An Interacting Multiple Model Estimator of LEO Satellite Clocks for Improved Positioning
Nadim Khairallah, Zaher M. Kassas
Abstract
An interacting multiple-model (IMM) estimator is developed to adaptively estimate the process noise covariance of low Earth orbit (LEO) satellite clocks for improved positioning. Experimental results are presented showing a stationary ground receiver localizing itself with carrier phase measurements from a single Orbcomm LEO satellite. The developed IMM is shown to reduce the localization error and improve filter consistency over two fixed mismatched extended Kalman filters (EKFs). Starting with an initial receiver position error of 1.45 km, the IMM yielded a final error of 111.26 m, while the errors of a conservative and optimistic EKFs converged to 254.71 m and 429.35 m, respectively.
Topics & Concepts
Kalman filterEstimatorSatellitePosition (finance)CovarianceComputer scienceControl theory (sociology)Orbit (dynamics)Low earth orbitConsistency (knowledge bases)AlgorithmMathematicsPhysicsStatisticsEngineeringArtificial intelligenceAerospace engineeringAstronomyEconomicsFinanceControl (management)GNSS positioning and interferenceAdvanced Frequency and Time StandardsInertial Sensor and Navigation