Scholarlink

<p>Sixth-generation radiocommunications (6G) systems are adopting language models for intent-driven control, context-aware adaptation to environmental and network dynamics, and end-to-end orchestration of communication. Conventional artificial intelligence (AI) typically lacks capabilities in task generalization, communication and reasoning; however, the rapid development of efficient large language models (LLMs) can automate mobile and network operations and inform the design of 6G. In this Perspective, we discuss the use of LLMs in 6G networks. We show how cloud LLMs can improve the self-organization, efficiency and local deployment of 6G networks. Next, we describe key techniques for implementing LLMs on devices for 6G. Finally, we propose LLMs for in multi-agent scenarios, emphasizing the importance of telecom-specific adaptations and security. LLMs have the potential to improve network design, operation and service delivery by enabling telecom stakeholders to convert intents into trustworthy, privacy-aware decisions that deliver more reliable services for users.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Large language models in 6G from standard to on-device networks

  • 期刊论文

  • 发表日期：

    2026年1月6日

查看全文
  • Hang Zou,
  • Qiyang Zhao,
  • Samson Lasaulce,
  • Chao Zhang,
  • Yu Tian,
  • Lina Bariah,
  • Faouzi Bader,
  • Merouane Debbah

摘要

Sixth-generation radiocommunications (6G) systems are adopting language models for intent-driven control, context-aware adaptation to environmental and network dynamics, and end-to-end orchestration of communication. Conventional artificial intelligence (AI) typically lacks capabilities in task generalization, communication and reasoning; however, the rapid development of efficient large language models (LLMs) can automate mobile and network operations and inform the design of 6G. In this Perspective, we discuss the use of LLMs in 6G networks. We show how cloud LLMs can improve the self-organization, efficiency and local deployment of 6G networks. Next, we describe key techniques for implementing LLMs on devices for 6G. Finally, we propose LLMs for in multi-agent scenarios, emphasizing the importance of telecom-specific adaptations and security. LLMs have the potential to improve network design, operation and service delivery by enabling telecom stakeholders to convert intents into trustworthy, privacy-aware decisions that deliver more reliable services for users.