Unleashing the Power of Edge-Cloud Generative AI in Mobile Networks: A Survey of AIGC Services

IEEE Communications Surveys & Tutorials, Journal Year: 2024, Volume and Issue: 26(2), P. 1127 - 1170

Published: Jan. 1, 2024

Artificial Intelligence-Generated Content (AIGC) is an automated method for generating, manipulating, and modifying valuable diverse data using AI algorithms creatively. This survey paper focuses on the deployment of AIGC applications, e.g., ChatGPT Dall-E, at mobile edge networks, namely that provide personalized customized services in real time while maintaining user privacy. We begin by introducing background fundamentals generative models lifecycle which includes collection, training, fine-tuning, inference, product management. then discuss collaborative cloud-edge-mobile infrastructure technologies required to support enable users access networks. Furthermore, we explore AIGC-driven creative applications use cases Additionally, implementation, security, privacy challenges deploying Finally, highlight some future research directions open issues full realization

Language: Английский

---

Task-Oriented Communications for 6G: Vision, Principles, and Technologies

IEEE Wireless Communications, Journal Year: 2023, Volume and Issue: 30(3), P. 78 - 85

Published: June 1, 2023

Driven by the interplay among artificial intelligence, digital twin, and wireless networks, 6G is envisaged to go beyond data-centric services provide intelligent immersive experiences. To efficiently support tasks with customized service requirements, it becomes critical develop novel information compression transmission technologies, which typically involve coupled sensing, communication, computation processes. this end, task-oriented communication stands out as a disruptive technology for system design exploiting task-specific structures folding goals into of task-level strategies. In article, developing extraction network resource orchestration strategies, we demonstrate effectiveness principles typical tasks, including federated learning, edge inference, semantic communication.

Language: Английский

---

Energy Efficient Semantic Communication Over Wireless Networks With Rate Splitting

IEEE Journal on Selected Areas in Communications, Journal Year: 2023, Volume and Issue: 41(5), P. 1484 - 1495

Published: Jan. 30, 2023

In this paper, the problem of wireless resource allocation and semantic information extraction for energy efficient communications over networks with rate splitting is investigated. considered model, a base station (BS) first extracts from its large-scale data, then transmits small-sized to each user which recovers original data based on local common knowledge. At BS side, probability graph used extract multi-level information. downlink transmission, scheme adopted, while private transmitted through message knowledge message. Due limited resource, both computation transmission are considered. This joint communication formulated as an optimization aiming minimize total consumption network under computation, latency, transmit power constraints. To solve problem, alternating algorithm proposed where closed-form solutions ratio frequency obtained at step. Numerical results verify effectiveness algorithm.

Language: Английский

---

Large AI Model-Based Semantic Communications

IEEE Wireless Communications, Journal Year: 2024, Volume and Issue: 31(3), P. 68 - 75

Published: June 1, 2024

Semantic communication (SC) is an emerging intelligent paradigm, offering solutions for various future applications like metaverse, mixed reality, and the Internet of everything. However, in current SC systems, construction knowledge base (KB) faces several issues, including limited representation, frequent updates, insecure sharing. Fortunately, development large AI model (LAM) provides new to overcome above issues. Here, we propose a LAM-based framework (LAM-SC) specifically designed image data, where first apply segment anything (SAM)-based KB (SKB) that can split original into different semantic segments by universal knowledge. Then, present attention-based integration (ASI) weigh generated SKB without human participation integrate them as semantic-aware image. Additionally, adaptive compression (ASC) encoding remove redundant information features, thereby reducing overhead. Finally, through simulations, demonstrate effectiveness LAM-SC possibility applying paradigms.

Language: Английский

---

Making Sense of Meaning: A Survey on Metrics for Semantic and Goal-Oriented Communication

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 45456 - 45492

Published: Jan. 1, 2023

Semantic communication (SemCom) aims to convey the meaning behind a transmitted message by transmitting only semantically-relevant information. This semantic-centric design helps minimize power usage, bandwidth consumption, and transmission delay. SemCom goal-oriented (or effectiveness-level SemCom) are therefore promising enablers of 6G developing rapidly. Despite surge in their swift development, design, analysis, optimization, realization robust intelligent as well fraught with many fundamental challenges. One challenges is that lack unified/universal metrics can stifle research progress on respective algorithmic, theoretical, implementation frontiers. Consequently, this survey paper documents existing – scattered references wireless SemCom, optical quantum SemCom. By doing so, inspire optimization wide variety systems. article also stimulates development performance assessment purely statistical hardly applicable reasoning-type tasks constitute heart beyond.

Language: Английский

---

Non-Orthogonal Multiple Access Enhanced Multi-User Semantic Communication

IEEE Transactions on Cognitive Communications and Networking, Journal Year: 2023, Volume and Issue: 9(6), P. 1438 - 1453

Published: Aug. 21, 2023

Semantic communication has served as a novel paradigm and attracted broad interest from researchers. One critical aspect of it is the multi-user semantic theory, which can further enhance its application for practical network environment. While most existing works have focused on design end-to-end single-user transmission, non-orthogonal multiple access (NOMA)-based system named NOMASC proposed in this study. The support transmission users with diverse modalities source information. To avoid high demand hardware, an asymmetric quantizer employed at end encoder discretizing continuous full-resolution feature. In addition, neural model used mapping discrete feature into self-learned symbols accomplishing intelligent detection (MUD) receiver side. Simulation results demonstrate that attains sound performance user signals outperforms other methods, particularly low-to-medium signal-to-noise ratios (SNRs). Moreover, shown robustness under various simulation settings mismatched testing scenarios.

Language: Английский

---

On the Utility-Informativeness-Security Trade-off in Discrete Task-Oriented Semantic Communication

IEEE Communications Letters, Journal Year: 2024, Volume and Issue: 28(6), P. 1298 - 1302

Published: March 28, 2024

Task-oriented semantic communication (ToSC) has been applied to edge inference tasks with limited and computing resources. By encoding the task-related features into a finite-size codebook transmitting index of codebook, ToSC can be compatible existing discrete systems. In such systems, shared by transmitter receiver contains information original data affects task performance, which may also obtained adversaries. Thus, there exists an inherent utility-informativeness-security (UIS) trade-off problem in This letter introduces novel framework, named UIS-ToSC, leverages vector quantized variational bottleneck (VQ-VIB) scheme for issue. Furthermore, we exploit adversarial learning (AL) train system against leakage. Comprehensive experiments demonstrate that proposed efficiently reduce overhead maintain security little influence on task-inference utility.

Language: Английский

---

Introduction and overview

Elsevier eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 3 - 9

Published: Jan. 1, 2025

Language: Английский

---

Learning to compress via robust information bottleneck with digital modulation

Elsevier eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 67 - 91

Published: Jan. 1, 2025

Language: Английский

---

Joint Source-Channel Coding System for 6G Communication: Design, Prototype and Future Directions

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 17708 - 17724

Published: Jan. 1, 2024

The emergence of the AI era signifies a shift in future landscape global communication networks, wherein robots are expected to play more prominent role compared humans. establishment novel paradigm for development next-generation 6G is utmost importance semantics task-oriented empowered communications. goal semantic lies integration collaborative efforts between intelligence transmission source and joint design coding channel coding. This characteristic represents significant benefit source-channel (JSCC), as it enables generation alphabets with diverse lengths achieves code rate unity. Therefore, we leverage not only quasi-cyclic (QC) characteristics facilitate utilization flexible structural hardware but also Unequal Error Protection (UEP) ensure recovery importance. In this study, feasibility using encoder/decoder that aware UEP can be explored based on existing JSCC system. approach aimed at protecting significance information. Additionally, deployment system facilitated by employing QC-Low-Density Parity-Check (LDPC) codes reconfigurable device. QC-LDPC layered decoding technique, which has been specifically optimized parallelism tailored applications, suitably adapted accommodate performance proposed evaluated conducting BER measurements both floating-point 6-bit quantization. done assess extent deterioration fair manner. fixed-point synthesized subsequently used feature reception across noisy channel, aim presenting prototype study concludes some insights potential research avenues context communication.

Language: Английский

---