1.广东工业大学自动化学院,广东 广州,510006
2.广东省智能系统与优化集成重点实验室,广东 广州,510006
3.粤港澳离散制造智能化联合实验室,广东 广州,510006
4.物联网智能信息处理与系统集成教育部重点实验室,广东 广州,510006
5.深空探测实验室,北京,100084
曾昆赣,博士,副教授,研究方向为人工智能、卫星导航定位。zengkungan@gmail.com
李珍妮,博士,教授。lizhenni2012@gmail.com
收稿:2026-04-01,
纸质出版:2026-07-05
移动端阅览
曾昆赣,唐健浩,李珍妮,等.面向动态环境表征的图Transformer GNSS非视距信号检测方法[J].武汉大学学报(信息科学版),2026,51(7):1299-1312.
ZENG Kungan,TANG Jianhao,LI Zhenni,et al.Graph Transformer for GNSS NLOS Signal Detection via Dynamic Environmental Representation[J].Geomatics and Information Science of Wuhan University,2026,51(7):1299-1312.
曾昆赣,唐健浩,李珍妮,等.面向动态环境表征的图Transformer GNSS非视距信号检测方法[J].武汉大学学报(信息科学版),2026,51(7):1299-1312. DOI:10.13203/j.whugis20260106
ZENG Kungan,TANG Jianhao,LI Zhenni,et al.Graph Transformer for GNSS NLOS Signal Detection via Dynamic Environmental Representation[J].Geomatics and Information Science of Wuhan University,2026,51(7):1299-1312. DOI:10.13203/j.whugis20260106
在低空经济快速发展的背景下,无人机在城市复杂环境中的物流配送、应急救援及巡检等任务对其连续、高精度的定位能力提出了严峻挑战。在高楼密集的城市峡谷环境中,全球导航卫星系统遭受严重的非视距(non-line-of-sight,NLOS)信号干扰,易导致定位误差增大甚至跳变,威胁无人机飞行安全。现有的深度神经网络检测方法在处理该问题时仍面临两大挑战:一是输入通常需固定维度,对可变数量观测值的零填充会引入非物理意义的伪观测;二是模型在不同时空场景下的泛化能力不足,难以适应动态变化的低空环境。为此,提出一种融合边特征编码与稀疏适配器的图Transformer检测模型。首先,构建天空卫星图,将历元内的卫星观测数据转化为图结构输入,在灵活表征卫星间几何拓扑关系的同时,避免了强制对齐维度带来的数据失真。其次,设计了一种节点-边联合注意力机制(graph node-edge attention,GNE-Attention),通过边特征动态调节卫星间的相似度评分,实现对环境特征的自适应提取。此外,设计了一个稀疏适配器,使卫星能够从GNE-Attention的输出中筛选关键的环境表征信息,从而显著增强了模型在异构场景和多径环境下的检测泛化能力。在真实城市峡谷数据集上的实验结果表明,所提方法在NLOS信号检测任务中的准确率超过93%,并且相较于各类现有神经网络方法,准确率提升约3%~20%。跨时段与跨区域的测试进一步验证了该方法在复杂环境中的稳健性,为无人机安全飞行提供了可靠的定位技术支撑。
Objectives
2
In the context of the rapidly developing low-altitude economy
unmanned aerial vehicle (UAV) are widely deployed for tasks such as urban logistics
emergency response
and infrastructure inspection
which impose stringent requirements for continuous and high-precision positioning in complex urban environments. In dense urban canyons
global navigation satellite system (GNSS) signals suffer from severe non-line-of-sight (NLOS) interference due to low flight altitudes and restricted sky visibility. This frequently leads to increased positioning errors and abrupt positioning jumps
threatening the safety and reliability of UAV operations. Although deep neural network-based methods have achieved progress in NLOS detection
two critical challenges remain. First
existing approaches typically require fixed-dimension inputs
necessitating zero-padding for variable numbers of satellite observations
which introduces non-physical pseudo-observations. Second
these models exhibit insufficient generalization across different spatiotemporal scenarios
making it difficult to adapt to the rapidly changing low-altitude environments encountered during UAV flight.
Methods
2
An edge-feature-encoded graph Transformer with sparse adapter (EGTSA) model is proposed to address the generalization challenges of GNSS NLOS detection. To avoid the distortion caused by forced dimension alignment
satellites observed within the same epoch are organized into a sky satellite graph. This allows variable-length GNSS measurements to be naturally represented through graph structures while flexibly characterizing inter-satellite geometric topological relationships. To enhance environmental representation learning
a graph node-edge attention (GNE-Attention) mechanism is introduced. By utilizing edge features to dynamically adjust inter-satellite similarity scores
the model can adaptively extract discriminative environmental features. Furthermore
a sparse adapter is designed to enable satellites to selectively attend to key environmental representations from the GNE-Attention output
thereby improving generalization in dynamic multipath environments.
Results
2
Extensive experiments were conducted using real-world GNSS datasets collected in complex urban canyon environments at different times and locations. The experimental results demonstrate that the proposed EGTSA model achieves NLOS detection accuracies exceeding 93%
representing an improvement of approximately 3% to 20% compared with existing state-of-the-art methods. Furthermore
cross-time and cross-location tests verify the robustness of the method in complex low-altitude urban environments
demonstrating its strong generalization ability.
Conclusions
2
A graph Transformer-based NLOS detection framework is proposed to address the issues of variable-length satellite measurements and poor cross-domain generalization. By modeling satellites as graph nodes and explicitly encoding geometric relationships via GNE-Attention
the proposed method preserves the intrinsic structure of GNSS observations and achieves adaptive environmental feature extraction. The superior detection accuracy and strong generalization performance across different urban environments
times
and locations provides reliable technical support for the stable and safe GNSS positioning of low-altitude UAVs.
李博峰 , 陈广鄂 . GNSS/INS组合车辆协同精密定位方法 [J]. 测绘学报 , 2022 , 51 ( 8 ): 1708 - 1716 .
LI Bofeng , CHEN Guang'e . Precise Cooperative Positioning for Vehicles with GNSS and INS Integration [J]. Acta Geodaetica et Cartographica Sinica , 2022 , 51 ( 8 ): 1708 - 1716 .
陈秀德 , 刘惠 , 蔚保国 , 等 . 北斗/GNSS广域精密定位技术与服务: 现状与展望 [J]. 武汉大学学报(信息科学版) , 2025 , 50 ( 3 ): 413 - 429 .
CHEN Xiude , LIU Hui , YU Baoguo , et al . BeiDou/GNSS Wide-Area Precise Positioning Technology and Service: Current Situation and Prospects [J]. Geomatics and Information Science of Wuhan University , 2025 , 50 ( 3 ): 413 - 429 .
ZENG K G , LI Z N , ZHAO H L , et al . A Spatiotemporal Information-Driven Cross-Attention Model with Sparse Representation for GNSS NLOS Signal Classification [J]. IEEE Internet of Things Journal , 2024 , 11 ( 19 ): 31892 - 31908 .
ZHAO H L , LI Z N , WANG Q M , et al . Improving Performances of GNSS Positioning Correction Using Multiview Deep Reinforcement Learning with Sparse Representation [J]. GPS Solutions , 2024 , 28 ( 3 ): 98 .
ZHANG G H , HSU L T . Intelligent GNSS/INS Integrated Navigation System for a Commercial UAV Flight Control System [J]. Aerospace Science and Technology , 2018 , 80 : 368 - 380 .
李新忠 , 熊永良 , 徐韶光 . 基于正则化与恒星日滤波的BDS多路径误差削减方法 [J]. 大地测量与地球动力学 , 2024 , 44 ( 2 ): 116 - 121 .
LI Xinzhong , XIONG Yongliang , XU Shaoguang . BDS Multipath Reduction Method Based on Regularization and Sidereal Filtering [J]. Journal of Geodesy and Geodynamics , 2024 , 44 ( 2 ): 116 - 121 .
杨宝山 . 利用恒星日滤波削减油田井架BDS多路径效应 [J]. 海洋测绘 , 2023 , 43 ( 5 ): 46 - 51 .
YANG Baoshan . Sidereal Filter Method to Reduce Multipath Effect of Oilfield Derrick BDS Carrier Phase [J]. Hydrographic Surveying and Charting , 2023 , 43 ( 5 ): 46 - 51 .
REN H X , LI G C , GENG J H , et al . Multipath Hemispherical Map Model with Geographic Cut-off Elevation Constraints for Real-Time GNSS Monitoring in Complex Environments [J]. GPS Solutions , 2023 , 27 ( 4 ): 188 .
屈利忠 , 郑冰雁 , 李竟超 , 等 . 利用多路径半天球图模型削弱GPS三频PPP时间传递中的多路径效应 [J]. 测绘通报 , 2025 ( 10 ): 82 - 86 .
QU Lizhong , ZHENG Bingyan , LI Jingchao , et al . Using a Multipath Hemispherical Map Model to Mitigate Multipatheffects in GPS Triple-Frequency PPP Time Transfer [J]. Bulletin of Surveying and Mapping , 2025 ( 10 ): 82 - 86 .
XIN S M , GENG J H , ZHANG G H , et al . 3D-Mapping-Aided PPP-RTK Aiming at Deep Urban Canyons [J]. Journal of Geodesy , 2022 , 96 ( 10 ): 78 .
ZHANG G H , NG H F , WEN W S , et al . 3D Mapping Database Aided GNSS Based Collaborative Positioning Using Factor Graph Optimization [J]. IEEE Transactions on Intelligent Transportation Systems , 2021 , 22 ( 10 ): 6175 - 6187 .
SUZUKI T , AMANO Y . NLOS Multipath Classification of GNSS Signal Correlation Output Using Machine Learning [J]. Sensors , 2021 , 21 ( 7 ): 2503 .
QIN Y H , LI Z N , XIE S L , et al . BDS Multipath Signal Classification Using Support Vector Machine [C]// International Conference on Industrial Artificial Intelligence , Shenyang, China , 2022 .
ZHANG G H , XU P H , XU H S , et al . Prediction on the Urban GNSS Measurement Uncertainty Based on Deep Learning Networks with Long Short-Term Memory [J]. IEEE Sensors Journal , 2021 , 21 ( 18 ): 20563 - 20577 .
LI Z N , ZENG K G , WANG L J , et al . NLOS/LOS Classification by Constructing Indirect Environment Interaction from GNSS Receiver Measurements Using a Transformer-Based Deep Learning Model [C]// International Technical Meeting of The Institute of Navigation , Long Beach, California, USA, 2023 .
SUN Y , FU L . Stacking Ensemble Learning for Non-Line-of-Sight Detection of Global Navigation Satellite System [J]. IEEE Transactions on Instrumentation and Measurement , 2022 , 71 : 3512510 .
MOHANTY A , GAO G . Learning GNSS Positioning Corrections for Smartphones Using Graph Convolution Neural Networks [J]. NAVIGATION: Journal of the Institute of Navigation , 2023 , 70 ( 4 ): navi. 622 .
ZHENG S L , ZENG K G , LI Z N , et al . Improving the Prediction of GNSS Satellite Visibility in Urban Canyons Based on a Graph Transformer [J]. NAVIGATION: Journal of the Institute of Navigation , 2024 , 71 ( 4 ): navi. 676 .
ZHOU R Y , HU Z G , ZHAO Q L , et al . Absolute Field Calibration of Receiver Antenna Phase Center Models for GPS/BDS-3 Signals [J]. Journal of Geodesy , 2023 , 97 ( 9 ): 83 .
HU Y F , WANG J , LI Z H , et al . Ground Surface Elevation Changes over Permafrost Areas Revealed by Multiple GNSS Interferometric Reflectometry [J]. Journal of Geodesy , 2022 , 96 ( 8 ): 56 .
TAO X L , LIU W K , WANG Y Z , et al . Smartphone RTK Positioning with Multi-Frequency and Multi-Constellation Raw Observations: GPS L1/L5, Galileo E1/E5a, BDS B1I/B1C/B2a [J]. Journal of Geodesy , 2023 , 97 ( 5 ): 43 .
ZHU N , BETAILLE D , MARAIS J , et al . GNSS Integrity Monitoring Schemes for Terrestrial Applications in Harsh Signal Environments [J]. IEEE Intelligent Transportation Systems Magazine , 2020 , 12 ( 3 ): 81 - 91 .
KIM S , BYUN J , PARK K . Machine Learning-Based GPS Multipath Detection Method Using Dual Antennas [C]// Asian Control Conference , Jeju, Korea , 2022 .
WANG Y , XU J W , YANG R , et al . GNSS Multipath Detection Based on Decision Tree Algorithm in Urban Canyons [C]// China Satellite Navigation Conference , Nanchang, China , 2021 .
LI X X , LI X , LIU G G , et al . Triple-Frequency PPP Ambiguity Resolution with Multi-constellation GNSS: BDS and Galileo [J]. Journal of Geodesy , 2019 , 93 ( 8 ): 1105 - 1122 .
GROVES P D , JIANG Z , SKELTON B , et al . Novel Multipath Mitigation Methods Using a Dual-Polarization Antenna [C]// International Technical Meeting of the Satellite Division of the Institute of Navigation , Portland, USA , 2010 .
HENDRYCKS D , GIMPEL K . Gaussian Error Linear Units (GELUs) [PP/OL]. V1. arXiv ( 2016-06-27 )[ 2026-01-07 ]. https://arxiv.org/abs/1606.08415 https://arxiv.org/abs/1606.08415 .
KINGA D , ADAM J B . A Method for Stochastic Optimization [C]// International Conference on Learning Representations , San Diego, USA , 2015 .
VARSHNEY M , SINGH P . Optimizing Nonlinear Activation Function for Convolutional Neural Networks [J]. Signal, Image and Video Processing , 2021 , 15 ( 6 ): 1323 - 1330 .
IOFFE S , SZEGEDY C . Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift [C]// International Conference on Machine Learning , Lille, France , 2015 .
MURPHY K , SCHÖLKOPF B , SRIVASTAVA N , et al . Dropout: A Simple Way to Prevent Neural Networks from Overfitting [J]. Journal of Machine Learning Research , 2014 , 15 ( 1 ): 1929 - 1958 .
MIN D C , KIM M , LEE J , et al . DNN-Based Approach to Mitigate Multipath Errors of Differential GNSS Reference Stations [J]. IEEE Transactions on Intelligent Transportation Systems , 2022 , 23 ( 12 ): 25047 - 25053 .
MOHANTY A , GAO G . Learning GNSS Positioning Corrections for Smartphones Using Graph Convolution Neural Networks [J]. NAVIGATION: Journal of the Institute of Navigation , 2023 , 70 ( 4 ): navi. 622 .
HSU L T , TOKURA H , KUBO N , et al . Multiple Faulty GNSS Measurement Exclusion Based on Consistency Check in Urban Canyons [J]. IEEE Sensors Journal , 2017 , 17 ( 6 ): 1909 - 1917 .
0
浏览量
1
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621