GPS based positioning of Satellite using cubature Kalman filter

Due to the nonlinearity of the equations governing the satellite estimation system, linear filters are not able to estimate the precise position, and then satellite tracking is associated with several errors. In this paper, satellite motion equations are investigated then, using GPS observations, Extended Kalman filter (EKF) and Cubature Kalman filter (CKF) the satellite’s position and velocity are determined. Cubature Kalman filter is an algorithm suitable for estimating noisy high dimension nonlinear systems, which is based on the Gaussian and Kalman filter. Simulation results and RMS of the positioning errors confirm that the Cubature Kalman filter has improved accuracy and performance compared to the Extended Kalman filter. Cubature Kalman filter shows 50 percent improvement in velocity estimation in y and z directions compared to EKF. Although, relative error improvement percentage for position is less than velocity and both filters have almost similar performance. However, CKF is slightly superior and has had 10 percent, 3 and 1.5 percent improvement in relative errors of x, y and z directions, respectively.