Planning with Language and Generative Models: Toward General Reward-Guided Wireless Network Design

TL;DR

This paper explores the use of diffusion-based generative models guided by a unified reward function for efficient and scalable indoor access point deployment in complex wireless network environments, outperforming large language models and other approaches.

Contribution

It introduces a diffusion sampling approach guided by a unified reward function for AP planning, demonstrating improved scalability and performance over existing methods.

Findings

Diffusion samplers outperform alternative generative approaches.

The approach effectively handles non-convex and fragmented objectives.

The method generalizes well across diverse floorplans.

Abstract

Intelligent access point (AP) deployment remains challenging in next-generation wireless networks due to complex indoor geometries and signal propagation. We firstly benchmark general-purpose large language models (LLMs) as agentic optimizers for AP planning and find that, despite strong wireless domain knowledge, their dependence on external verifiers results in high computational costs and limited scalability. Motivated by these limitations, we study generative inference models guided by a unified reward function capturing core AP deployment objectives across diverse floorplans. We show that diffusion samplers consistently outperform alternative generative approaches. The diffusion process progressively improves sampling by smoothing and sharpening the reward landscape, rather than relying on iterative refinement, which is effective for non-convex and fragmented objectives. Finally, we introduce a large-scale real-world dataset for indoor AP deployment, requiring over $50k$ CPU hours to train general reward functions, and evaluate in- and out-of-distribution generalization and robustness. Our results suggest that diffusion-based generative inference with a unified reward function provides a scalable and domain-agnostic foundation for indoor AP deployment planning.