A Stochastic Model Based on Optimal Satellite Subset Selection Strategy for Smartphone Pseudorange Relative...

Area of Science:

• Geomatics Engineering
• Satellite Navigation Systems
• Signal Processing

Background:

• Traditional stochastic models for smartphone Global Navigation Satellite System (GNSS) positioning face limitations due to observation quality constraints.
• Mitigating these constraints is crucial for enhancing the accuracy and reliability of phone-based GNSS applications.

Purpose of the Study:

• To introduce a novel double-difference code pseudorange residual (DDPR)-dependent stochastic model for smartphone GNSS positioning.
• To improve positioning accuracy and reliability by optimizing satellite selection and observation weighting.

Main Methods:

• Developed a DDPR-dependent stochastic model utilizing an optimal satellite subset selected based on carrier-to-noise density ratio (C/N0).
• Estimated DDPRs for satellites and used them as prior information to build the observation stochastic model.
• Applied pseudorange differential positioning to determine approximate terminal locations.

Main Results:

• The DDPR-dependent model significantly improved positioning accuracy in occluded environments for Huawei Mate40 and P40 terminals compared to the C/N0-dependent model.
• Achieved approximately 30-34% accuracy improvements in North, East, and Up directions for Mate40, and 26-33% for P40.
• The model effectively identified and down-weighted multipath and non-line-of-sight (NLOS) signals, enhancing overall positioning performance.

Conclusions:

• The proposed DDPR-dependent stochastic model offers a computationally efficient and straightforward method for enhancing smartphone GNSS positioning.
• This approach is particularly suitable for complex urban environments with signal obstructions.
• The model successfully mitigates gross errors from multipath and NLOS signals, leading to more accurate and reliable positioning.