LAUSR

The indoor pseudolite (PL) code-based solution is not sufficiently precise, and conducting prolonged static carrier phase observations is impractical due to the static PL geometric distribution. Thus, a high-precision indoor PL positioning, which is generally based on the known point initialization (KPI) method and adopts the extended Kalman filter (EKF), is used to obtain the PL float ambiguity solution. However, the first-order linear truncation error of the EKF cannot be neglected because the indoor space is small, and its performance is quite dependent on the accuracy of the initial coordinates in the KPI. In this study, a well-suited nonlinear parameter estimation method, which is expected to have high precision and low dependence on the initial coordinate values, namely the unscented Kalman filter (UKF), is introduced and applied in PL positioning. Based on the relative positioning model and the KPI method, the positioning performance of the EKF and UKF under code-based differential pseudolite (DPL) positioning and phase-based real-time kinematic (RTK) positioning modes is compared and analyzed. Numerical results indicate that the computational efficiencies of the EKF and UKF are of the same level, though the former is slightly superior to the latter. In terms of the DPL positioning results, the precision of the UKF is higher with a decimeter-level improvement compared with that of the EKF. The dependence on the accuracy of the initial coordinates of UKF is reduced to some extent, which makes it a convenient technique, especially in the PL ambiguity resolution of the RTK positioning with fast convergence speed. The UKF outperforms the EKF and is more practicable in indoor PL positioning. The UKF can also achieve a decimeter-level positioning precision in DPL and a centimeter-level positioning precision in RTK.