Joint and Streamwise Distributed MIMO Satellite Communications with Multi-Antenna Ground Users

TL;DR

This paper explores distributed MIMO satellite communication strategies with multi-antenna ground users, proposing joint and streamwise transmission modes, optimizing spectral efficiency, and demonstrating significant performance improvements through simulations.

Contribution

It introduces a novel iterative transceiver design for joint transmission and an eigenmode-based association for streamwise transmission, addressing fronthaul constraints in satellite MIMO systems.

Findings

Joint transmission achieves higher spectral efficiency with full fronthaul.

Streamwise transmission reduces fronthaul load with near-optimal performance under certain conditions.

Numerical results show substantial gains over conventional methods.

Abstract

We consider a low Earth orbit downlink communication, where multiple satellites jointly serve multi-antenna ground users, transmitting multiple spatial streams per user. Using a line-of-sight-dominant satellite channel model with statistical channel state information, including angular information and large-scale fading, we study two distributed transmission modes with different fronthaul requirements. First, for joint transmission, where all satellites transmit all user streams, we formulate a sum spectral efficiency (SE) maximization problem under general convex power constraints and address the intractability of the exact ergodic SE expression by adopting a tractable approximation. Exploiting the equivalence between sum SE maximization and weighted sum mean square error minimization, we derive a novel iterative transceiver design. Second, to reduce fronthaul load, we propose streamwise transmission, where each stream is sent by a single satellite, and develop an eigenmode-based stream-satellite association using participation factors and a maximum-weight bipartite matching problem solved by the Hungarian algorithm. Numerical simulations evaluate the validity of the SE approximation, demonstrate conditions under which streamwise transmission performs nearly optimally or trades SE for lower overhead, highlight the impact of stream/user loading, and show substantial performance gains over conventional benchmarks.