Prediction of IRNSS User Position using Regression Algorithms

Indian Regional Navigation Satellite System (IRNSS) is a regional satellite-based navigation system used over the Indian region for computing the user position and for navigation. This system is used by the military, civilian and commercial receivers. However, if the transmitted signals of IRNSS satellites are lost temporarily due to the atmospheric conditions, satellite health issues, receiver located at the place where there is no satellite visibility then the user position estimation may not be possible at the receiver. To get rid of this problem the Regression algorithms are used for the prediction of user position. Here, Regression tree and Bagged tree algorithms are utilized for the purpose. The performance measure for these algorithms is Root Mean Square error (RMSE). The user position is estimated in Earth-centered, Earth-Fixed (ECEF) coordinate system. It is observed that the RMSE values for the predicted ${(x, y, z)}$ using Bagged tree are is better when compared to RMSE values for the predicted ${(x, y, z)}$ using Regression tree.