Autonomous and Decentralized Orbit Determination and Clock Offset Estimation of Lunar Navigation Satellites Using GPS Signals and Inter-Satellite Ranging

The development of a robust navigation infrastructure in cis-lunar space is crucial for the coming new era of advanced lunar exploration. NASA has set a goal to develop an extensible and scalable lunar communication and navigation architecture called LunaNet. For the flexibility and robustness of the system, the architecture is desired to have an autonomous and decentralized operation capability. This paper presents and demonstrates a decentralized and autonomous state estimation algorithm for SmallSats that provides positioning, navigation, and timing service, each equipped with a GNSS receiver, chip-scale atomic clock, and inter-satellite communication module. In the framework, each satellite individually estimates its own state and clock offset with a decentralized Schmidt Extended Kalman Filter by processing weak GNSS signal and inter-satellite range measurements. The algorithm is validated with Monte-Carlo state estimate simulations of five satellites at the hybrid constellation of lunar frozen orbit and the Near-Rectilinear Halo Orbit (NRHO). The proposed filter converged in all Monte-Carlo cases with clock offset errors below 0.15 microseconds and 1.0 microseconds in the frozen orbit and the NRHO, respectively.