A Pragmatic Approach to Massive MIMO for Broadband Communication Satellites

Abstract

In this paper, we analyze the performance of Massive Multiple Input Multiple Output (M-MIMO) techniques aiming at increasing the throughput of broadband satellites. In particular, we investigate a 'pragmatic' approach to the design of the M-MIMO to ease its implementation both at system and satellite payload level. We compare the performance of optimized classical Minimum Mean Square Error (MMSE), Zero Forcing (ZF), Matched Filter (MF) schemes with the proposed 'pragmatic' M-MIMO one dubbed fixed Multi-Beam (MB). To further boost the M-MIMO performance, a novel radio resource management approach based on Mixed Integer Quadratic Programming (MIQDP-RRM) is proposed. The adoption of MIQDP-RRM is shown to greatly enhance the M-MIMO throughput performance. It is shown that the MB scheme closely approximate the MF, ZF and MMSE performance with a much simpler active antenna-based payload architecture and without requiring any user channel estimation. The MB MIQDP-RRM M-MIMO pragmatic solution allows to achieve higher satellite broadband throughput compared to a conventional four colors frequency reuse scheme (CFR) with affordable complexity for both space and ground segments. At the same time we show that by non-conventional CFR multi-beam array design, the performance gap compared to MB MIQDP-RRM M-MIMO can be significantly reduced.