AgenticPrecoding: LLM-Empowered Multi-Agent System for Precoding Optimization

2026-05-07

AI Summary

This work addresses the limited generalizability and adaptability of conventional precoding methods in the heterogeneous and dynamic environments envisioned for 6G. To overcome this challenge, we propose AgenticPrecoding, a multi-agent, end-to-end precoding optimization framework that decomposes the task into four collaborative stages: problem modeling, solver selection, prompt upsampling, and code generation. For the first time, it integrates multi-agent systems with domain-specific precoding knowledge by injecting expert insights through LoRA-finetuned reasoning agents, enabling large language models to automatically derive executable precoding solutions from high-level user requirements. A feedback-driven refinement mechanism is further introduced to enhance solution quality. Extensive experiments across ten representative scenarios demonstrate that our approach significantly outperforms traditional optimization algorithms and existing LLM-based baselines, exhibiting exceptional cross-scenario adaptability.

Abstract

Precoding is a key technique for interference management and performance improvement in multi-antenna wireless systems. However, existing precoding methods are typically developed for specific system models, objectives, and constraint sets, which limits their adaptability to the heterogeneous and evolving scenarios expected in future 6G networks. To address this limitation, we propose AgenticPrecoding, a universal multi-agent framework that automates end-to-end precoding derivation directly from user-level communication requirements. Specifically, AgenticPrecoding decomposes the derivation process into four coordinated stages: problem formulation, solver selection, prompt upsampling, and code generation, assigning each stage to a specialized agent tailored to its specific reasoning demands. We employ two LoRA-adapted reasoning agents to inject precoding-specific domain knowledge for problem formulation and solver selection, while two general-purpose Large Language Models (LLMs) handle prompt refinement and executable code generation. Furthermore, a feedback-driven refinement mechanism is incorporated to enhance code executability, constraint feasibility, and solution quality. Extensive experiments across 10 representative precoding scenarios demonstrate that AgenticPrecoding achieves superior cross-scenario adaptability compared to conventional optimization-based and LLM-based baselines.

Problem

• Precoding
• 6G networks
• Adaptability
• Multi-agent system
• Interference management

Innovation

• multi-agent system
• large language models
• precoding optimization
• LoRA adaptation
• feedback-driven refinement