Comparing non-linear filters for aided inertial navigators

The performance of the Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF) in an aided-inertial navigator for terrestrial applications is analyzed. Parameters affecting the quality of state estimate provided by these two non-linear filters are identified. It is shown that for a GNSS-aided navigator, the performance of these two filters is very similar and, thus, argues that the EKF is adequate for this application. However, for applications where the aiding for the inertial sensors comes from ranging sources other than GNSS, it is shown that the performance between these two filters can be different.