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

TL;DR

AgenticPrecoding introduces a multi-agent framework leveraging LLMs to automate and adapt precoding optimization for diverse 6G wireless scenarios, enhancing flexibility and performance.

Contribution

It presents a universal, multi-stage LLM-based system that automates precoding derivation from user requirements, improving adaptability over traditional methods.

Findings

Outperforms conventional optimization and LLM baselines across 10 scenarios.

Achieves superior cross-scenario adaptability and solution quality.

Incorporates feedback-driven refinement for executable and feasible code generation.

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.