



network working group                                             B. Liu
Internet-Draft                                                  L. Zhang
Intended status: Informational                       Huawei Technologies
Expires: 23 April 2026                                            Y. Liu
                                                             Independent
                                                         20 October 2025


    Agents Networking Scenarios in Enterprise and Broadband Networks
                draft-zl-agents-networking-scenarios-00

Abstract

   This document describes agents networking scenarios in enterprise and
   home broadband networks.  These scenarios differ from 6G and Internet
   scenarios.  Since the agentic service is still at the emerging stage,
   especially in enterprise and home broadband networks, the scenarios
   are mostly based on reasonable assumptions.

Status of This Memo

   This Internet-Draft is submitted in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF).  Note that other groups may also distribute
   working documents as Internet-Drafts.  The list of current Internet-
   Drafts is at https://datatracker.ietf.org/drafts/current/.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   This Internet-Draft will expire on 23 April 2026.

Copyright Notice

   Copyright (c) 2025 IETF Trust and the persons identified as the
   document authors.  All rights reserved.











Liu, et al.               Expires 23 April 2026                 [Page 1]

Internet-Draft         Agents Networking Scenarios          October 2025


   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents (https://trustee.ietf.org/
   license-info) in effect on the date of publication of this document.
   Please review these documents carefully, as they describe your rights
   and restrictions with respect to this document.  Code Components
   extracted from this document must include Revised BSD License text as
   described in Section 4.e of the Trust Legal Provisions and are
   provided without warranty as described in the Revised BSD License.

Table of Contents

   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   2
     1.1.  Requirements Language . . . . . . . . . . . . . . . . . .   2
   2.  Enterprise Scenarios  . . . . . . . . . . . . . . . . . . . .   3
     2.1.  Smart Office  . . . . . . . . . . . . . . . . . . . . . .   3
     2.2.  Agentic Digitalization  . . . . . . . . . . . . . . . . .   3
     2.3.  Industrial Automation . . . . . . . . . . . . . . . . . .   4
   3.  Home Broadband Network Scenarios  . . . . . . . . . . . . . .   5
   4.  Operational Considerations  . . . . . . . . . . . . . . . . .   6
     4.1.  Self-contained Private Deployment . . . . . . . . . . . .   6
     4.2.  Economical Integrated Implementation  . . . . . . . . . .   6
   5.  Security Considerations . . . . . . . . . . . . . . . . . . .   7
   6.  IANA Considerations . . . . . . . . . . . . . . . . . . . . .   7
   7.  Acknowledgements  . . . . . . . . . . . . . . . . . . . . . .   7
   8.  Normative References  . . . . . . . . . . . . . . . . . . . .   7
   Authors' Addresses  . . . . . . . . . . . . . . . . . . . . . . .   7

1.  Introduction

   With the rapid development of artificial intelligence, single-agent
   systems have gradually revealed their limitations in handling
   complex, multi-task, and cross-domain scenarios.  Agent networking,
   as a core paradigm for breaking through individual capabilities and
   achieving collective intelligence, has become a key trend in the
   future development of agents.

   This document describes agents networking scenarios in enterprise and
   home broadband networks.

1.1.  Requirements Language

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
   "OPTIONAL" in this document are to be interpreted as described in
   BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all
   capitals, as shown here.





Liu, et al.               Expires 23 April 2026                 [Page 2]

Internet-Draft         Agents Networking Scenarios          October 2025


2.  Enterprise Scenarios

   The AI Agent is progressively becoming the key technological unit
   supporting both internal enterprise process automation and cross-
   enterprise ecosystem collaboration.  Compared to the traditional API
   interface integration model, the AI Agent is capable of achieving
   more flexible and loosely coupled collaboration methods through
   abilities such as semantic perception, strategy-driven execution, and
   autonomous decision-making.  The essence of this change is an upgrade
   from "system integration" to an "Intranet of Agents."

2.1.  Smart Office

   In the Smart Office scenario, AI Agents enhance daily productivity
   and collaboration.

   Agent Collaboration:

   *  Personal Assistant Coordination: A user's personal AI Agent
      (residing on a mobile or PC) coordinates with other specialized
      Agents, such as the Meeting Room Agent, IT Service Agent etc..
      This collaboration facilitates automated task execution, including
      scheduling meetings, reserving room resources, preparing necessary
      documents, and initiating video conference sessions.

   *  Team Task Alignment: Multiple project-related Agents, representing
      different team members or functions, engage in real-time data
      sharing, automatic summarization of meeting minutes, dynamic task
      prioritization, and collaborative document editing.

   Network Requirements:

   *  Low Latency and High Reliability: Crucial for seamless video
      conferencing and real-time collaborative applications.

   *  Unified Agent Discovery and Invocation: Agents must rely on
      standard protocols to locate and invoke the capabilities of other
      Agents within the enterprise network (e.g., printing or storage
      services).

   *  Secure Authentication and Authorization: to ensure Agent actions
      are strictly compliant with the authorized scope granted by the
      human user or project manager.

2.2.  Agentic Digitalization

   This scenario focuses on automating and optimizing business processes
   using autonomous Agents.



Liu, et al.               Expires 23 April 2026                 [Page 3]

Internet-Draft         Agents Networking Scenarios          October 2025


   Agent Collaboration:

   *  Business Workflow Automation: A Process Management Agent
      collaborates with various Business System Agents (e.g., CRM Agent,
      ERP Agent).  It autonomously translates high-level business
      intents (e.g., "maximize customer retention") into executable
      plans, decomposes the overall task, and executes complex workflows
      across disparate business systems by invoking their respective
      Agents.

   *  Data-Driven Decision Making: Data Collection Agents aggregate
      information from various sources, feeding it to Data Analysis
      Agents.  The resulting insights guide Decision Agents, which
      autonomously take prescribed actions, such as dynamically
      adjusting cloud resource allocations.

   Networking and Communication Requirements:

   *  Heterogeneous Interoperability: Agent communication protocols must
      be designed to bridge communication gaps between legacy IT system
      APIs and modern AI Agent platforms.

   *  High Concurrency and Scalability: The network infrastructure must
      be capable of supporting numerous Agents engaged in frequent,
      complex, and high-volume interactions across the enterprise.

2.3.  Industrial Automation

   This scenario applies Agents to the operational domain, demanding
   strict performance guarantees.

   Agent Collaboration:

   *  High Efficient Producing: On the production floor, Sensors,
      Robots, Controllers collaborate to continuously monitor
      environmental variables and production status.  This tight
      collaboration allows for dynamic optimization of manufacturing
      processes, maximizing efficiency and minimizing downtime.

   *  High Quality Producing: Remote Diagnostics Agents collaborate with
      local Equipment Agents to transmit high-definition video feeds and
      machine logs.  This enables sophisticated remote fault diagnosis
      and highly accurate predictive maintenance operations.

   Networking and Communication Requirements:






Liu, et al.               Expires 23 April 2026                 [Page 4]

Internet-Draft         Agents Networking Scenarios          October 2025


   *  Ultra-Low Latency and Jitter: Critical control loops require
      communication that adheres to industrial-grade standards,
      specifically Ultra-Reliable Low-Latency Communication.

   *  Deterministic Networking (DetNet): The underlying network may need
      to incorporate DetNet technologies to guarantee bounded and
      predictable latency for communications among critical control
      Agents.

   *  OT/IT Convergence: A key requirement is defining how Agents can
      securely and reliably communicate with gateway functions
      connecting the Operational Technology (OT) domain to the
      Information Technology (IT) domain.

   *  Mobility Management: the robots and AGVs are constently moving, it
      is critical to maintain the session for the moving nodes.

3.  Home Broadband Network Scenarios

   The home network scenario is characterized by user experience
   optimization and device collaboration.

   Agent Collaboration:

   *  Home Security: For instance, a smart camera may stream footage to
      a NAS Agent for storage of anomalous events, while simultaneously
      alerting to the user's Mobile Agent.

   *  User Experience Guarantee: The Home Gateway (e.g., on the home
      router/ONT, or on the BNG) collaborates with Terminal Agents
      (e.g., on gaming consoles, smart TVs) to classify traffic.  This
      allows the Gateway to dynamically prioritize and allocate
      bandwidth to latency/bandwidth-sensitive applications (like online
      gaming or 4K streaming), ensuring a consistently high Quality of
      Experience (QoE).

   Networking and Communication Requirements:

   *  Lightweight and Efficient Protocols: Given the variety of
      resource-constrained consumer devices, the Agent communication
      protocol must be designed to be lightweight and computationally
      efficient.

   *  Mobility Management: The network architecture must provide
      mechanisms to maintain session persistence and support seamless
      handover for Agents when a user transitions between the home Wi-Fi
      network and a mobile cellular network.




Liu, et al.               Expires 23 April 2026                 [Page 5]

Internet-Draft         Agents Networking Scenarios          October 2025


4.  Operational Considerations

   There are two primary operational paradigms in varying enterprise and
   home requirements: the fully isolated Self-contained Private
   Deployment and the resource-optimized Economical Integrated
   Implementation.

4.1.  Self-contained Private Deployment

   This model emphasizes security, performance, and control, typically
   favored by large enterprises, industrial environments (Smart
   Manufacturing), or highly security-conscious users.

   *  Isolated Infrastructure: All core AI Agent components, including
      the Large Language Model (LLM) inference engines, Agent
      orchestrators, Agent naming/discovery services etc., are deployed
      entirely within the private network domain (e.g., enterprise LAN
      or private cloud).

   *  Data Sovereignty: No Agent communication data, task context, or
      locally sensitive information leaves the private network.  This is
      essential for scenarios involving proprietary data or compliance
      with strict regulatory requirements.

   *  On-Premises Compute Resources: Requires substantial dedicated
      compute and storage resources (e.g., GPUs for inference) within
      the private domain, which must be connected via high-speed
      interfaces.

   *  Controlled Network Egress: Strict policies are applied to prevent
      Agents from accessing unauthorized external services, minimizing
      the external attack surface.  The network needs mechanisms for
      precise monitoring of all egress traffic initiated by Agents.

4.2.  Economical Integrated Implementation

   This model prioritizes cost-efficiency, and leverages existing cloud
   and public infrastructure, often seen in home networks (Home
   Broadband) and smaller Smart Office setups.

   *  Hybrid Architecture: The core intelligence (e.g., the LLM) and
      central coordination services (e.g., Agent discovery) are
      typically hosted in a public or operator cloud environment.  Local
      Agents (e.g., device Agents, edge Agents on the home gateway)
      handle sensing, local actuation, and interface with the cloud-
      based central Agent.





Liu, et al.               Expires 23 April 2026                 [Page 6]

Internet-Draft         Agents Networking Scenarios          October 2025


   *  Tool/API Delegation: Cloud-based Agents often delegate local tasks
      by invoking APIs exposed by local Agents, or by sending compressed
      instructions to the edge device.

   *  Cost Optimization: This model reduces the need for comprehensive
      system that combined by various components, but rather, some
      integrated model of providing the service (e.g. through an Agent
      Gateway).

5.  Security Considerations

   TBD

6.  IANA Considerations

   This document has no IANA actions.

7.  Acknowledgements

   TBD

8.  Normative References

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119,
              DOI 10.17487/RFC2119, March 1997,
              <https://www.rfc-editor.org/info/rfc2119>.

   [RFC8174]  Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
              2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
              May 2017, <https://www.rfc-editor.org/info/rfc8174>.

Authors' Addresses

   Bing Liu
   Huawei Technologies
   No. 156 Beiqing Road
   Beijing
   China
   Email: leo.liubing@huawei.com


   Li Zhang
   Huawei Technologies
   No. 156 Beiqing Road
   Beijing
   China
   Email: zhangli344@huawei.com



Liu, et al.               Expires 23 April 2026                 [Page 7]

Internet-Draft         Agents Networking Scenarios          October 2025


   Yuxing Liu
   Independent
   Beijing
   China
   Email: 83772966@qq.com














































Liu, et al.               Expires 23 April 2026                 [Page 8]
