Upcoming Release: Enhanced Precision in GNSS

Key points

Highlighting Integration Milestones: Showcasing the latest advancements of RISE in autonomously implementing Precise Point Positioning (PPP) with the use of the Galileo High Accuracy Service (HAS) on streamlined GNSS receivers.

Pioneering Efficient Algorithms: The implementation of powerful yet efficient algorithms is optimized for low-power hardware, showcasing our commitment to energy-efficient and high-performance solutions.

Achieving Centimeter Precision: Our technology brings centimeter-level accuracy to stand-alone OEM positioning devices. This elevates the performance of a wide array of vehicles, including rovers, drones, and logistic trackers, as well as enhances operations in mining, construction, and precision agriculture.

Electronic circuit board of the OrbFix GNSS receiver module

Introduction

Precise Point Positioning (PPP) is an advanced GNSS (Global Navigation Satellite System) technique designed to pinpoint a receiver’s location with exceptional accuracy, often down to the centimeter. This level of precision is achieved through correcting various satellite biases, including clock errors, code, and phase discrepancies. While traditionally these corrections are accessed via paid subscriptions, the Galileo High Accuracy Service (HAS) offers them at no cost.

Understanding PPP: Precise Point Positioning (PPP) uses a sophisticated algorithm to improve location accuracy beyond standard GNSS capabilities. It is essential for applications requiring high precision, such as geodesy, surveying, and autonomous vehicle navigation.

Role of Corrections in PPP: The accuracy of PPP heavily relies on external corrections related to the satellite’s clock, code, and phase biases. These corrections adjust for errors that can affect the precision of the positioning data.

Advantages of Galileo HAS: Unlike other services that require a subscription, Galileo HAS provides these crucial corrections, independent of external services (i.e. internet, radio connection). It increases autonomy and use cases to remote locations covered only by satellite. This accessibility can significantly lower the barrier to entry for high-accuracy GNSS applications, making precise positioning more attainable for users and industries worldwide.

How it works

PPP employs sophisticated filtering techniques in real-time to estimate a comprehensive state vector. This vector not only captures the receiver’s precise location but also integrates critical contributing factors, such as tropospheric characteristics and phase biases, which are essential for achieving an accurate solution. In our specific application, we have implemented a robust adaptive Extended Kalman Filter (EKF).

Our advanced EKF effectively harnesses both pseudorange and phase measurements from the receiver. By processing these inputs, the filter meticulously refines the state vector, enhancing the precision and reliability of the positioning data. This precision is attained by modeling a spectrum of effects that influence GNSS accuracy.

We incorporate adjustments for path relativistic effects, which account for the influence of Earth’s gravity on the signal path. The Sagnac effect, caused by the rotation of the Earth, is also precisely calculated to ensure the accuracy of our system. To counteract the impact of the ionosphere on signal propagation, we utilize an iono-free combination technique for tropospheric corrections.

The algorithm further compensates for phase wind-up effects, ensuring that the periodic changes in the relative position of satellites and the receiver do not degrade the signal quality. Additionally, meticulous care is taken to correct for antenna phase-center offsets and variations, which are critical for maintaining consistent measurement quality.

Our system is robust against geophysical disturbances, including the solid Earth tides and eclipse effects on satellites, maintaining operational consistency under various environmental conditions. Enhanced cycle-slip detection is integral to our approach, quickly identifying and correcting discontinuities in the satellite signal to ensure the integrity of our positioning solutions.

Flowchart illustrating the process flow within a GNSS system: Galileo HAS Message Decoder providing Orbit & Code bias corrections and Clock corrections, which feed into a Robust and Adaptive Navigation Filter featuring an Extended Kalman Filter for enhanced satellite navigation accuracy.
Workflow of a GNSS Receiver utilizing the Galileo HAS to decode messages and apply corrections for precise positioning, highlighting the integration of an EKF for advanced navigation accuracy

The state vector encapsulates several crucial elements that contribute to determining the receiver’s precise location and time. It includes:

  1. Receiver’s Position: This is the core component, detailing the exact geographical coordinates of the receiver.
  2. Receiver’s Clock Bias and Clock Drift: These metrics address any discrepancies between the receiver’s internal clock and the actual time, ensuring timing precision.
  3. Inter-System Bias: Arising from the use of both GPS and Galileo satellite constellations, this factor compensates for the differences between the systems to improve accuracy.
  4. Zenith Wet Delay and Tropospheric Gradients: These elements account for atmospheric delays (particularly humidity effects) that impact signal transmission, with specific adjustments for the east and north components.
  5. Phase Biases: These corrections are vital for refining the position data derived from satellite signals, mitigating errors introduced by phase inconsistencies.

By integrating these components, the state vector in PPP allows for an exceptionally precise positioning solution, crucial for applications requiring the utmost accuracy.

Our EKF filter uses state of the art algorithms designed to enhance accuracy by constructing an adaptive covariance matrix and dynamically adjusting the Kalman Gain based on the analysis of measurement residuals.

Key to our technique is the computation of a standardized residual that consolidates data from both GPS and Galileo satellite systems. These residuals are crucial for calculating a robust classification factor using the t-test statistic, which evaluates the consistency and reliability of the measurement data. This factor plays a pivotal role in recalibrating the Kalman Gain, which subsequently refines the updated state covariance and the state vector.

This process not only boosts the filter’s sensitivity to discrepancies and anomalies in data quality but also tailors the measurement noise covariance matrix to reflect the varying precision of different satellite systems and signal types. The result is a significantly more reliable and accurate PPP solution, ensuring enhanced positioning output across varying conditions.

Results

The figure depicted below showcases the long-term performance of our proprietary algorithm, optimized for a static receiver setup. The convergence time of our filter can vary based on the desired level of accuracy. Due to our robust filtering approach, it typically achieves a 3D error of less than 20 centimeters within 30 to 40 minutes. Importantly, while the algorithm excels in static scenarios, we have specifically tailored it to perform well in dynamic applications, adapting to the needs of mobile environments.

Graph showcasing the Extended Kalman Filter (EKF) convergence over time in a static setup, with lines representing the X, Y, and Z axis errors, demonstrating error reduction and stabilization.
Error convergence graph for a static GNSS receiver setup, depicting the performance of the EKF as it achieves precision correction in the X, Y, and Z coordinates

Another test was conducted in a kinematic scenario, utilizing the Galileo HAS corrections. The results displayed below compare the receiver’s SPP solution with the output from our navigation filter.

Satellite image overlaid with GNSS tracking data, showing the EKF solution in a kinematic setup with two paths: a red dotted line for Receiver SPP and a blue line marked by triangles for the GnXact, through urban areas indicating path accuracy
Satellite view of a kinematic testing environment with superimposed GNSS tracking paths: the Receiver SPP path shown in red dots, versus the GnXact path in blue triangles, highlighting the fidelity of the EKF solution in an urban landscape
Aerial map highlighting GNSS receiver tracks over mixed terrain with a red dotted path for Receiver SPP and a blue path for GnXact.
Comparative GNSS track analysis on a rural to urban transition landscape, with Receiver SPP’s path depicted in red dots and the GnXact demonstrated with the blue line