GPS Satellite Position Estimation from Ephemeris Data by Minimum Mean Square Error Filtering Under Conditions of Selective Availability

This paper investigates GPS satellite position estimation and infers conclusions from the analysis performed. The position estimation is done by evaluating a system of equations under both ideal and noisy conditions. A minimum mean squared error filter is then applied to the resulting position data. Additive noise is then applied to the received ephemeris data to simulate a selective availability condition. The results from the minimum mean squared error filter are compared to the results from the standard algorithm under noisy conditions and normal conditions. From this analysis quantitative judgment is used to determine this methods feasibility. This analysis is performed on selected data gathered from the International GNSS service. This data is from a single satellite and the time period is a single week. Analysis using the above method shows that an improvement of satellite position estimation is possible only under the noisiest of conditions.