.
- Edge Cloud-Metaverse Architecture: Self-balancing federated learning assigns users to edge nodes according to data distributions or AI-task types to address statistical heterogeneity.
- Mobile Edge Cloud-Metaverse Architecture: MEC places cloud services near mobile users and can combine multiple edge nodes to reduce delay caused by user mobility.
- Mobile Edge Cloud-Metaverse Architecture: The FL-based MEC architecture transmits models rather than user data between edge nodes and uses a DAG to improve computational efficiency.
- Scalability: Growing online-user populations increase infrastructure computing workloads, creating scalability demands for cloud servers and edge nodes.
- Decentralized Metaverse Architecture: The proposed decentralized architecture addresses insufficient model performance caused by privacy restrictions in blockchain-based Metaverse systems.
- Decentralized Metaverse Architecture: Blockchain-based federated learning lets users with similar data distributions contribute local models, average them on-chain, and record transactions for traceability.
4.2 Challenges and Advanced Methods
The survey examines privacy, latency, and resource-allocation challenges in the Metaverse, together with advanced methods addressing them. These methods include privacy mechanisms, latency-reduction incentives, blockchain-based allocation, and decentralized edge learning.
- Privacy: Metaverse privacy challenges include risks to users’ data, locations, activities, and goals in an open virtual environment.
- Privacy: “Clone cloud” uses moving user substitutes to confuse private information, while “private copy” confines experiences to virtual private storage.
- Delay: Insufficient bandwidth and network latency can cause visual jitter and delays that undermine immersive Metaverse experiences.
- Delay: A DRL-based double Dutch auction uses perceived quality to evaluate immersion while improving communication efficiency and reducing auction cost.
- Resource Allocation: Resource allocation methods include Metachain sharding with Stackelberg incentives and decentralized edge learning using cooperative offloading and multiagent DRL.
5 OPEN RESEARCH TOPICS AND FUTURE DIRECTIONS
The future directions address perceptual realization, ethical complexity, identity privacy, and the balance between virtual and real life. Proposed research must improve immersion while managing precision, security, social behavior, and overindulgence.
- Perceptual Realization: BCI could augment wearable devices by enabling brain-device communication, but future systems must balance invasive precision against non-invasive convenience.
- Ethical Issues: Metaverse avatars create ethical challenges because unconstrained virtual behavior can produce social relationships more complex than those in the real world.
- Identity Privacy: Wearables collect identity data such as fingerprints, iris patterns, habits, and voice, creating leakage risks when Metaverse servers fail or are attacked.
- Balance Between Virtual and Reality: The survey proposes access limits, reduced virtual realism, and regulatory mechanisms to prevent excessive immersion from disrupting users’ real lives.
6 CONCLUSION
The survey integrates 6G-oriented edge intelligence with the Metaverse, investigates systems and challenges, and reviews advanced methods and future research directions. It aims to guide further work on Metaverse development with 6G edge intelligence.
- The survey introduces the integration of 6G-oriented edge intelligence into the Metaverse and investigates three Metaverse systems based on 6G edge intelligence.
- It reviews core Metaverse challenges, advanced methods, and open topics and directions for future research.
- The survey aims to provide insights and guidance for researchers and practitioners expanding the Metaverse with 6G edge intelligence.