Multi-Device Cooperative Fine-Tuning of Foundation Models at the Network Edge DOI
Wu Hai, Xu Chen,

Kaibin Huang

et al.

Published: Aug. 7, 2024

Language: Английский

LLM-Based Edge Intelligence: A Comprehensive Survey on Architectures, Applications, Security and Trustworthiness DOI Creative Commons
Othmane Friha, Mohamed Amine Ferrag, Burak Kantarcı

et al.

IEEE Open Journal of the Communications Society, Journal Year: 2024, Volume and Issue: 5, P. 5799 - 5856

Published: Jan. 1, 2024

Language: Английский

Citations

9

Positioning Using Wireless Networks: Applications, Recent Progress, and Future Challenges DOI
Yang Yang, Mingzhe Chen,

Yufei Blankenship

et al.

IEEE Journal on Selected Areas in Communications, Journal Year: 2024, Volume and Issue: 42(9), P. 2149 - 2178

Published: July 10, 2024

Language: Английский

Citations

8

HARGPT: Are LLMs Zero-Shot Human Activity Recognizers? DOI

Sijie Ji,

Xinzhe Zheng,

Chenshu Wu

et al.

Published: May 13, 2024

Language: Английский

Citations

7

From screens to scenes: A survey of embodied AI in healthcare DOI
Yihao Liu, Xu Cao, Tingting Chen

et al.

Information Fusion, Journal Year: 2025, Volume and Issue: unknown, P. 103033 - 103033

Published: Feb. 1, 2025

Language: Английский

Citations

0

Towards AI-Powered Applications: The Development of a Personalised LLM for HRI and HCI DOI Creative Commons
Khashayar Ghamati, Maryam Banitalebi Dehkordi, Abolfazl Zaraki

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(7), P. 2024 - 2024

Published: March 24, 2025

In this work, we propose a novel Personalised Large Language Model (PLLM) agent, designed to advance the integration and adaptation of large language models within field human-robot interaction human-computer interaction. While research in has primarily focused on technical deployment LLMs, critical academic challenges persist regarding their ability adapt dynamically user-specific contexts evolving environments. To address fundamental gap, present methodology for personalising LLMs using domain-specific data tests NeuroSense EEG dataset. By enabling personalised interpretation, our approach promotes conventional implementation strategies, contributing ongoing AI adaptability user-centric application. Furthermore, study engages with broader ethical dimensions PLLM, critically discussing issues generalisability privacy concerns research. Our findings demonstrate usability PLLM scenario real-world settings, highlighting its applicability across diverse domains, including healthcare, education, assistive technologies. We believe proposed system represents significant step towards personalisation, offering substantial benefits range fields.

Language: Английский

Citations

0

Leveraging Large Language Models for Intelligent Control of 6G Integrated TN-NTN With IoT Service DOI
Bo Rong, Humphrey Rutagemwa

IEEE Network, Journal Year: 2024, Volume and Issue: 38(4), P. 136 - 142

Published: April 1, 2024

With the advent of sixth generation (6G) Internet Things (IoT), integrated terrestrial network (TN) and non-terrestrial (NTN) will play a vital role in enabling new applications services. However, realizing potential 6G TN-NTN requires addressing key challenges like intelligent optimized control mechanisms for resource management, interference cancellation, handover management. This paper explores large language models (LLMs) TN-NTN. LLMs can learn complex relationships patterns from large-scale data, then be fine-tuned on small labeled datasets, significantly reducing training time cost. study examines main obstacles integration IoT systems, further discusses how may effectively address those issues. Our suggested approach utilizes to create efficient anaptive algorithms that handle diverse, ever-changing, decentralized characteristics

Language: Английский

Citations

3

A Survey of 5G Core Network User Identity Protections, Concerns, and Proposed Enhancements for Future 6G Technologies DOI Creative Commons
Paul Scalise, Michael Hempel, Hamid Sharif

et al.

Future Internet, Journal Year: 2025, Volume and Issue: 17(4), P. 142 - 142

Published: March 25, 2025

Fifth-Generation (5G) cellular networks extensively utilize subscriber identifiers throughout the protocol stack, thereby linking subscribers to their activities on network. With inherent use of linked comes potential capability track subscribers’ location and behavior, which poses critical challenges for user identity protections privacy in sensitive applications like military or healthcare operating over public 5G infrastructure. The reliance such personal threatens a user’s right brings light importance proper mechanisms mitigate these risks current future network technologies. In this paper, we explore specifications understand most important list across Virtual Network Functions (VNF), points exposure within Core (CN). We also examine existing literature regarding efforts concerns targeted CN. Findings include need trust relationship between users providers protect safeguard identity. While technology has greater compared previous generations, our analysis shows that several areas concern remain, particularly exchange metadata. This work finds new technologies adopted add further complexity maintaining strict posture safeguarding protections. paper reviews scientific community’s proposed enhancements 6G networks’ protections, with focus emerging Artificial Intelligence (AI) Machine Learning (ML) applications. ethical implications private anonymous communications are carefully weighed examined multifaceted nature topic. Our is concluded by proposing research reduce prevalence as SUPI (Subscription Permanent Identifier) operations help better propose replacing widespread VNFs ephemeral identifiers, building upon 3GPP aiming from eavesdroppers.

Language: Английский

Citations

0

Efficient Inference Offloading for Mixture-of-Experts Large Language Models in Internet of Medical Things DOI Open Access
Xiaoming Yuan,

Weixuan Kong,

Zhenyu Luo

et al.

Electronics, Journal Year: 2024, Volume and Issue: 13(11), P. 2077 - 2077

Published: May 27, 2024

Despite recent significant advancements in large language models (LLMs) for medical services, the deployment difficulties of LLMs e-healthcare hinder complex applications Internet Medical Things (IoMT). People are increasingly concerned about risks and privacy protection. Existing face providing accurate questions answers (Q&As) meeting resource demands IoMT. To address these challenges, we propose MedMixtral 8x7B, a new LLM based on mixture-of-experts (MoE) architecture with an offloading strategy, enabling IoMT, improving protection users. Additionally, find that factors affecting latency include method device interconnection, location servers, speed disk.

Language: Английский

Citations

1

Multi-Device Cooperative Fine-Tuning of Foundation Models at the Network Edge DOI
Wu Hai, Xu Chen,

Kaibin Huang

et al.

Published: Aug. 7, 2024

Language: Английский

Citations

0