Interface on demand: Towards AI native Control interfaces for 6G

2025-08-21

AI Summary

Traditional standardized interfaces in 6G networks suffer from vendor lock-in, rigid design, and poor adaptability. To address these fundamental limitations, this paper proposes an AI-native dynamic control interface generation method. Our approach employs a large language model (LLM)-based multi-agent framework comprising a functional matching agent and a code generation agent, enabling semantic alignment and automatic API construction across heterogeneous network functions (e.g., gNB and WLAN AP) from different vendors. Crucially, we pioneer the integration of LLM-based reasoning into a closed-loop interface generation pipeline, breaking away from static, pre-defined protocol paradigms. Evaluated in a multi-vendor simulation environment, the framework demonstrates on-demand, low-latency generation of interoperable control interfaces for heterogeneous devices. It achieves an effective trade-off among interface correctness, generation efficiency, and model inference cost—significantly enhancing 6G network interoperability, dynamic adaptability, and deployment flexibility.

Technology Category

Natural Language Processing: GenerationHumans and AI: Intelligent User InterfacesMachine Learning: Large Multimodal Models (LMMs)

Application Category

Social Networks and Social Media: Generative AI / large language models and their impact on social systemsSystems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applicationsEconomics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions

Abstract

Traditional standardized network interfaces face significant limitations, including vendor-specific incompatibilities, rigid design assumptions, and lack of adaptability for new functionalities. We propose a multi-agent framework leveraging large language models (LLMs) to generate control interfaces on demand between network functions (NFs). This includes a matching agent, which aligns required control functionalities with NF capabilities, and a code-generation agent, which generates the necessary API server for interface realization. We validate our approach using simulated multi-vendor gNB and WLAN AP environments. The performance evaluations highlight the trade-offs between cost and latency across LLMs for interface generation tasks. Our work sets the foundation for AI-native dynamic control interface generation, paving the way for enhanced interoperability and adaptability in future mobile networks.

Problem

• Addressing vendor incompatibilities in network interfaces
• Generating adaptive control interfaces using LLMs
• Enhancing interoperability in multi-vendor network environments

Innovation

• Multi-agent framework using LLMs for interfaces
• Matching agent aligns NF capabilities with requirements
• Code-generation agent creates API servers dynamically