TL;DR
This paper investigates how Japan's Quasi-Zenith Satellite System (QZSS) can enhance the positioning accuracy of GPS and BeiDou (BDS) using single-frequency, single-epoch measurements in the Asia-Pacific region. The findings demonstrate that QZSS significantly improves availability, reduces position dilution of precision (PDOP) and ambiguity dilution of precision (ADOP), and increases the success rate of ambiguity resolution, particularly for longer baselines.
What Problem Does This Paper Address?
Global Navigation Satellite Systems (GNSS) like GPS and BDS are crucial for many applications, but achieving high-accuracy positioning, especially with single-frequency, single-epoch (SFSE) measurements, can be challenging. Factors like satellite geometry (PDOP), atmospheric delays, and multipath effects can degrade accuracy. In the Asia-Pacific region, while GPS and BDS offer coverage, their performance can be further improved. The Quasi-Zenith Satellite System (QZSS), a regional system developed by Japan, is designed to augment existing GNSS constellations, particularly for users in the Asia-Pacific. This paper aims to quantify the actual performance enhancement QZSS provides to GPS and BDS in this specific region, using real-world data, to understand its potential for improving SFSE positioning.
Methods
The study employs a single-frequency, double-differenced (DD) mathematical model for positioning. This approach is standard for high-accuracy GNSS positioning and involves differencing measurements between two receivers and two satellites to eliminate common errors. The researchers developed specific DD models for integrating QZSS with GPS (GPS/QZSS) and with BDS (BDS/QZSS). They also incorporated a stochastic model based on elevation angle weighting to account for varying observation quality.
For evaluation, the paper uses two key metrics:
- Availability and PDOP: Assessing the number of visible satellites and their geometric distribution.
- ADOP and Success Rate: Quantifying the quality of integer ambiguity resolution, which is critical for high-accuracy positioning. The bootstrapped success rate (Bsrt) and actual success rate (Asrt) are used.
Two real-data experiments were conducted using different baseline lengths (12.5 km and 4.8 km) in Hong Kong to evaluate the performance under various conditions.
Datasets
The study utilizes actual QZSS, GPS, and BDS data collected from stations in Hong Kong. Specific details on the exact satellite constellations and receiver types used are provided. The data covers a week-long period for simulation-based analysis and specific time intervals for the two experimental cases. No new benchmark datasets were introduced; rather, existing GNSS data was used to evaluate the performance of integrated systems.
Key Insights and Ideas
- QZSS as a powerful augmenter: QZSS is not a standalone system but a significant enhancer for GPS and BDS, especially in the Asia-Pacific region.
- Improved satellite geometry: By adding QZS satellites, the overall number of visible satellites increases, leading to a better geometric distribution and lower PDOP.
- Enhanced ambiguity resolution: QZSS significantly improves the success rate of integer ambiguity resolution (higher Asrt and Bsrt), which is crucial for achieving centimeter-level accuracy in SFSE positioning.
- BDS advantage in Asia-Pacific: BDS generally performs better than GPS in terms of satellite availability and elevation in the Asia-Pacific region, even without QZSS augmentation.
- Synergistic benefits: Combining BDS with QZSS offers substantial improvements, and the integration of all three systems (BDS+GPS+QZSS) yields the best overall performance.
- More QZS, better results: The more QZS satellites are utilized, the greater the enhancement in positioning performance.
Key Results
- Availability and PDOP: For a 35° mask angle, GPS/QZSS improved the mean number of satellites (NOS) and reduced PDOP compared to single GPS. Similarly, BDS/QZSS showed improvements over single BDS.
- ADOP and Success Rate: In a 10 km baseline simulation, GPS/QZSS showed a significant decrease in ADOP and a substantial increase in Bsrt compared to single GPS. BDS/QZSS also improved over single BDS, though the relative improvement was smaller.
- Experimental Case 1 (12.5 km baseline):
- GPS/QZSS improved the actual success rate (Asrt) from 37.40% to 94.51% with three QZS.
- BDS/QZSS improved Asrt to 99.61% with three QZS.
- The combined BDS+GPS+QZSS achieved 100% Asrt and the highest accuracy (1.25 cm precision).
- Experimental Case 2 (4.8 km baseline):
- GPS/QZSS improved Asrt from 75.79% to 94.78% with two QZS.
- BDS/QZSS improved Asrt to 99.99% with two QZS.
- The combined BDS+GPS+QZSS achieved 100% Asrt and the highest accuracy (1.03 cm precision).
Limitations and Future Work
- Inter-System Biases (ISBs): The paper acknowledges that ISBs between different GNSS systems (e.g., GPS and BDS, or BDS and QZSS) can occur and vary. While they adopted a loosely combined model and noted that ISBs between GPS L1 and QZSS L1 are stable and can be corrected, further research into the characteristics of ISBs between overlapping and different frequencies (BDS B1 and QZSS L1) is suggested for future work.
- Bsrt vs. Asrt: The paper notes that the bootstrapped success rate (Bsrt) can sometimes be inaccurate, especially with longer baselines or larger DD residuals, and that the actual success rate (Asrt) is a more reliable indicator. This suggests that further investigation into the factors affecting Bsrt accuracy is warranted.
- Limited geographical scope: While the study focuses on the Asia-Pacific region, the specific experiments were conducted in Hong Kong. Expanding the analysis to other areas within the QZSS service region could provide a more comprehensive understanding.
Summary updated