Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Transfer to Sky: Unveil Low-Altitude Route-Level Radio Maps via Ground Crowdsourced Data

Wenlihan Lu, Huacong Chen, Ruiyang Duan, Weijie Yuan, Shijian Gao

TL;DR

This work tackles predicting route-level $RSRP$ for UAVs along planned trajectories by leveraging abundant ground crowdsourced measurements. It introduces a three-stage transfer learning framework: (1) pretraining on ray-traced synthetic data to build propagation priors using a dual-transmitter U-Net, (2) adversarial domain adaptation to align simulated and real feature spaces at a bottleneck, and (3) decoder-only finetuning with sparse UAV measurements to calibrate to real propagation. The approach yields substantial improvements over baselines, achieving up to 51.9% accuracy gain on real Meituan UAV data and enabling scalable, proactive route-level connectivity planning in urban environments. Overall, the method bridges the sim-to-real gap for route-level radio maps and provides a practical tool for reliable low-altitude UAV operations in the growing low-altitude economy.

Abstract

The expansion of the low-altitude economy is contingent on reliable cellular connectivity for unmanned aerial vehicles (UAVs). A key challenge in pre-flight planning is predicting communication link quality along proposed and pre-defined routes, a task hampered by sparse measurements that render existing radio map methods ineffective. This paper introduces a transfer learning framework for high-fidelity route-level radio map prediction. Our key insight is to leverage abundant crowdsourced ground signals as auxiliary supervision. To bridge the significant domain gap between ground and aerial data and address spatial sparsity, our framework learns general propagation priors from simulation, performs adversarial alignment of the feature spaces, and is fine-tuned on limited real UAV measurements. Extensive experiments on a real-world dataset from Meituan show that our method achieves over 50% higher accuracy in predicting Route RSRP compared to state-of-the-art baselines.

Transfer to Sky: Unveil Low-Altitude Route-Level Radio Maps via Ground Crowdsourced Data

TL;DR

This work tackles predicting route-level for UAVs along planned trajectories by leveraging abundant ground crowdsourced measurements. It introduces a three-stage transfer learning framework: (1) pretraining on ray-traced synthetic data to build propagation priors using a dual-transmitter U-Net, (2) adversarial domain adaptation to align simulated and real feature spaces at a bottleneck, and (3) decoder-only finetuning with sparse UAV measurements to calibrate to real propagation. The approach yields substantial improvements over baselines, achieving up to 51.9% accuracy gain on real Meituan UAV data and enabling scalable, proactive route-level connectivity planning in urban environments. Overall, the method bridges the sim-to-real gap for route-level radio maps and provides a practical tool for reliable low-altitude UAV operations in the growing low-altitude economy.

Abstract

The expansion of the low-altitude economy is contingent on reliable cellular connectivity for unmanned aerial vehicles (UAVs). A key challenge in pre-flight planning is predicting communication link quality along proposed and pre-defined routes, a task hampered by sparse measurements that render existing radio map methods ineffective. This paper introduces a transfer learning framework for high-fidelity route-level radio map prediction. Our key insight is to leverage abundant crowdsourced ground signals as auxiliary supervision. To bridge the significant domain gap between ground and aerial data and address spatial sparsity, our framework learns general propagation priors from simulation, performs adversarial alignment of the feature spaces, and is fine-tuned on limited real UAV measurements. Extensive experiments on a real-world dataset from Meituan show that our method achieves over 50% higher accuracy in predicting Route RSRP compared to state-of-the-art baselines.
Paper Structure (11 sections, 10 equations, 7 figures, 3 tables)

This paper contains 11 sections, 10 equations, 7 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Problem Statement
  3. Proposed Framework
  4. Pretrain via Synthetic Augment Data
  5. Adversarial Domain Adaptation
  6. Real Data Finetuning
  7. Experiments
  8. Data Augmentation via Raytracing
  9. Performance Comparison
  10. Ablation Studies
  11. Conclusions

Figures (7)

  • Figure 1: Illustration of a single-cell scenario with UAV and ground users. The right figure shows the 3D distribution of RSRP for aerial and ground measurements.
  • Figure 2: An illustration of the proposed framework.
  • Figure 3: U-Net with a shared encoder and attention-enhanced, transmitter-specific decoders for masked radio map recovery.
  • Figure 4: Complementary 3D coverage of two adjacent transmitters and their combined grid representation.
  • Figure 5: Visualization of a subset of the Meituan UAV collected aerial RSRP data along with ground crowdsourced RSRP measurements in an urban area.
  • ...and 2 more figures