RadioKMoE: Knowledge-Guided Radiomap Estimation with Kolmogorov-Arnold Networks and Mixture-of-Experts

2025-11-21

AI Summary

To address low accuracy and poor robustness of radio map estimation (RME) in complex wireless propagation environments, this paper proposes RadioKMoE, a knowledge-guided framework. Methodologically, it innovatively integrates physical priors with data-driven modeling: a Kolmogorov–Arnold network captures global radio propagation characteristics, while a Mixture-of-Experts (MoE) architecture with gating mechanisms models local signal heterogeneity—jointly optimizing global consistency and fine-grained accuracy. A hierarchical estimation mechanism further enhances generalization by jointly leveraging environmental semantics and coarse-grained coverage maps. Extensive experiments on both multi-band and single-band RME tasks demonstrate that RadioKMoE significantly outperforms state-of-the-art methods, validating its high accuracy, strong robustness under highly dynamic and interference-intensive conditions, and effective cross-band generalization capability.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Radiomap serves as a vital tool for wireless network management and deployment by providing powerful spatial knowledge of signal propagation and coverage. However, increasingly complex radio propagation behavior and surrounding environments pose strong challenges for radiomap estimation (RME). In this work, we propose a knowledge-guided RME framework that integrates Kolmogorov-Arnold Networks (KAN) with Mixture-of-Experts (MoE), namely RadioKMoE. Specifically, we design a KAN module to predict an initial coarse coverage map, leveraging KAN's strength in approximating physics models and global radio propagation patterns. The initial coarse map, together with environmental information, drives our MoE network for precise radiomap estimation. Unlike conventional deep learning models, the MoE module comprises expert networks specializing in distinct radiomap patterns to improve local details while preserving global consistency. Experimental results in both multi- and single-band RME demonstrate the enhanced accuracy and robustness of the proposed RadioKMoE in radiomap estimation.

Problem

• Estimating radiomaps in complex wireless environments
• Improving accuracy of signal propagation predictions
• Enhancing local details while maintaining global consistency

Innovation

• Uses Kolmogorov-Arnold Networks for initial coarse prediction
• Employs Mixture-of-Experts for precise radiomap estimation
• Combines environmental data with specialized expert networks