Quaternion based heading estimation with handheld MEMS in indoor environments

Abstract

Pedestrian Dead-Reckoning (PDR) is the prime candidate for autonomous navigation with self-contained sensors. Nevertheless with noisy sensor signals and high hand dynamics, estimating accurate attitude angles remains a challenge for achieving long term positioning accuracy. A new attitude estimation algorithm based on a quaternion parameterization directly in the state vector and two opportunistic updates, i.e. magnetic angular rate update and acceleration gradient update, is proposed. The benefit of this method is assessed both at the theoretical level and at the experimental level. The error on the heading, estimated only with the PDR navigation algorithms, is found to less than 7° after 1 km of walk.