当智能体AI遇见通感一体化
文章背景与核心概要
本文探讨了智能体人工智能(Agentic AI)与通感一体化(ISAC)的深度融合,并引入了一个名为 AISAC 的统一范式。长期以来,基于学习的感知、资源分配、可重构智能表面(RIS)以及边缘智能等技术,在很大程度上作为面向功能的物理层技术被孤立地进行研究。
为了弥合物理层原语与完全自主系统之间的鸿沟,作者提出了一个全面的六阶段闭环框架以及五级智能体成熟度模型。通过对代表性研究的审视,本综述揭示了宣称的智能体成熟度与实际演示之间的巨大差距,并随后指出了物理到语义基础、预测性世界模型以及资源高效自主性等方面面临的开放性挑战。
元数据 (Metadata)
- arXiv ID:
arXiv:2608.05792[cs.AI] - 学科领域: 人工智能 (
cs.AI) - 提交日期: 2026年8月6日
- 作者:
- Kai Li
- Conggai Li
- Sarah Ali Siddiqui
- Syed Sohail Ahmed
- Xin Yuan
- Shenghong Li
- Wei Ni
- 文档统计: 35页,132篇参考文献,10张表格,9幅图表
摘要 (Abstract)
智能体人工智能(Agentic AI)正在将通感一体化(ISAC)从一种面向功能的物理层技术转变为一个目标驱动的闭环智能系统,我们将这一范式称为 AISAC。
Agentic artificial intelligence (AI) is transforming Integrated Sensing and Communication (ISAC) from a function-oriented physical-layer technology into a goal-driven, closed-loop intelligent system, a paradigm we term AISAC.
现有关于基于学习的感知、资源分配、可重构智能表面(RIS)、边缘智能、多智能体协同以及弹性网络的研究,大都在很大程度上处于孤立状态。本综述将相关文献统一在一个包含以下六个阶段的闭环框架中: 1. 感知(Observation) 2. 情境化(Contextualization) 3. 推理与预测(Reasoning and prediction) 4. 规划与编排(Planning and orchestration) 5. 执行与协作(Execution and collaboration) 6. 反馈与弹性(Feedback and resilience)
Existing work on learning-based sensing, resource allocation, reconfigurable intelligent surfaces (RIS), edge intelligence, multi-agent coordination, and resilient networking has developed largely in isolation. This survey unifies the literature within a six-stage closed-loop framework comprising: 1. Observation 2. Contextualization 3. Reasoning and prediction 4. Planning and orchestration 5. Execution and collaboration 6. Feedback and resilience
本文还引入了五级智能体成熟度,涵盖从物理层原语到完全闭环的智能体 ISAC。我们利用该框架回顾了多模态智能、大语言模型、强化学习、联邦学习、RIS 辅助控制、无人机(UAV)与车联网以及 AI 原生网络管理方面的进展,并分析了隐私、安全、弹性和可持续性作为完整“感知-推理-动作”循环中跨领域的需求。
The paper also introduces five levels of agentic maturity, ranging from physical-layer primitives to fully closed-loop agentic ISAC. We use this framework to review advances in multimodal intelligence, large language models, reinforcement learning, federated learning, RIS-assisted control, Unmanned Aerial Vehicle (UAV) and vehicular networks, and AI-native network management, and analyze privacy, security, resilience, and sustainability as cross-cutting requirements of the full perception-reasoning-action loop.
对照九项特定于智能体的评估标准对代表性研究进行的审核表明,没有任何系统报告超过其中一到两项标准,这暴露出宣称的智能体成熟度与实际展示的成熟度之间存在差距。最后,我们指出了在物理到语义基础、预测性世界模型、实时智能体-物理层(Agent-PHY)交互、安全的工具使用、异构多智能体协作、基准测试以及资源高效的自主性等方面的开放性挑战。
An audit of representative studies against nine agentic-specific evaluation criteria shows that no system reports more than one or two of them, exposing a gap between claimed and demonstrated agentic maturity. Finally, we identify open challenges in physical-to-semantic grounding, predictive world models, real-time agent-PHY interaction, safe tool use, heterogeneous multi-agent collaboration, benchmarking, and resource-efficient autonomy.
全文与参考链接 (Full-Text & Reference Links)
- 访问论文: 查看 PDF | HTML(实验性) | TeX 源码
- 数字对象唯一标识符 (DOI): 10.48550/arXiv.2608.05792
- 外部引用: 谷歌学术 | Semantic Scholar | NASA ADS