6G-enabled Edge AI for Metaverse: Challenges, Methods, and Future Research Directions

Chang, Luyi; Zhang, Zhe; Li, Pei; Xi, Shibo; Guo, Wei; Shen, Yukang; Xiong, Zehui; Kang, Jiawen; Niyato, Dusit; Qiao, Xiuquan; Wu, Yi

doi:10.48550/arxiv.2204.06192

Cited by 13 publications

(17 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It will be of interest to investigate how blockchain and mobile edge computing can help improve the system in [30]. A very recent survey [3] presents 6G-enabled edge AI for the Metaverse, but blockchain is not highlighted. In addition, [31] focuses on 6G for the Metaverse; [32] presents the Metaverse for wireless systems; and [33] explains edge-enabled Metaverse.…”

Section: Convergence Of the Technologies: Ai And Blockchain Assisted ...mentioning

confidence: 99%

“…The sixth generation of mobile communication systems (6G) meet the requirements of a comprehensive connected and smart digital environment [3], [4], supporting complete network coverage, enhanced intelligent applications and stronger security. Compared to previous generations of mobile communication technology, 6G will have a radical change from "connected things" to "connected intelligence", i.e., going beyond personalized communication to realize intelligent connections between people, devices, and resources.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mobile Edge Computing, Metaverse, 6G Wireless Communications, Artificial Intelligence, and Blockchain: Survey and Their Convergence

Wang¹,

Zhao²

2022

Preprint

View full text Add to dashboard Cite

With the advances of the Internet of Things (IoT) and 5G/6G wireless communications, the paradigms of mobile computing have developed dramatically in recent years, from centralized mobile cloud computing to distributed fog computing and mobile edge computing (MEC). MEC pushes computeintensive assignments to the edge of the network and brings resources as close to the endpoints as possible, addressing the shortcomings of mobile devices with regard to storage space, resource optimisation, computational performance and efficiency. Compared to cloud computing, as the distributed and closer infrastructure, the convergence of MEC with other emerging technologies, including the Metaverse, 6G wireless communications, artificial intelligence (AI), and blockchain, also solves the problems of network resource allocation, more network load as well as latency requirements. Accordingly, this paper investigates the computational paradigms used to meet the stringent requirements of modern applications. The application scenarios of MEC in mobile augmented reality (MAR) are provided. Furthermore, this survey presents the motivation of MEC-based Metaverse and introduces the applications of MEC to the Metaverse. Particular emphasis is given on a set of technical fusions mentioned above, e.g., 6G with MEC paradigm, MEC strengthened by blockchain, etc.

show abstract

Section: Convergence Of the Technologies: Ai And Blockchain Assisted ...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Mobile Edge Computing, Metaverse, 6G Wireless Communications, Artificial Intelligence, and Blockchain: Survey and Their Convergence

Wang¹,

Zhao²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…As its four pillar technologies (Figure 1), Digital Twins [2], [24] collects real-time data from the real-world industrial environment and creates digital avatars of machines and managers in the virtual world. AI (especially FL and Blockchain) [5], [25]- [27] learn from these digital data and assist managers in making decisions or executing automated controls, where FL and Blockchain provide privacy and security guarantees respectively. In addition, actions performed in the virtual world will also be synchronized with the real world in real time.…”

Section: Introductionmentioning

confidence: 99%

“…In addition, actions performed in the virtual world will also be synchronized with the real world in real time. 5G/6G High-Speed Communication and Scheduling and Mobile Edge Computing [27]- [31] infrastructures guarantee low latency and high reliability for real-time data stream synchronization, aiming to provide users with seamless and immersive interactive experiences. Federated Learning and Heterogeneity.…”

Section: Introductionmentioning

confidence: 99%

HFedMS: Heterogeneous Federated Learning with Memorable Data Semantics in Industrial Metaverse

Zeng¹,

Li²,

Yu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Federated Learning (FL), as a rapidly evolving privacy-preserving collaborative machine learning paradigm, is a promising approach to enable edge intelligence in the emerging Industrial Metaverse. Even though many successful use cases have proved the feasibility of FL in theory, in the industrial practice of Metaverse, the problems of non-independent and identically distributed (non-i.i.d.) data, learning forgetting caused by streaming industrial data, and scarce communication bandwidth remain key barriers to realize practical FL. Facing the above three challenges simultaneously, this paper presents a highperformance and efficient system named HFEDMS for incorporating practical FL into Industrial Metaverse. HFEDMS reduces data heterogeneity through dynamic grouping and training mode conversion (Dynamic Sequential-to-Parallel Training, STP). Then, it compensates for the forgotten knowledge by fusing compressed historical data semantics and calibrates classifier parameters (Semantic Compression and Compensation, SCC). Finally, the network parameters of the feature extractor and classifier are synchronized in different frequencies (Layer-wise Alternative Synchronization Protocol, LASP) to reduce communication costs. These techniques make FL more adaptable to the heterogeneous streaming data continuously generated by industrial equipment, and are also more efficient in communication than traditional methods (e.g., Federated Averaging). Extensive experiments have been conducted on the streamed non-i.i.d. FEMNIST dataset using 368 simulated devices. Numerical results show that HFEDMS improves the classification accuracy by at least 6.4% compared with 8 benchmarks and saves both the overall runtime and transfer bytes by up to 98%, proving its superiority in precision and efficiency.

show abstract

“…M ETAVERSE is the next generation Internet based platform which allows the Metaverse users (MUs) to virtually interact with the living and non-living entities of the physical world in real-time [1], [2]. Recent technologies such as Virtual Reality (VR), Augmented Reality (AR), Digital Twin (DT), Beyond 5G (B5G) communication, Artificial Intelligence (AI), and Blockchain help the MUs to have realtime and seamless interaction with the three dimensional (3D) virtual world [3]- [5]. Recently, Metaverse gains attention of both academic researchers and industrialists because of its huge applicability in different domains such as smart industry, online -education, gaming, meeting or conference, shopping, virtual driving, social interaction, and to name a few [6]- [10].…”

Section: Introductionmentioning

confidence: 99%

D2R2: Distributed and Dynamic Reputation based Resource Allocation in Metaverse

Kalita¹,

Gurusamy²

2022

Preprint

View full text Add to dashboard Cite

<p>Metaverse is the fusion of physical and virtual world, where Metaverse users (MUs) interact with both physical and virtual world in real-time. To improve the Quality of Experience (QoE) of the MUs, it is essential to feed the whole Metaverse system with high quality real-world data such as higher sampling rate and bit rate to create better and up-to-date replication of the physical world i.e., digital (or mirror) world. On the other hand, the owner of the smart Internet of Things (IoT) devices or sensors that collect data from the real-world on behalf of the Metaverse systems should be well paid for collecting such high quality data. Therefore, in this work, to maintain the balance between the required quality of data, and the profit and cost associated with such data collection in Metaverse systems, we design a Distributed and Dynamic Reputation based Resource Allocation (D2R2) model. This D2R2 model leads the rational smart devices to feed the Metaverse systems with high quality data by maintaining its reputation while earning higher payoffs. To find the optimal solution of the proposed D2R2 model, we first fit the D2R2 into the Stackelberg gaming model, and then prove that proposed D2R2 model have Stackelberg Equilibrium point. Later, we use Alternating Direction Method of Multipliers (ADMM) algorithm for its faster convergence to find the optimal solutions of the Stackelberg gaming model in a distributed manner. Finally, we implement the proposed D2R2 model in MATLAB and provide numerical results to demonstrate the effectiveness and efficiency of the proposed model.</p>

show abstract

6G-enabled Edge AI for Metaverse: Challenges, Methods, and Future Research Directions

Cited by 13 publications

References 49 publications

Mobile Edge Computing, Metaverse, 6G Wireless Communications, Artificial Intelligence, and Blockchain: Survey and Their Convergence

Mobile Edge Computing, Metaverse, 6G Wireless Communications, Artificial Intelligence, and Blockchain: Survey and Their Convergence

HFedMS: Heterogeneous Federated Learning with Memorable Data Semantics in Industrial Metaverse

D2R2: Distributed and Dynamic Reputation based Resource Allocation in Metaverse

Contact Info

Product

Resources

About