One of the most important issues in a design of navigation system is to estimate the system state accurately with limited sensors. This study aims at verifying the navigation system of an experimental airframe model of a new space plane concept. The most widely used algorithm to estimate the system state in the navigation system is the extended Kalman filter (EKF). However, using the EKF is not enough for the experimental airframe model to fill the required estimate accuracy. Therefore, in this study, the unscented Kalman filter (UKF) which has higher accuracy compared with the EKF is used for the state estimation of the nonlinear system. These navigation systems which use UKF or EKF are evaluated by numerical simulation.