Bayesian multisensor navigation incorporating pseudoranges and multipath model

Indoor and urban canyons are application areas that are becoming increasingly important for navigation application. However, achieving the required accuracy and availability is still a challenge. Multisensor navigation is one of the techniques that has shown promising results in addressing the challenges of such areas. Being able to incorporate raw low-level sensor data is advantageous as it allows to incorporate all available information and more accurate estimation models. In this paper a Particle Filter based multisensor positioning system is extended to use pseudorange measurements of a GPS sensor instead of a calculated position solution. Using pseudoranges, any number of visible satellites can improve positioning accuracy, when combined with measurements from other sensors like an electronic compass, a barometric altimeter or a foot mounted inertial measurement unit (IMU). Additionally, statistical error models for pseudoranges are integrated and tested. Our results show that by using the models within the Bayesian framework yields promising results in terms of error mitigation.