Enhancing satellite positioning with LEO-PNT: A robust PVT estimator
Researchers from the Faculty of Applied Engineering have developed a refined and robust adaptation of the Position, Velocity, and Timing (PVT) estimator for Low Earth Orbit Positioning, Navigation, and Timing (LEO-PNT).
Situation before
Current Global Navigation Satellite Systems (GNSS) operate from Medium Earth Orbit (MEO) at ~20,000 km altitude, and are widely used to determine absolute position and time in outdoor environments. However, they face several limitations—particularly due to the high orbital altitude of the satellites, which results in weak signal strength upon arrival on Earth. This makes GNSS signals vulnerable to jamming, less reliable for indoor and urban environments, and more energy-intensive for battery-powered devices.
LEO satellites, orbiting closer to Earth, offer a promising alternative. Their proximity ensures stronger signals, and their high relative velocity allows for positioning based on Doppler shifts. However, traditional GNSS-based PVT estimators, which rely on Least Squares algorithms, perform poorly in cold-start scenarios when only LEO signals are available. In such cases, the algorithm may begin its search from the Earth's center, leading to divergence.
Technology
The research group has developed an enhanced PVT estimation method that enables cold-start positioning using only LEO signals, only GNSS signals, or a combination of both. This solution uses existing range and Doppler measurements but modifies the initialisation procedure in the Least Squares algorithm, without increasing computational complexity or requiring a complete overhaul of the system.
The result is a drop-in enhancement that can be integrated into existing receiver architectures, enabling them to seamlessly support positioning with satellites from different orbital altitudes.
Key advantages of this solution:
- Solves LEO Cold Start Convergence: It ensures convergence of iterative descent algorithms, even without upfront position knowledge, enabling receivers to determine location without prior rough location.
- Faster Convergence & Higher Computational Efficiency: It dramatically increases the convergence rate, reducing the number of iterations and thus fewer computationally complex matrix inversions.
- Reduced Energy Consumption: The lower computational burden translates to less energy use, beneficial for handheld consumer devices and extending battery life.
- Easy & Broad Implementation: The method can be easily integrated into existing geospatial receivers and iterative descent algorithms. It is signal-independent, applicable in all regular GNSS receivers and future LEO-PNT receivers, and works with various observables such as pseudo range, carrier phase, and doppler shift.
Partners we search for
The technology can be implemented in geospatial receiver hardware (navigation circuitry), as software instructions for a receiver or computing system, or distributed on a computer-readable medium.
We are seeking partnerships with leading geospatial receiver and chipset manufacturers to explore licensing opportunities and/or further adaptation of this solution.
About the research team
Our innovation of “Earth-Step Adjustment for Geospatial Positioning” is a collaborative effort by researchers from the IDLab research group. Key researchers involved are Wout Van Uytsel, Andres Maria Majorana, Dr. Thomas Janssen and Prof. Dr. Maarten Weyn. The goal was to develop a solution that could be practically deployed and easily adopted by the existing GNSS ecosystem.
IDLab is a core research group within imec, with part of its activities embedded at the University of Antwerp’s Faculty of Applied Engineering and Faculty of Science. The UAntwerp branch of IDLab conducts leading research in wireless networks and artificial intelligence.
More information
University of Antwerp
Valorisation Office
Middelheimlaan 1
2020 Antwerp - Belgium