Abstract: Techniques are provided for non-linear satellite state modeling. First global navigation satellite systems (GNSS) signal data is obtained from a set of GNSS satellites. First satellite state data is obtained. The first satellite state data includes orbit data for the set of GNSS satellites. A non-linear satellite state model that includes a plurality of model parameters is generated. The non-linear satellite state model is generated by adjusting the plurality of model parameters based on the first GNSS signal data and the first satellite state data. The non-linear satellite state model outputs satellite state data based on GNSS signal data. Second GNSS signal data is obtained from the set of GNSS satellites. A set of updated satellite state data is calculated using the non-linear satellite state model and the second GNSS signal data.

Abstract: A neural network is used to process raw, uncorrected signals received by a Global Positioning System (GPS) receiver to obtain signal corrections which are used to correct the raw signals and obtain highly accurate position coordinate data. The neural network is trained with a particular GPS receiver. Once the neural network is trained, the weight matrices used for the particular GPS receiver may be used in different GPS receivers without requiring training of the different GPS receivers.