Environment-Aware and 3D Map-Aided GNSS

This program uses 3D city models, LiDAR maps, sky visibility, ray tracing, digital twins, and urban geometry to improve GNSS positioning and reliability assessment in dense cities. The urban environment is not only an obstacle — it can become prior information for signal prediction, NLOS detection, multipath modeling, and integrity-aware positioning.

Why it matters. Smartphone- and vehicle-grade GNSS receivers operating in urban canyons see severe NLOS and multipath errors. Using a known 3D model of the surrounding buildings turns the environment from noise source into measurement aid — and is the foundation of the lab’s longest-running industry partnership.

Funding & industry partners.

• Huawei — 3-year HK$10M collaborative agreement signed July 2020, with 9+ contracts under the framework on 3DMA GNSS, deep integration, 5G indoor positioning, collaborative GNSS, and multi-sensor fusion (since 2018)
• Google Research Award — US$40k, 2024, 3DMA GNSS Precise Positioning for Vehicular Applications (PI)
• Google Android Location — Visiting Research Scientist at Mountain View (20 months, 2022–2023); Pixel team talk Aug 2024

Recognition. TechConnect 2021 Innovation Award for 3D LiDAR Aided GNSS for L4 AV; US12123961B2 patent on 3D LiDAR Aided GNSS NLOS Detection and Correction (Hsu, Wen, granted 2024).

Representative publications. 3D-mapping-aided PPP-RTK aiming at deep urban canyons (Xin et al., Journal of Geodesy 2022); 3D Vision Aided GNSS RTK for Autonomous Systems in Urban Canyons (Wen, Bai, Hsu, NAVIGATION 2023); 3D Mapping Database Aided GNSS RTK (Ng, Hsu, IEEE TAES 2021).