Dual-Hop LEO Satellite-Terrestrial Communication With Joint Beamforming, Movable-Antenna Control, and Nonlinear Energy Harvesting

This paper investigates a dual-hop low-Earth-orbit (LEO) satellite to ground communication system that employs rate-splitting multiple access (RSMA), multi-antenna transmission at both the satellite and ground stations (GSs), movable antennas at the GSs, and a practical nonlinear radio-frequency (RF) energy-harvesting model at the users. The end-to-end throughput of such systems is fundamentally constrained by the minimum rate across the satellite-to-GS and GS-to-user hops, creating a tightly coupled design problem in which beamforming, common private stream allocation, antenna locations, and power-splitting ratios must be jointly optimized. To address this challenge, we formulate an end-to-end sum rate maximization problem under rate, power, movement, and energy-harvesting constraints. Auxiliary variables are introduced to decouple the hop-dependent rates, and the overall problem is solved through an alternating optimization framework that incorporates semidefinite relax ation, successive convex approximation, and penalty-based rank one refinement for the precoding blocks, as well as convex position and power-splitting updates. Simulation results verify the effectiveness of the proposed design and show that optimizing the GS antenna locations provides significant performance benefits. In particular, the results demonstrate a sum rate gain of up to 41% compared with systems employing fixed antenna locations.