Uplink Angle of Departure Estimation via Joint Sensing of NGSO Satellites

To mitigate co-frequency interference (CFI) with primary users (PUs), secondary systems use spectrum sensing to identify spatial and temporal spectrum holes. In satellite communications, directional antennas create sparsity in the angle domain, offering additional spectrum availability to the secondary system. Therefore, it becomes crucial for secondary systems to accurately understand the Uplink Angle of Departure (UL-AoD) or Downlink Angle of Arrival (DL-AoA) of the PU. However, in nongeostationary orbit (NGSO) systems, this information is time-varying and generally not shared with noncooperative secondary systems. To cope with this, we propose a novel UL-AoD estimation method. First, we leverage spot beams from secondary system satellites to jointly collect the PU’s signal. Then, a two-phase algorithm is designed to select a high-quality signal sample set from the collected signal samples and utilize the set to estimate the UL-AoD. Given the varying processing capabilities of secondary systems with respect to the signal of the primary system, we employ matched filtering (MF) to process the collected signal for scenarios with sufficient prior knowledge of the primary system and energy detection (ED) for scenarios with insufficient knowledge. Finally, combined with the ephemeris data of the primary system, the estimation result is then used to infer the most probable actual UL-AoD. Simulation results show the MF-based method approaches the Cramér-Rao lower bound (CRLB) and achieves error-free AoD estimation with fewer samples using ephemeris data. The ED-based method, with the assistance of the ephemeris data, attains over 85% AoD accuracy in large-scale constellations using more samples.