Spatial separation of closely-spaced users in measured distributed massive MIMO channels

TL;DR

This paper evaluates the spatial separation capabilities of distributed massive MIMO systems in indoor 6G scenarios through measurements at 5.6 GHz, demonstrating improved user orthogonality and capacity with distributed antennas, especially in line-of-sight conditions.

Contribution

It provides an experimental assessment of distributed massive MIMO's ability to separate closely spaced users, highlighting the effectiveness of distributed antenna topologies in different propagation conditions.

Findings

Distributed MIMO achieves better user orthogonality in LOS conditions.

Zero forcing precoding performs close to DPC capacity with fewer complexities.

More antennas and distributed topologies improve capacity and fairness in LOS, limited in NLoS.

Abstract

Aiming for the sixth generation (6G) wireless communications, distributed massive multiple-input multiple-output (MIMO) systems hold significant potential for spatial multiplexing. In order to evaluate the ability of a distributed massive MIMO system to spatially separate closely spaced users, this paper presents an indoor channel measurement campaign. The measurements are carried out at a carrier frequency of 5.6 GHz with a bandwidth of 400 MHz, employing distributed antenna arrays with a total of 128 elements. Multiple scalar metrics are selected to evaluate spatial separability in line-of-sight, non line-of-sight, and mixed conditions. Firstly, through studying the singular value spread, it is shown that in line-of-sight conditions, better user orthogonality is achieved with a distributed MIMO setup compared to a co-located MIMO array. Furthermore, the dirty-paper coding (DPC) capacity and zero forcing (ZF) precoding sum-rate capacities are investigated across varying numbers of antennas and their topologies. The results show that in all three conditions, the less complex ZF precoder can be applied in distributed massive MIMO systems while still achieving a large fraction of the DPC capacity. Additionally, in line-of-sight conditions, both sum-rate capacities and user fairness benefit from more antennas and a more distributed antenna topology. However, in the given NLoS condition, the improvement in spatial separability through distributed antenna topologies is limited.