Aircraft need reliable altitude above ground level (AGL) for safe flight, but conventional systems can be constrained by active transmission, weather, coverage and single-point geometry. This field study tests whether reflected Global Navigation Satellite System (GNSS) signals can provide a complementary AGL measurement. Using upward- and downward-looking antennas and raw intermediate-frequency data, the team processed multi-constellation signals with a software-defined receiver. They propose a Short-Time Shortest-Path retracking method that uses short integrations and selects the shortest path delay when reflected waveforms show multiple peaks. The study assesses retrieval availability and accuracy over flat rural, urban/suburban, mountainous and undulating rural terrain.
Radio altimeters remain the primary real-time source of altitude above ground level (AGL), but their range is often limited and accuracy can be affected by aircraft attitude and radio-frequency interference from fifth-generation (5G) systems. Barometric altimeters and standard Global Navigation Satellite System (GNSS) receivers measure pressure-based or ellipsoidal height, not local terrain clearance. Global Navigation Satellite System-Reflectometry (GNSS-R) offers a passive alternative by comparing direct and surface-reflected signals, with multi-point observations and low sensitivity to aircraft attitude. Spaceborne and airborne campaigns show promise over ocean, ice, soil and vegetation, but systematic validation over complex land and dynamic flight remains scarce. Given these challenges, in-depth research is needed on robust GNSS-R AGL retrieval across diverse landforms.
A joint research team led by Professor Qin Weiwei of Rocket Force University of Engineering and Professor Xu Tianhe of Shandong University tested a GNSS-R system for aircraft AGL monitoring. Published (DOI: 10.1186/s43020-026-00228-6) on 27 October 2026 in Satellite Navigation , the study used an Aurora SA60L aircraft on a 288 km flight from Dongying to Tai'an. The team collected raw intermediate-frequency (IF) data from BeiDou Navigation Satellite System (BDS), Galileo, Global Positioning System (GPS) and Quasi-Zenith Satellite System (QZSS) signals and processed them with a software-defined receiver (SDR). It compared GNSS-R-derived AGL with references from GNSS positioning and a Digital Elevation Model (DEM), and proposed a Short-Time Shortest-Path (STSP) retracking method.
The campaign collected about 1.7 hours of raw flight data and produced tens of thousands of valid path-delay observations across 31 satellite signal types. Availability was highest over flat rural areas, followed by undulating rural, urban/suburban and mountainous terrain, showing that surface roughness and signal type strongly shape performance. Wideband civil signals from BDS, Galileo and GPS/QZSS generally outperformed older or narrower-band alternatives, while Galileo E1C was weakest. The central innovation is the STSP method. Rather than heavy averaging that can smooth or distort land-reflected waveforms, it uses short integration windows and, when multiple peaks appear, selects the earliest peak if peaks are close or compares peak amplitudes if the delay is larger. Raising the cutoff elevation angle improved accuracy, especially in mountains, but also reduced the number of usable observations, so the study identifies a practical balance. The reference combines precise aircraft positioning with a DEM and geoid model, allowing the team to isolate retrieval residuals across terrain types. The authors report accuracy better than 4 m over flat terrain and about 15 m in mountainous areas, and find that multipath and surface roughness dominate the remaining error budget.
The authors said the flight campaign moves GNSS-R AGL measurement from proof-of-concept toward practical terrain-aware monitoring. They said the STSP approach is especially useful over land, where reflections from slopes, vegetation and buildings can create multiple peaks that ocean-altimetry methods were not designed to handle. They said wideband civil signals gave the most dependable results, while higher elevation cutoff angles helped reduce multipath and improve accuracy. They said the system remains an offline, proof-of-concept platform, but its passive, low-power design supports future miniaturized and real-time receivers.
GNSS-R could complement radar, barometric and GNSS altimeters by providing passive, multi-point terrain clearance without active transmission. Its low sensitivity to aircraft attitude and use of freely available satellite signals make it attractive for manned aircraft, unmanned aircraft systems (UAS) and low-altitude operations, including terrain-awareness and warning systems, auto-land support and go-around decisions. The study suggests that BDS, Galileo and GPS/QZSS wideband civil signals, combined with a higher elevation cutoff, offer a practical accuracy–availability balance. Before operational use, the team says receivers must become smaller, lighter and capable of real-time processing, for example through graphics processing unit (GPU)-based software-defined receivers. Higher-resolution DEMs and better vegetation and slope modeling could further improve performance.
