A Hybrid Approach for GPS-Based Relative Navigation of Formation Flying Satellites in Remote Sensing Mission

The paper deals with the use of dual-frequency Carrier-phase Differential GPS for autonomous relative navigation of formation flying satellites. The problem is specifically analyzed for remote sensing applications in which high accuracy in the knowledge of the relative positions is required and the inter-satellite distance largely varies during the mission due to scientific application requirements and orbital constrains. To guarantee high accuracy and robustness a hybrid approach is proposed by combining in cascade an augmented-state Extended Kalman Filter with a kinematic filter. Specifically, the dynamic filter provides robust floating estimates of double-difference GPS ambiguities, which are then processed in the kinematic filter. The approach is numerically tested considering relative orbit scenarios relevant to existing or proposed remote sensing applications. Results show that the proposed filtering scheme is robust and provides estimation accuracy ranging from the millimetre to the centimetre level as the baseline grows from a few to hundreds kilometres.