Layered architecture of enterprise AI agent infrastructure: an analytical framework based on cloud computing practice

Tian Yiqing; Shen Yang

doi:10.11959/j.issn.1000-0801.DXKX260135

您当前的位置：

首页 >

文章列表页 >

Layered architecture of enterprise AI agent infrastructure: an analytical framework based on cloud computing practice

更新时间：2026-07-21

- Layered architecture of enterprise AI agent infrastructure: an analytical framework based on cloud computing practice
- Telecommunications Science Vol. 42, Issue 6, Pages: 66-77(2026)
- 作者机构：
  
  1.北京邮电大学电子工程学院，北京 100876
  2.北京欧珀通信有限公司，北京 100026
- 作者简介：
- 基金信息：
- DOI：10.11959/j.issn.1000-0801.DXKX260135
  CLC： TN915.5
- Received：28 February 2026，
  
  Revised：2026-04-10，
  
  Accepted：16 April 2026，
  
  Online First：21 July 2026，
  
  Published：20 June 2026
- 稿件说明：
移动端阅览
田一晴,沈洋.企业AI智能体基础设施的层次化架构：基于云计算实践的分析框架[J].电信科学,2026,42(06):66-77.

Tian Yiqing,Shen Yang.Layered architecture of enterprise AI agent infrastructure: an analytical framework based on cloud computing practice[J].Telecommunications Science,2026,42(06):66-77.
田一晴,沈洋.企业AI智能体基础设施的层次化架构：基于云计算实践的分析框架[J].电信科学,2026,42(06):66-77. DOI： 10.11959/j.issn.1000-0801.DXKX260135.

Tian Yiqing,Shen Yang.Layered architecture of enterprise AI agent infrastructure: an analytical framework based on cloud computing practice[J].Telecommunications Science,2026,42(06):66-77. DOI： 10.11959/j.issn.1000-0801.DXKX260135.

摘要

大语言模型（large language model，LLM）的工程化落地催生了以自主决策与工具调用为核心的智能体系统。然而，将智能体从受控原型推向企业生产环境仍面临执行隔离、身份治理、工具生态治理、多体协同、可观测性和生态分发6类结构性挑战。现有研究多聚焦于智能体的推理架构或单一平台能力，缺乏面向企业全生命周期部署需求的系统性基础设施分析框架。从上述6类工程挑战出发，提出覆盖执行运行时、身份与安全、工具生态集成、多智能体编排、可观测性与评估、市场与生态分发的6层参考框架，并以主流云计算平台和开源社区项目的工程实践为参照进行双路径验证。在此基础上，将新框架与3GPP SA2#173会议关于6G核心网人工智能（AI）架构演进的研究进展进行对比，发现两者在智能体身份治理、工具调用机制、多智能体协作、安全管控等维度上呈现演进性趋同倾向（evolutionary convergence tendency）。新框架为理解面向6G的智能体原生网络架构演进提供了来自云计算实践的分析视角。

Abstract

The engineering deployment of large language model (LLM) has given rise to agent systems with autonomous decision-making and tool-calling capabilities. However

transitioning from controlled prototypes to enterprise production still faces six structural challenges: execution isolation

identity governance

tool ecosystem governance

multi-agent coordination

observability

and ecosystem distribution. Existing research primarily focuses on agent reasoning architectures or individual platform capabilities

lacking a systematic infrastructure framework for full-lifecycle enterprise deployment. A six-layer reference framework was proposed covering agent runtime

identity and security

tool integration

multi-agent orchestration

observability and evaluation

and marketplace distribution. The design logic of each layer was validated through dual-path verification against both enterprise cloud platforms and open-source community projects. An evolutionary convergence tendency across agent identity

tool invocation

multi-agent coordination

and governance security was revealed by a systematic comparison with 3GPP SA2#173 research. The framework provides an analytical lens from cloud computing practice for understanding agent-native 6G network architecture evolution.

关键词

Keywords

references

Gartner . Gartner predicts 40% of enterprise apps will feature task-specific AI agents by 2026 [R ] . 2025-08-26 .

Cleanlab . AI agents in production 2025: enterprise trends and best practices [R ] . 2026-03-01 .

Wang L , Ma C , Feng X Y , et al . A survey on large language model based autonomous agents [J ] . Frontiers of Computer Science , 2024 , 18 ( 6 ): 186345 .

Weng L . LLM-powered autonomous agents [EB ] . 2023 .

Guo T , Chen X , Wang Y , et al . Large language model based multi-agents: a survey of progress and challenges [C ] // Proceedings of the 33rd IJCAI . Burlington, Massachusetts : Morgan Kaufmann , 2024 : 8048 - 8057 .

Xi Z H , Chen W X , Guo X , et al . The rise and potential of large language model based agents: a survey [J ] . Science China Information Sciences , 2025 , 68 ( 2 ): 121101 .

Zheng L M , Chiang W L , Sheng Y , et al . Judging LLM-as-a-judge with MT-bench and chatbot arena [C ] // Proceedings of the 37th International Conference on Neural Information Processing Systems . New York : ACM Press , 2023 : 46595 - 46623 .

ITU-R M . 2160: 2023 Framework and overall objectives of the future development of IMT for 2030 and beyond [S ] .

Ehtesham A , Singh A , Gupta G K , et al . A survey of agent interoperability protocols: model context protocol (MCP), agent communication protocol (ACP), agent-to-agent protocol (A2A), and agent network protocol (ANP) [PP ] . V2. arXiv ( 2025-05-23 )[ 2026-01-23 ] . arXiv: 2505.02279 .

Anthropic . Model context protocol specification [EB ] . 2024 .

Google . Agent-to-agent (A2A) protocol specification [EB ] . 2025 .

OpenID Foundation . Identity management for agentic AI [R ] . 2026-02-15 .

Amazon Web Services . Amazon Bedrock AgentCore documentation [EB ] . 2025 .

OpenClaw . OpenClaw: self-hosted gateway for AI agents [EB ] . 2026 .

ClawHub . Public skill registry for OpenClaw [EB ] . 2026 .

Agache A , Brooker M , Iordache A , et al . Firecracker: lightweight virtualization for serverless applications [C ] // Proceedings of the 17th USENIX NSDI . Berkeley : USENIX Association , 2020 : 419 - 434 .

ETSI GS ZSM 002 V1.1.1: 2019 Zero-touch network and service management (ZSM) reference architecture [S ] .

3 GPP . S2-2601624 : KI#18: 2026 AI for 6G architecture (moderator merged tdoc)[Z ] .

3 GPP . S2-2601625 : KI#19: 2026 Solution variants[Z ] .

3 GPP . S2-2600070: 2026 Solution for AI governance and symbolic validation [Z ] .

3GPP. TS 33.127: 2026 Lawful interception (LI) architecture and functions. V18.12.0 [S ] .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Research on multi-agent latent space strategies based on frozen main branch dual-loop evolution

A collaborative behavior security architecture for 6G agent networks

Endogenous intelligent agent architecture and intent-driven subnetwork generation mechanism for 6G core networks

Research on the technical system and key technologies of integrated communication, sensing, computing, and intelligence for 6G

Heterogeneous computing power scheduling system and applications under the background of computing- network convergence

Related Author

田一晴

沈洋

Zhao Yuanhe

Wang Yu

Li Haochen

Yang Shiduo

Guo Kenan

Gao Chen

Related Institution

北京欧珀通信有限公司¹

北京邮电大学¹

Tianjin Research Institute for Water Transport Engineering, M.O.T

School of Control and Computer Engineering, North China Electric Power University

National Engineering Research Center of Advanced Network Technologies, School of Electronic and Information Engineering, Beijing Jiaotong University

⁰