Precise Orbit Determination of CubeSasts Using a Proposed Observations Weight Model

CubeSats are small low-cost and low-power satellites that can be used for many space missions. Some missions require precise location determination of the cubeSats, such as radio-occultation, Interferometric Synthetic Aperture Radar (InSAR), satellite altimetry, and gravity field recovery. Therefore, precise orbits and clocks of CubeSats is essential to achieve the required accuracy. They are also essential for future mega-constellations Low Earth Orbit (LEO) satellites that are proposed as augmentation systems for positioning and navigation. The Precise Orbit Determination (POD) methods are well developed for large LEO satellites during the last two decades. However, CubeSats are mainly built from commercial off-the-shelf (COTS) components and have their own characteristics, which need new investigations. In this paper, seventeen 3U-CubeSats, launched in different orbits in the Spire Global Constellation, are analyzed in terms of precise orbits and clocks. The orbits generated from both the reduced-dynamic and the kinematic POD methods are validated internally with the overlap analysis, the posterior covariance factors, and the residuals. One-month processing of these CubeSats revealed that around 90% of precise orbits have decimeter accuracy, while 50% are at centimeter-level. This accuracy fulfills most of the abovementioned space and earth science applications. The limitations in using elevation-dependant weighting models for CubeSats POD are discussed and, as an alternative, a weighting model based on signal-to-noise ratio has been proposed and tested. The impact of crossing CubeSats from the eclipse region, as well as the near-filed multipath due to the CubeSat structure and the signal direction in space, are also considered in the proposed weighting model