anima M. Han, Ed. Internet-Draft China Unicom Intended status: Standards Track T. Huang, Ed. Expires: 4 January 2027 CNIC, CAS J. Zhao, Ed. China Unicom 3 July 2026 Problem Statement and Gap Analysis for Autonomic Networking with AI- powered Autonomic Service Agent draft-han-anima-gap-analysis-ai-asa-00 Abstract This document presents a problem statement and gap analysis of the technical challenges faced by the GeneRic Autonomic Signaling Protocol (GRASP) [RFC8990] when it is used, within the Autonomic Networking Infrastructure (ANI) defined by the ANIMA Working Group [RFC8993], to support enhanced Autonomic Service Agents that incorporate Large Language Model (LLM) capabilities, referred to in this document as AI-ASAs. 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 4 January 2027. Copyright Notice Copyright (c) 2026 IETF Trust and the persons identified as the document authors. All rights reserved. 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. Han, et al. Expires 4 January 2027 [Page 1] Internet-Draft Problem Statement and Gap Analysis for A July 2026 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 and Characteristics . . . . . . . . . . . . . . 2 2. Problem Statement and Gap Analysis . . . . . . . . . . . . . 3 2.1. Insufficient Payload Capacity for Large-Scale Semantic Data Transfer . . . . . . . . . . . . . . . . . . . . . . . . 3 2.2. Objective Semantics Are Not Flexible . . . . . . . . . . 3 3. Potential Areas for Future Work . . . . . . . . . . . . . . . 3 3.1. Large-Scale Information Transfer . . . . . . . . . . . . 4 3.2. Self-Describing Metadata for Objectives . . . . . . . . . 4 4. Security Considerations . . . . . . . . . . . . . . . . . . . 4 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 4 6. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 4 7. References . . . . . . . . . . . . . . . . . . . . . . . . . 4 7.1. Normative References . . . . . . . . . . . . . . . . . . 4 7.2. Informative References . . . . . . . . . . . . . . . . . 5 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 5 1. Introduction and Characteristics Within the Autonomic Networking Infrastructure (ANI) framework defined by [RFC8993], the Autonomic Control Plane (ACP) [RFC8994], and the GeneRic Autonomic Signaling Protocol (GRASP) [RFC8990], the Autonomic Service Agent (ASA) is a core logical component running on an autonomic node. As embedded AI and edge inference capabilities are introduced into the network control plane, enhanced ASAs with LLM-based reasoning capabilities, referred to as AI-powered Autonomic Service Agent (AI- ASAs) [I-D.han-anima-ai-asa], are beginning to emerge. Communication between these agents has peer-to-peer characteristics that differ from those of traditional hard-coded network components, and this creates new workload requirements for the existing signaling plane: * Large information transfers: Signaling exchanges may need to carry variable-length context sequences, such as reasoning traces, prompt variables, or abstract intent descriptions. A single Objective payload may reach several megabytes. * Objective-oriented interaction: The information exchanged and negotiated between nodes is no longer limited to traditional structured resource parameters, such as the IPv6 prefix length Han, et al. Expires 4 January 2027 [Page 2] Internet-Draft Problem Statement and Gap Analysis for A July 2026 defined in [RFC8992]. It may instead consist of abstract semantic intents describing high-level network Objectives, or natural- language interactions. 2. Problem Statement and Gap Analysis 2.1. Insufficient Payload Capacity for Large-Scale Semantic Data Transfer Problem: In current GRASP implementations and deployments, Objective payloads are usually tens or hundreds of bytes in size. The protocol was not considered for contexts in which payloads may reach the megabyte level. In practical AI-ASA collaboration, the amount of data exchanged may increase sharply, while the current mechanism does not provide application-layer support for advertising data size, negotiating a receive window, or transferring data in chunks. This may cause memory overflow at the receiver and may also lead to head- of-line blocking within the ACP. Gap:The data transfer pattern also needs to be distinguished. AI- ASAs may use short request-response exchanges, but they may also share information continuously, for example by streaming real-time inference-state updates. The existing bulk-transfer draft [I-D.carpenter-anima-grasp-bulk] focuses mainly on one-time transfers of large data blocks. Continuous, low-latency delivery in streaming scenarios therefore requires further discussion. 2.2. Objective Semantics Are Not Flexible Problem: Each Objective must also be registered in advance. This static model cannot meet the need for AI-ASAs to generate new Objectives dynamically or to describe the meaning of an Objective by themselves. It therefore limits flexible and open semantic interaction between LLM-based agents. Gap: The current GRASP specification relies on strictly predefined CBOR encodings and fixed semantic mappings for Objectives, as illustrated by [RFC8992]. 3. Potential Areas for Future Work To minimize the development and deployment cost for the working group and existing networks, while maintaining the stability of the current ANI control plane, future work could examine backward-compatible extensions at the application layer. Such work should not change the state-transition logic of the GRASP core state machine, modify the underlying ACP transport protocol, or define how specific intent semantics are interpreted. Han, et al. Expires 4 January 2027 [Page 3] Internet-Draft Problem Statement and Gap Analysis for A July 2026 3.1. Large-Scale Information Transfer This issue is related to the ANIMA Working Group draft on bulk data transfer [I-D.carpenter-anima-grasp-bulk]. That draft proposes a general method for transferring bulk data by using GRASP itself and is intended for cases in which the data is too large to fit in a normal GRASP message. It is also related to the large-scale information delivery requirements discussed in [I-D.ietf-anima-grasp-distribution]. 3.2. Self-Describing Metadata for Objectives Option 1: Add optional self-describing metadata to an Objective. The sender could include machine-readable type information, such as objective-type and semantic-version, together with the Objective, allowing the receiving AI-ASA to identify and parse a new Objective that has not been predefined. Option 2: One approach is to define a generic Objective and allow its content to be generated by AI based on the existing Objective format. Another approach is to allow AI agents to communicate with each other in natural language. At the gap-analysis stage, both approaches, as well as other possible approaches, can be discussed. 4. Security Considerations TBD. 5. IANA Considerations TBD. 6. Acknowledgements TBD 7. References 7.1. Normative References [RFC8990] Bormann, C., Carpenter, B., Ed., and B. Liu, Ed., "GeneRic Autonomic Signaling Protocol (GRASP)", RFC 8990, DOI 10.17487/RFC8990, May 2021, . Han, et al. Expires 4 January 2027 [Page 4] Internet-Draft Problem Statement and Gap Analysis for A July 2026 [RFC8993] Behringer, M., Ed., Carpenter, B., Eckert, T., Ciavaglia, L., and J. Nobre, "A Reference Model for Autonomic Networking", RFC 8993, DOI 10.17487/RFC8993, May 2021, . [RFC8494] Wilson, D. and A. Melnikov, Ed., "Multicast Email (MULE) over Allied Communications Publication (ACP) 142", RFC 8494, DOI 10.17487/RFC8494, November 2018, . 7.2. Informative References [RFC8992] Jiang, S., Ed., Du, Z., Carpenter, B., and Q. Sun, "Autonomic IPv6 Edge Prefix Management in Large-Scale Networks", RFC 8992, DOI 10.17487/RFC8992, May 2021, . [I-D.carpenter-anima-grasp-bulk] Carpenter, B. E., Jiang, S., and B. Liu, "Transferring Bulk Data over the GeneRic Autonomic Signaling Protocol (GRASP)", Work in Progress, Internet-Draft, draft- carpenter-anima-grasp-bulk-05, 9 January 2020, . [I-D.ietf-anima-grasp-distribution] Jiang, S., Liu, B., Xiao, X., Hecker, A., and X. Zheng, "Information Distribution over GRASP", Work in Progress, Internet-Draft, draft-ietf-anima-grasp-distribution-12, 11 December 2024, . [I-D.han-anima-ai-asa] Han, M., Zhang, N., and J. Zhao, "Considerations of AI- powered Autonomic Service Agent Communication", Work in Progress, Internet-Draft, draft-han-anima-ai-asa-01, 15 January 2026, . Authors' Addresses Mengyao Han (editor) China Unicom Beijing China Email: hanmy12@chinaunicom.cn Han, et al. Expires 4 January 2027 [Page 5] Internet-Draft Problem Statement and Gap Analysis for A July 2026 Tianyi Huang (editor) CNIC, CAS Beijing China Email: tyhuang@cnic.cn Jing Zhao (editor) China Unicom Beijing China Email: zhaoj501@chinaunicom.cn Han, et al. Expires 4 January 2027 [Page 6]