Iterated Extended Kalman Filter Application to Real-time GPS Data Processing

The Kalman filtering is a method with which the raw data with noise can be cleaned. The standard KF can be extended to a non-linear model. In the extended Kalman filter (EKF), Taylor proximate formula has been applied to both state equations and measurement equations, in order to estimate linearized dynamical models. But if the initial value is incorrect or the noise is very strong, the linearized models may not be good anymore. The iterated extended Kalman filter (IEKF) therefore has been applied to GPS raw data processing, and the results are satisfactory.