Intent2Tc: Automated Intent-to-Traffic Control Translation with Language Models

2026-09-25

AI Summary

This study addresses the complexity and error-proneness of translating network intents into traffic control policies by proposing a closed-loop language model framework that automatically converts high-level traffic shaping intents into verified Linux tc configurations. Methodologically, the framework integrates an active queue management digital twin, critique-driven optimization, and retrieval-augmented generation (RAG) to coordinate large and small language models for knowledge reuse and policy iteration. Experimental results demonstrate that the framework achieves high semantic fidelity and deployment readiness, with Claude Sonnet 4.6 attaining a semantic similarity score of 0.98. Furthermore, RAG effectively reduces inference latency, significantly enhancing both the semantic consistency and reliability of configuration generation.

Abstract

Automated and highly usable Quality-of-Service (QoS) enforcement requires translating high-level service intents into deployable traffic-management policies. Although intent-based networking (IBN) has simplified policy specification, bridging the gap between business-level intents and executable network configurations remains complex, error-prone, and difficult to automate. This paper presents Intent2Tc, a closed-loop language-model-driven framework that translates business-level traffic-shaping intents into declarative sub-intents and subsequently into validated, executable Linux traffic control (tc) configurations. The framework integrates an Active Queue Management (AQM)-based digital twin (DT) semantic model, automated metadata extraction, critique-driven refinement, and Retrieval-Augmented Generation (RAG)-based knowledge reuse to improve semantic consistency and configuration reliability. We evaluate multiple open-source large language models (LLMs) and small language models (SLMs), together with Claude Sonnet-4.6, on 100 Request for Comments (RFC) 9315-compliant traffic-shaping intents. Across both translation stages, Intent2Tc achieves high semantic fidelity, configuration accuracy, and deployment readiness, with Claude Sonnet-4.6 reaching 0.98 semantic similarity, 1.0 semantic unit coverage, and 0.045 normalized edit distance. Furthermore, RAG reduces token consumption and inference latency while enabling compact models such as Phi-4-mini to approach the performance of substantially larger models. Linux tc serves as the target configuration platform, demonstrating the practical applicability of the proposed framework.

Problem

• Intent-based networking
• Quality-of-Service
• traffic control
• intent translation
• network configuration

Innovation

• Intent-based Networking
• Large Language Models
• Retrieval-Augmented Generation
• Digital Twin
• Traffic Control