On the Spectral Efficiency of D-MIMO Networks under Rician Fading

TL;DR

This paper evaluates the spectral efficiency of Distributed-MIMO networks under Rician fading, revealing how channel correlation affects performance and identifying optimal configurations for antenna deployment.

Contribution

It provides a comprehensive analysis of D-MIMO performance under Rician fading, highlighting the impact of channel correlation and optimal antenna arrangements.

Findings

Channel correlation increases with the Rician factor, reducing spectral efficiency.

An optimal number of APs and antennas per AP maximizes performance.

Performance depends on the Rician factor and receive combining scheme.

Abstract

Contemporary wireless communications systems adopt the Multi-User Multiple-Input Multiple-Output (MU-MIMO) technique: a single base station or Access Point (AP) equipped with multiple antenna elements serves multiple active users simultaneously. Aiming at providing a more uniform wireless coverage, industry and academia have been working towards the evolution from centralized MIMO to Distributed-MIMO. That is, instead of having all the antenna elements co-located at a single AP, multiple APs, each equipped with a few or a single antenna element, jointly cooperate to serve the active users in the coverage area. In this work, we evaluate the performance of different D-MIMO setups under Rician fading, and considering different receive combining schemes. Note that the Rician fading model is convenient for MU-MIMO performance assessment, as it encompasses a wide variety of scenarios. Our numerical results show that the correlation among the channel vectors of different users increases with the Rician factor, which leads to a reduction on the achievable Spectral Efficiency (SE). Moreover, given a total number of antenna elements, there is an optimal number of APs and antenna elements per AP that provides the best performance. This "sweet spot" depends on the Rician factor and on the adopted receive combining scheme.