Dual-Frequency Kinematic PPP in Free-Multipath Open-Sky: CSRS-PPP vs. PPP-WIZARD Vs. APPS

Abstract

This study assesses the performance of kinematic Precise Point Positioning (PPP) using three freely available online services: the Canadian Spatial Reference System Precise Point Positioning (CSRS-PPP), PPP-WIZARD, and the Automatic Precise Positioning Service (APPS). The evaluation was carried out under open-sky conditions to minimize multipath effects and considered two processing configurations, namely GPS-alone and GPS+GLONASS. Dual-frequency GNSS observations were collected at 10 spatially well-distributed reference points across Libya, with each station observed continuously for 24 hours and processed in kinematic mode. The performance evaluation relied on four statistical indicators: Average Absolute Error (AAE), Maximum Absolute Error (MAE), Root Mean Square Error (RMSE), and the percentage of outliers exceeding 5 cm in the Easting (E) and Northing (N) components and 10 cm in the Height (H) component. The results indicate that the integration of GLONASS with GPS generally leads to improvements in GPS-only kinematic PPP solutions and a marked reduction in the percentage of gross errors. These improvements are slight in case of CSRS-PPP and considerable with PPP-WIZARD. Among the evaluated services, CSRS-PPP demonstrates the best overall performance, achieving 0% outliers, AAE and RMSE values below 2 cm in all coordinate components, and 3D-MAE values of approximately 10 cm and 7 cm for GPS-alone and GPS+GLONASS processing, respectively. PPP-WIZARD ranks second, with the multi-constellation solution reducing RMSE values in the E, N, and H components by approximately 9 cm, 15 cm, and 42 cm, respectively, and decreasing the proportion of outliers by about 25%, 7.5%, and 11.5%, in the same order. In addition, the AAE and MAE of kinematic PPP-WIZARD solutions decrease significantly in the E, N, and H components by approximately 1 m, 2.2 m, and 2.8 m, respectively, when integrated constellations are employed. APPS, which supports GPS observations only, exhibits the lowest accuracy and the highest proportion of outliers, even when compared with GPS-only solutions from CSRS-PPP and PPP-WIZARD. Specifically, APPS produces 3D-AAE, 3D-MAE, and 3D-RMSE values of approximately 0.65 m, 3.5 m, and 0.63 m, respectively, with outlier percentages of 64% in E, 38% in N, and 45% in H. Overall, the findings clearly demonstrate the significant advantages of multi-constellation GNSS processing over single-system approaches, particularly in terms of increased observation redundancy, improved satellite geometry, and enhanced capability to mitigate weak or unreliable measurements. The comparative analysis confirms that CSRS-PPP delivers the most reliable performance under both GPS-alone and GPS+GLONASS configurations, with a slight advantage observed for the latter. PPP-WIZARD provides the second-best performance, achieving acceptable accuracy under multi-constellation processing and thus meeting the requirements of certain engineering applications that demand decimeter-level precision. In contrast, the GPS-only PPP-WIZARD and APPS kinematic solutions remain unstable, less precise, and characterized by a high proportion of gross errors.

Introduction

This study investigates the performance of kinematic Precise Point Positioning (PPP) using GPS-only and combined GPS+GLONASS observations, with a focus on evaluating three free online PPP services: CSRS-PPP, PPP-WIZARD, and APPS. GNSS technology enables global positioning with varying accuracy levels, where code-based positioning offers meter-level accuracy, while advanced carrier-phase techniques such as DGNSS and PPP achieve centimetre-level precision. Unlike DGNSS, PPP does not require nearby reference stations and relies on precise satellite orbit and clock products.

Using ten dual-frequency GNSS datasets collected across Libya under open-sky conditions, the study first validates static PPP solutions against official reference coordinates to identify systematic errors. The same datasets are then processed in kinematic mode, and performance is assessed using statistical indicators including AAE, RMSE, MAE, and outlier percentages. The primary objective is to quantify the benefits of integrating GLONASS observations with GPS for kinematic PPP, particularly in reducing outliers and improving solution robustness.

Results show that CSRS-PPP delivers consistently high-quality kinematic solutions at the centimetre level for both GPS-only and GPS+GLONASS processing, with zero outliers and modest improvements (13–21%) when GLONASS is included. PPP-WIZARD exhibits poor and unreliable performance in GPS-only mode, but demonstrates substantial improvement with GPS+GLONASS integration, reducing errors from the metre to the decimetre–centimetre level and significantly lowering outlier rates. In contrast, APPS (GPS-only) produces unstable kinematic solutions with large errors and a high proportion of outliers, making it unsuitable for precise kinematic applications.

Overall, the study confirms that integrating GLONASS with GPS significantly enhances kinematic PPP accuracy and reliability by increasing satellite availability and improving geometry, leading to reduced dilution of precision and better atmospheric modelling. The findings highlight the superiority of multi-constellation PPP, particularly when using CSRS-PPP, and provide a robust foundation for future research on PPP performance in challenging environments, drone-based surveying, and integrated navigation systems.

Conclusion

This study evaluated the performance of kinematic Precise Point Positioning (PPP) using three free online services, namely: the Canadian Spatial Reference System Precise Point Positioning (CSRS-PPP), PPP-WIZARD, and the Automatic Precise Positioning Service (APPS). The evaluation was conducted under open-sky conditions to avoid multipath effects and under two processing configurations: GPS-alone and GPS+GLONASS. Dual-frequency GNSS observations were collected at 10 well-distributed reference points across Libya, with each site observed continuously for 24 hours and processed in kinematic mode. The performance assessment was based on four statistical indicators: Average Absolute Error (AAE), Maximum Absolute Error (MAE), Root Mean Square Error (RMSE), and the percentage of outliers exceeding 5 cm in the Easting (E) and Northing (N) components and 10 cm in the Height (H) component. The results show that, in general, integrating GLONASS with GPS improves the quality of GPS-alone kinematic PPP solutions and significantly reduces the percentage of gross errors. This improvement was slight in case of CSRS-PPP and considerable with PPP-WIZARD. CSRS-PPP provides the best overall performance among the three services, achieving 0% outliers, AAE and RMSE values of less than 2 cm in all components, and 3D-MAE values of less than 10 cm and 7 cm for GPS-alone and GPS+GLONASS, respectively. PPP-WIZARD ranks second, where the combined solution reduces RMSE values in E, N, and H by approximately 9 cm, 15 cm, and 42 cm, respectively, and decreases the proportion of outliers by about 25%, 7.5%, and 11.5%, in the same order. In addition, AAE and MAE values of kinematic PPP-WIZARD decrease substantially in E, N, and H by approximately 1 m, 2.2 m, and 2.8 m, respectively, when using the integrated constellations. APPS supports GPS observations only and even when compared with GPS-alone results from CSRS-PPP and PPP-WIZARD, it exhibits the lowest accuracy and the highest percentage of outliers. APPS yields 3D-AAE, 3D-MAE, and 3D-RMSE values of approximately 0.65 m, 3.5 m, and 0.63 m, respectively, with outlier percentages of 64% in E, 38% in N, and 45% in H. Overall, the results clearly highlight the significant advantages of multi-constellation GNSS processing over single-system solutions, particularly in terms of increased observation redundancy, improved satellite geometry, and enhanced capability to mitigate weak or unreliable observations. The comparative results confirm that CSRS-PPP provides the best performance using both GPS-alone and GPS+GLONASS with slight differences for the benefit of the second. PPP-WIZARD came second with acceptable performance under the integrated constellation processing which can be used for some engineering applications that require decimeter level of accuracy. However, GPS-alone PPP-WIZARD and APPS kinematic solutions remain unstable, less precise, and characterized by a high percentage of gross errors.

References

[1] Hofmann-Wellenhof B, Lichtenegger H, Wasle E. GNSS global navigation satellite systems: GPS, GLONASS, Galileo, and More. 2007 [2] Amami M. The Advantages and Limitations of Low-Cost Single Frequency GPS/MEMS-Based INS Integration. Global Journal of Engineering and Technology Advances. Feb. 2022; 10(2): 018-031. [3] Amami M. Investigations into the quality of final-product dual frequency kinematic PPP in open sky using CSRS-PPP free online service: GPS Vs. GPS+GLONASS. Global Journal of Engineering and Technology Advances (GJETA). November 2025; 25(2):138-143. [4] Amami M. Investigations into the quality of final-product dual-frequency static PPP over fixing time using CSRS-PPP free online service: GPS Vs. GPS + GLONASS. World Journal of Advanced Engineering Technology and Sciences (GJETA). November 2025; 17(2): 438-443. [5] Amami M. Kinematic PPP Using PPP-WIZARD Free Online Service: GPS Vs. GPS + GLONASS. International Research Journal of Modernization in Engineering Technology and Science. December 2025; 7(12): 3445-3451. [6] Amami M, Smith M, Kokkas N. Low-Cost Vision Based Personal Mobile Mapping System. ISPRS- International Archives of The Photogrammetry, Remote Sensing and Spatial Information Sciences. 2014; XL-3/W1: 1-6. [7] Amami M. A Novel Design Concept of Cost-Effective Permanent Rail-Track Monitoring System. World Journal of Advanced Research and Reviews. March 2022; 13(3): 451-473. [8] Amami M. The Integration of Image-Based Navigation and GPS Carrier Phase Delta Positioning for Aerial Triangulation using Low-Cost Drones with Limited Number of Ground Control Points. Quest Journals. 2022; 7(5): 23-31. [9] Amami M, EL-Turki A, Rustum A, EL-Amaari I, Jabir T. Topographic Surveying using low-cost amateur drones and 4K ultra-high-definition videos. Open Access Research Journal of Science and Technology. April 2022; 4(2): 072-082. [10] Amami M, Elmehdwi A, Borgaa A, Buker A, Alareibi A. Investigations into utilizing low-cost amateur drones for creating ortho-mosaic and digital elevation model. International research journal of modernization in engineering technology and science. March 2022; 4(3): 2107-2118. [11] Amami M. Quality of Static PPP over observation time using PPP-WIZARD free online service: GPS Vs. GPS + GLONASS. Open Access Research Journal of Science and Technology. December 2025; 15(2): 195-200 [12] Amami M, Fadeil A. Investigations into the quality of Kinematic PPP in open-sky using final-product APPS free online GPS-alone service. GSC Advanced Research and Reviews. January 2026; 26(1): 023:028 [13] Amami M. Enhancing Stand-Alone GPS Code Positioning Using Stand-Alone Double Differencing Carrier Phase Relative Positioning. Journal of Duhok University (Pure and Eng. Sciences). 2017; 20(1): 347 355. [14] Amami M. The Integration of Stand-Alone GPS Code Positioning, Carrier Phase Delta Positioning and MEMS-Based INS. International Research Journal of Modernization in Engineering Technology and Science. March 2022; 4(3): 700-715. [15] Amami M. The Integration of Time-Based Single Frequency Double Differencing Carrier Phase GPS/ Micro-Elctromechanical System-Based INS. International Journal of Recent Advances in Science and Technology. Dec. 2018; 5(4): 43-56. [16] Amami M, Fadeil A. Evaluating The Quality Of Static PPP Over Observation Time Using APPS Free Online GPS-Alone Service. International Journal of Engineering Development and Research. January 2026; 14(1): 787-792 [17] Amami M. Testing Patch, Helix and Vertical Dipole GPS Antennas with/without Choke Ring Frame. International Journal for Research in Applied Sciences and Engineering Technology. Feb. 2022; 10(2): 933-938. [18] Amami M. Fast and Reliable Vision-Based Navigation for Real Time Kinematic Applications. International Journal for Research in Applied Sciences and Engineering Technology. Feb. 2022; 10(2): 922-932. [19] Amami M. Comparison Between Multi and Single Epipolar Geometry-Based Filters for Optical Robot Navigation. International Research Journal of Modernization in Engineering Technology and Science. March 2022; 4 (3): 476-485. [20] Amami M. Comparison Between Multi Epipolar Geometry and Conformal 2D Transformation-Based Filters for Optical Robot Navigation. International Journal for Research in Applied Sciences and Engineering Technology. March 2022; 10(3): 388-398. [21] Amami M. Comparison Between Multi Epipolar Geometry and Affine 2D Transformation-Based Filters for Optical Robot Navigation. International Journal for Research in Applied Sciences and Engineering Technology. March 2022; 10(3): 399-409. [22] Amami M. Speeding up SIFT, PCA-SIFT and SURF Using Image Pyramid. Journal of Duhok University, [S.I]. July 2017; 20(1): 356-362. [23] Amami M. Multi and Single Epipolar Geometry-Based Filters Vs. Affine and conformal 2D Transformation-Based Filters. Global Journal of Engineering and Technology Advances. March 2022; 10(3): 032-051.

Copyright

Copyright © 2026 Mustafa M. Amami. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.