Bayesian-based Transmission and Channel Knowledge Map Reconstruction in LEO Satellite-terrestrial Communication

Due to limited spectrum resources, fine-grained spectrum monitoring is crucial for spectrum sharing and interference avoidance in Low Earth orbit (LEO) satellite networks, which exhibits superior coverage potential compared to terrestrial network infrastructures. Satellite-level channel awareness along with transmission recovery is important for the monitoring and management of spectrum resources. In this paper, we present an efficient satellite space-ground channel knowledge dictionary construction scheme based on sparse Bayesian learning (SBL), which aims to recover meteorological attenuation and terrestrial transmission information with limited sampling data. By exploiting the sparsity of transmission distributions and channel characteristics, an SBL-based spectrum data hierarchical reconstruction framework has been established. A meteorological attenuation estimation model is also devised to quantify the impact of weather conditions. Extensive evaluations show that our method achieves better reconstruction performance than competing baseline, especially with limited samples.