A Low-Complexity Hybrid Precoding for Wideband Satellite Communication Systems

This paper investigates a hybrid beamforming structure for wideband satellite communication (SatCom) systems and proposes a novel compression-based hybrid precoding (CHP) method for low computational complexity. Unlike cellular systems with multipath fading channels, SatCom systems, which have a single path channel, show high correlation between precoding vectors generated at adjacent frequencies. The key idea of our approach is to leverage this unique correlation characteristic. By leveraging this, we first derive two compressed precoding matrices that have marginal compression loss and are significantly smaller in size compared to the original precoding matrix. We then design the analog and digital precoding matrices to implement these compressed matrices, considering the constant modulus constraint and computational complexity in the base-band processing. Simulation results verify the superiority of the proposed solution in terms of average spectral efficiency, system throughput, and computational complexity. In particular, the results show that the proposed solution can reduce the computational complexity to 15.5% of what is required by existing methods.