RMTransformer: Accurate Radio Map Construction and Coverage Prediction

TL;DR

This paper introduces RMTransformer, a hybrid transformer-convolution model that significantly improves radio map prediction accuracy, reducing RMSE by over 30% compared to existing methods, thereby enhancing wireless network modeling.

Contribution

The paper presents a novel hybrid transformer-convolution architecture for radio map prediction, combining multi-scale transformer encoding with convolution decoding for better accuracy.

Findings

Over 30% reduction in RMSE compared to SOTA methods

Enhanced prediction accuracy for radio maps

Effective feature extraction with transformer components

Abstract

Radio map, or pathloss map prediction, is a crucial method for wireless network modeling and management. By leveraging deep learning to construct pathloss patterns from geographical maps, an accurate digital replica of the transmission environment could be established with less computational overhead and lower prediction error compared to traditional model-driven techniques. While existing state-of-the-art (SOTA) methods predominantly rely on convolutional architectures, this paper introduces a hybrid transformer-convolution model, termed RMTransformer, to enhance the accuracy of radio map prediction. The proposed model features a multi-scale transformer-based encoder for efficient feature extraction and a convolution-based decoder for precise pixel-level image reconstruction. Simulation results demonstrate that the proposed scheme significantly improves prediction accuracy, and over a 30% reduction in root mean square error (RMSE) is achieved compared to typical SOTA approaches.