Multi‐satellite autonomous orbit determination by game‐theoretic optimization in space service volume

Global navigation satellite systems (GNSSs) are increasingly used for multi‐satellite missions in space service volume (SSV). However, the performance of GNSS‐based autonomous orbit determination (AOD) for SSV users is severely constrained by limited measurement accuracy and signal availability. Relative measurement and orbital dynamic models can be used to improve the AOD performance when the estimator is prone to divergence due to the cross‐correlation of the distributed formation network. In this paper, a sequential game‐theoretic (SGT) optimization algorithm is proposed to deal with cross‐correlation issues when fusing the GNSS and relative measurements for dynamic AOD in the SSV field. A four geostationary Earth orbit satellite mission is simulated as the case study for the AOD performance analysis. The conventional sequential covariance intersection (SCI) is also presented for comparison purposes. The simulation results show that with the combination of GNSS and relative measurements with 10 m and 0.1 m precision, the SGT‐based AOD can achieve about 1.9 m three‐dimensional positioning accuracy, which is 80% better than in the GNSS‐only case. It is also demonstrated that the SGT‐based AOD is superior to the SCI‐based AOD by 50% in the fusion of GNSS and relative measurements in the range of 1–10 m and with 0.1–0.001 m accuracy.