Multi-Antenna UAV Attitude Determination Method Based on LEO Satellite Communication Signal

With advancements in low Earth orbit (LEO) communication technology, the potential for an attitude determination system utilizing LEO satellite constellations has been explored. We introduce a method for Uncrewed Air Vehicle (UAV) attitude determination employing multiple antennas and sufficient LEO communication downlink signals. By estimating the time difference of arrival (TDOA) of signals with various encoding and modulation schemes received by different antennas, the attitude of the UAV can be determined. We propose an attitude determination framework based on LEO communication signals, explaining the relationships between satellite distribution, antenna baseline distribution, and LEO signal characteristics. We derive single-epoch and multi-epoch attitude determination algorithms and the Cramér-Rao lower bound (CRLB) from this framework. The findings indicate that the highest accuracy of attitude measurement is achieved with the orthogonal main baselines and increases with the number of redundant baselines. With both baselines at 1 m, a signal bandwidth of 20 MHz and SNR of -20 dB, an accuracy better than 1$^{\circ }$ can be achieved. Moreover, the multi-epoch factor graph optimization (FGO)-based attitude estimation algorithm significantly outperforms the single-epoch algorithm.