Name: Dynamic Multi-agents Secured Collaboration(DMSC)

Description

The DMSC BoF aims to discuss the foreseeable AI agents use cases, overall infrastructure architecture, necessary components in each manageable domain (network/cloud operators) and the interaction among them.

The aims of the DMSC WG are to produce the guideline documents and necessary protocol extensions to build one operable infrastructure for the future of AI agents collaboration, which include how to accomplish the AI agent registration and classification identification, how to discovery the most suitable peer agents, how to intelligent routing the AI agent request, how to keep the secure inter-agent communication，which (if any) security mechanisms and models are required: interoperable agent identity binding, authN/authZ, and a verifiable trust anchor for cross-domain collaboration, and how to keep the sustain high performance and operational visibility.

Required Details

• Status: "WG forming"
• Responsible AD: Andy Newton, Orie Steele
• BOF proponents: wangaj3@chinatelecom.cn, leo.liubing@huawei.com, enge.seg@alibaba-inc.com, liujun@bupt.edu.cn, arashmid.akhavain@huawei.com
• Number of people expected to attend: 150
• Length of session (1 or usually 2 hours): 2 hours
• Conflicts (whole Areas and/or WGs)
• Chair Conflicts: TBD
• Technology Overlap: TBD
• Key Participant Conflict: ART/WIT/SEC AD

Information for IAB/IESG

To allow evaluation of your proposal, please include the following items:

• Any protocols or practices that already exist in this space: A2A/MCP, but they focus only the protocol layer, and there is no document to describe how to build one operable, manageable infrastructure to accomplish the AI agents’ collaboration.
• Which (if any) modifications to existing protocols or practices are required: AI agents’ registrations
• Which (if any) entirely new protocols or practices are required: AI agents’ classification, AI agents’ discovery, AI agents request(semantic) routing, AI agents’ communication operational visibility etc.
• Open source projects (if any) implementing this work: ANP/AIP

Agenda

• Use cases, scenarios and requirements(20min)
  1-2 representative presentations(TBD)

• Internet of Agents Vision(20 min)
  1 presentation(TBD)

• Overall Architecture Consideration(15 min)
  1 presentation(TBD)

• Network Infrastructure Evolution Considerations(15 min)
  1 presentation(TBD)

• Interested DMSC Solution Proposal(10 min)
  1 presentation(TBD)

• DMSC Security Considerations (10 min)

• Charter Discussion(30 min)
  Items, Internet-Drafts, speakers, timing

Links to the mailing list, draft charter if any (for WG-forming BoF), relevant Internet-Drafts, etc.

• Mailing List: dmsc@ietf.org (subscribe the list at: https://mailman3.ietf.org/mailman3/lists/dmsc@ietf.org/)

• Draft charter: https://github.com/ietf-dmsc/Charter/blob/main/dmsc-charter.md

• Relevant Internet-Drafts:

  • Use Cases:

    1. AI Agent Use Cases and Requirements in 6G Network(https://datatracker.ietf.org/doc/draft-yu-dmsc-ai-agent-use-cases-in-6g/)
    2. Network AI Agent Use Cases and Requirements in 6G(https://datatracker.ietf.org/doc/html/draft-tong-network-agent-use-cases-in-6g-00)
       - ... ...

  • Solutions

    3. Dynamic Multi-agents Secured Collaboration Infrastructure architecture(https://datatracker.ietf.org/doc/draft-li-dmsc-inf-architecture/)

    4. Agent Collaboration Protocols Architecture for Internet of Agents https://datatracker.ietf.org/doc/draft-liu-dmsc-acps-arc/

    5. Internet of Agents Task Protocol (IoA Task Protocol) for Heterogeneous Agent Collaboration https://datatracker.ietf.org/doc/draft-yang-dmsc-ioa-task-protocol/

    6. Requirements for Agent Gateway(https://datatracker.ietf.org/doc/draft-liu-dmsc-gw-requirements/)

    7. Security Analysis of Multi-agents Secured Communication and Limitations of Existing Protocols(https://datatracker.ietf.org/doc/draft-zhang-dmsc-mas-communication/)

    8. Multi Agents Discovery in Cloud(Alibaba, to be submitted)

    9. Semantic-Aware Discovery and Invocation for AI Agents(draft-cui-dmsc-ai-agent-discovery-invocation/To be submitted)

    10. Using Natural Language for Universal Coordination in Multi Agent Systems(IBM/Dinesh/To be submitted)