Abstract
In low-altitude economy (LAE), fast and accurate beam prediction between high-mobility unmanned aerial vehicles (UAVs) and ground base stations is essential for seamless coverage and reliable communications. However, existing deep learning-based beam prediction methods lack high-level semantic understanding of dynamic environments, resulting in poor generalization. Emerging large language model (LLM)-based approaches show promise for generalization but lack rich environmental perception and fine-grained spatial semantics for precise beam alignment. To address these limitations, we propose BeamVLM, an end-to-end generative framework that casts beam prediction as a visual question answering task using existing vision-language models (VLMs). By projecting raw visual patches directly into the language domain and judiciously designing an instructional prompt, the proposed BeamVLM enables the VLM to jointly reason over UAV trajectories and environmental context. Finally, experimental results on real-world datasets demonstrate that the proposed BeamVLM outperforms state-of-the-art methods in prediction accuracy and also exhibits superior generalization for other scenarios such as vehicle-to-infrastructure (V2I) beam prediction.
| Original language | English |
|---|---|
| Journal | IEEE Transactions on Vehicular Technology |
| DOIs | |
| Publication status | Accepted/In press - 2026 |
| Externally published | Yes |
Keywords
- Low-altitude economy
- beam prediction
- unmanned aerial vehicles
- vision-language model
Fingerprint
Dive into the research topics of 'BeamVLM for Low-Altitude Economy: Generative Beam Prediction via Vision-Language Models'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver