Item

Abstract

Precise satellite orbit determination is commonly carried out by exploiting range measurements between ground-fixed stations and Earth-orbiting satellites. During the last decade an increasing number of Earth satellites, operating mainly in low-orbit mode (LEO), are equipped with GNSS receivers thus enabling the linkage of different satellites in space via inter-satellite range measurements. The usage of such observations provides a compelling option for estimating the orbits of Earth satellites without the need to employ ground-fixed tracking stations. To explore the potential of this option, a simulator has been developed in SYRTE for studying the satellite-to-satellite orbit determination problem. The simulator is designed to deal with different satellite constellations and parameterization schemes by considering a variety of physical dynamical models for the orbital motion. Furthermore, a least-squares adjustment strategy is implemented on simulated satellite-to-satellite range measurements to estimate the initial conditions (positions-velocities) of the considered satellite constellation.Using the aforesaid simulator, our aim here is to investigate the orbit determination problem through inter-satellite range measurements between GNSS and LEO satellites. In the present study we chose to work with the combination of three different dynamical models, namely the central gravity field, the J₂ acceleration force and the attractive forces coming from the planets and the Moon. The objective is to analyze the rank-deficiency of the normal equations system derived from different GNSS+LEO satellite constellations, and to assess the quality of the estimated state parameters (initial satellite positions-velocities) by applying different types of constraints to overcome the inherent datum deficiency of the underlying problem.