Decentralized Cooperative Localization with Relative Pose Estimation for a Spacecraft Swarm

This paper presents a decentralized cooperative localization approach to relative state estimation for a team of spacecraft to estimate inertial position, orientation, velocity, and angular velocity of each spacecraft. The solution uses an Extended Kalman Filter (EKF) running on each spacecraft to estimate the full state of every other spacecraft in the swarm. A Multiplicative EKF is used to address limitations of quaternion representation of attitude; non-quaternion states are estimated by a standard EKF. Each spacecraft is equipped with a gyroscope, a star tracker, and a relative pose measurement sensor. Only pose measurements between spacecraft are shared among the team. Simulation results are provided which demonstrate an improvement in state estimation for the swarm when shared relative pose measurements are incorporated into the filters running on each spacecraft. Estimation performance in various sensing graph configurations is also explored. The methodology and results presented in this paper provide a foundation for future research by introducing the concept of cooperative localization into spacecraft swarm state estimation.