Massive MIMO Uplink Transmission for Multiple LEO Satellite Communication

In this article, we investigate massive multiple-input–multiple-output (MIMO) uplink (UL) transmission for multiple low-Earth-orbit satellite communication. The signal and channel models are established for the UL massive MIMO multisatellite system. We reveal that the signals transmitted from the user terminals (UTs), which are intended for different satellites, are typically asynchronous at each satellite. We propose linear precoding with frequency and time precompensations at each UT for each satellite so that the signals intended for a specific satellite from different UTs can be synchronized. We formulate the ergodic sum rate maximization problem by considering the minimum mean square error successive interference cancellation receiver at each satellite with perfect instantaneous channel state information (iCSI). We propose a Riemannian conjugate gradient (RCG)-based precoding vector design algorithm within the manifold optimization framework, which does not include any matrix inversion. Then, in order to reduce the complexity in implementation at satellites, we reformulate the UL ergodic sum rate maximization problem by considering the linear receiving at each satellite with statistical CSI (sCSI) only. We devise an RCG-based joint precoding and receiving vector design to maximize the sum rate, which avoids any matrix inversion, thereby reducing the computational complexity. Simulation results indicate the effectiveness of our proposed approaches.