A comparison between different step length estimation algorithms for pedestrian dead reckoning is presented. This work covers theoretic evaluation of the estimators' performance and presents a comparison based on measurement data. Measurement data were taken from a group of five adults walking at three different velocities. For reference, the sensors were placed according to the recommendation given for each algorithm. In respect to everyday usability the performance of the estimators is furthermore evaluated for arbitrary placement of the sensors, as it is the case when using a mobile measurement platform like a smartphone

For a microwave locating system virtual synchronisation of low-cost free-running transmitter clocks is proposed. The synchronisation strategy is based on modelling the transmitter clocks. A stable stochastic model was designed by analysing the correlation function of the observables in the locating system i.e. time of arrival and carrier phase. Exploiting this model in a Kalman filter, a reference receiver estimates the current clock states and distributes them periodically to the mobile receivers enabling them to predict the transmitter clock deviations between updates. The impact of this synchronization scheme on the position accuracy using both simulated and real data is analysed

No abstract

An inexpensive, flexible and precise pseudolite positioning system, developed at Fraunhofer IIS is presented. For this pseudolite system a real-time kinematic (RTK) method similar to the one employed in GNSS is applied. Because of the use of free-running clocks in the pseudolites and the characteristics of the signal, the application of RTK to the proprietary pseudolite system is challenging. A description of these challenges e.g., synchronization of transmitters and receivers, and methods to overcome them are addressed in this paper. The application of RTK allows accurate localization of a rover. The rover position is estimated through an extended Kalman filter (EKF) and the carrier phase integer ambiguity is resolved through Least-squares AMBiguity Decorrelation Adjustment (LAMBDA) method

A derivation of phase smoothing for a pseudolite system is given. The distinction to phase smoothing known from literature arises from the setup of the system. As of the employed free-running pseudolites, the main challenge is to properly estimate the pseudolite clocks. The better the pseudolite clocks are estimated, the more precise are the phase measurements and the more effective is phase smoothing. Two different methods for clock estimation are examined in terms of their applicability to phase smoothing. Simulated as well as real data are used in order to evaluate the precision gain by phase smoothing. Optimal solutions according to system restrictions are presented