Invariant extended Kalman filter for tracking the bishop frame using radar measurements

This paper proposes a target model for radar state estimation based on the Bishop or Parallel Transport frame, extending upon previous work using the Frenet-Serret frame to describe target motion. The Bishop frame is embedded into an element of the SE(3) Lie group which allows for the partial implementation of the invariant extended Kalman filter (IEKF) to estimate the the target pose, unit norm velocity and Bishop curvatures. The performance of the proposed filter is compared to the original Frenet-Serret IEKF and industry standard algorithms. Results demonstrate comparable performance to the Frenet-Serret IEKF and the choice of kinematic model is discussed.