Scholarlink

<p>Multipath and diffraction errors significantly impact the accuracy of precise point positioning (PPP), particularly in challenging environments. Traditional mitigation techniques mainly focus on the multipath, whereas the diffraction is often ignored. This paper proposes a Moran-index-driven classification and correction approach that integrates improved multipath hemispherical map (MHM) and sidereal filtering (SF) models to address both error sources. Specifically, the proposed method leverages a relative Moran index consisting of global and local Moran indices to classify multipath-dominated background grids and diffraction-like clustered residual grids according to their spatial clustering characteristics within the MHM framework. The designed edge- and aperture diffraction experiments show that diffraction-related residuals exhibit stronger local spatial autocorrelation than ordinary multipath in the MHM representation, although strongly clustered anomalies are not assumed to arise exclusively from diffraction. A refined MHM is then applied to grids that are suitable for stable spatial averaging, while a denoised window-matching SF model based on cross-correlation methods is used for grids exhibiting stronger local clustering, local irregularity, and poorer suitability for direct MHM averaging. To validate the effectiveness of the proposed method, two 48-h field experiments were conducted in edge and aperture diffraction environments, complemented by an 8-day real monitoring test. The results from both static and kinematic PPP experiments demonstrate significant enhancements in positioning accuracy, convergence efficiency, and overall reliability. Specifically, compared with the uncorrected and traditional methods, the proposed method achieves average 3-dimensional (3D) root mean square error (RMSE) reductions of 48.46% and 27.33% across the two typical diffraction environments. Consistent results are also obtained in real monitoring applications, where the 3D RMSEs are improved by 46.21% and 23.97%, respectively. In addition, both the convergence analysis and residual evaluation confirm the superior reliability of the proposed method compared with the uncorrected and traditional methods. These findings indicate that the proposed method can effectively identify diffraction-dominated or diffraction-like clustered residual regions from the ordinary multipath-dominated background and mitigate the associated errors, providing an accurate and reliable solution for GNSS applications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Classification and mitigation of multipath and diffraction based on Moran-index-driven spatiotemporal corrections

  • 期刊论文

  • 发表日期：

    2026年7月18日

查看全文
  • Zhetao Zhang,
  • Ling Wang,
  • Bofeng Li

摘要

Multipath and diffraction errors significantly impact the accuracy of precise point positioning (PPP), particularly in challenging environments. Traditional mitigation techniques mainly focus on the multipath, whereas the diffraction is often ignored. This paper proposes a Moran-index-driven classification and correction approach that integrates improved multipath hemispherical map (MHM) and sidereal filtering (SF) models to address both error sources. Specifically, the proposed method leverages a relative Moran index consisting of global and local Moran indices to classify multipath-dominated background grids and diffraction-like clustered residual grids according to their spatial clustering characteristics within the MHM framework. The designed edge- and aperture diffraction experiments show that diffraction-related residuals exhibit stronger local spatial autocorrelation than ordinary multipath in the MHM representation, although strongly clustered anomalies are not assumed to arise exclusively from diffraction. A refined MHM is then applied to grids that are suitable for stable spatial averaging, while a denoised window-matching SF model based on cross-correlation methods is used for grids exhibiting stronger local clustering, local irregularity, and poorer suitability for direct MHM averaging. To validate the effectiveness of the proposed method, two 48-h field experiments were conducted in edge and aperture diffraction environments, complemented by an 8-day real monitoring test. The results from both static and kinematic PPP experiments demonstrate significant enhancements in positioning accuracy, convergence efficiency, and overall reliability. Specifically, compared with the uncorrected and traditional methods, the proposed method achieves average 3-dimensional (3D) root mean square error (RMSE) reductions of 48.46% and 27.33% across the two typical diffraction environments. Consistent results are also obtained in real monitoring applications, where the 3D RMSEs are improved by 46.21% and 23.97%, respectively. In addition, both the convergence analysis and residual evaluation confirm the superior reliability of the proposed method compared with the uncorrected and traditional methods. These findings indicate that the proposed method can effectively identify diffraction-dominated or diffraction-like clustered residual regions from the ordinary multipath-dominated background and mitigate the associated errors, providing an accurate and reliable solution for GNSS applications.