Towards End-to-End Network Intent Management with Large Language Models

2025-04-18

AI Summary

This study addresses intent-driven networking (IBN) by investigating large language models (LLMs) for end-to-end translation of natural-language network intents into deployable configurations in 5G/6G mobile networks—including both RAN and core network domains. Methodologically, it integrates proprietary (Gemini 1.5 Pro, GPT-4) and open-weight LLMs (Llama, Mistral), enhanced by domain-specific prompt engineering and rigorous configuration syntax/semantic validation. To holistically assess performance, we propose FEACI—a novel multi-dimensional metric encompassing Format compliance, Explainability, Accuracy, Inference latency, and Cost. Experimental results demonstrate that optimized lightweight open-weight LLMs achieve accuracy and explainability on par with or exceeding proprietary models, while substantially reducing hardware requirements and deployment costs. This validates the feasibility of cost-efficient, trustworthy LLMs for autonomous telecom network operations.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Large Language Models (LLMs) are likely to play a key role in Intent-Based Networking (IBN) as they show remarkable performance in interpreting human language as well as code generation, enabling the translation of high-level intents expressed by humans into low-level network configurations. In this paper, we leverage closed-source language models (i.e., Google Gemini 1.5 pro, ChatGPT-4) and open-source models (i.e., LLama, Mistral) to investigate their capacity to generate E2E network configurations for radio access networks (RANs) and core networks in 5G/6G mobile networks. We introduce a novel performance metrics, known as FEACI, to quantitatively assess the format (F), explainability (E), accuracy (A), cost (C), and inference time (I) of the generated answer; existing general metrics are unable to capture these features. The results of our study demonstrate that open-source models can achieve comparable or even superior translation performance compared with the closed-source models requiring costly hardware setup and not accessible to all users.

Problem

• Translate human intents into network configurations using LLMs
• Assess LLM performance in 5G/6G network configuration generation
• Compare open-source and closed-source LLMs for intent translation

Innovation

• Leverage LLMs for intent-based networking translation
• Introduce FEACI metrics for performance assessment
• Compare open-source and closed-source LLMs in networking