Development and Performance Analysis of Algorithm for Joint Processing of Measurements from UWB ToF/AoA LPS and PDR Algorithm Estimates for Known User Height

This paper is devoted to the development of observation models used in the algorithm for joint processing of measurements from ultra-wideband time-of-flight/angle-of-arrival local positioning system and pedestrian dead reckoning algorithm estimates. In previously published papers the algorithm for joint processing of measurements from ultra-wideband time-of-flight/angle-of-arrival local positioning system and pedestrian dead reckoning algorithm estimates was developed and described in details, however it doesn't take into account the height difference between the user and the radio navigation reference point of local positioning systems. The developed models for considered algorithm has been tested in simulation and real experiment, results of which allow us to conclude that height difference between the radio navigation reference point and wearable module causes a systematic error in y-coordinate estimates. To compensate for this error, it is proposed to include the pre-measured value of abovementioned height difference in the observation model of proposed filter. For the trajectory traveled in the real experiment, the height-aware model showed a gain in the accuracy of y-coordinate estimate by a factor of more than 4 over the height-unaware model.