Abstract
Massive multiple-input multiple-output (m-MIMO) is considered as an essential technique to meet the high data rate requirements of future sixth generation (6G) wireless communications networks. The vast majority of m-MIMO research has assumed that the channels are uncorrelated. However, this assumption seems highly idealistic. Therefore, this study investigates the m-MIMO performance when the channels are correlated and the base station employs different antenna array topologies, namely the uniform linear array (ULA) and uniform rectangular array (URA). In addition, this study develops analyses of the mean square error (MSE) and the regularized zero-forcing (RZF) precoder under imperfect channel state information (CSI) and a realistic physical channel model. To this end, the MSE minimization and the spectral efficiency (SE) maximization are investigated. The results show that the SE is significantly degraded using the URA topology even when the RZF precoder is used. This is because the level of interference is significantly increased in the highly correlated channels even though the MSE is considerably minimized. This implies that using a URA topology with relatively high channel correlations would not be beneficial to the SE unless an interference management scheme is exploited.
Highlights
This study explores the impact of having arbitrary array geometries on the performance of massive multiple-input multiple-output (m-MIMO) systems by using different array topologies, namely uniform linear array (ULA) and uniform rectangular array (URA)
The results show that a substantial improvement in the spectral efficiency (SE) performance of the m-MIMO system can be achieved with a ULA topology in comparison to the performance obtained using a URA topology
The results are presented for the ULA and URA with realistic configurations in order to emphasize the importance of our results in a realistic channel model
Summary
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Investigating mMIMO systems in a more general fading scenario, which takes into account the correlation conditions, is an interesting research topic that should be considered This motivates us to investigate the m-MIMO system performance in terms of mean square error (MSE) and SE in a more realistic scenario where the channels are correlated. This study seeks to answer the following question: Which array configuration maximizes the SE of m-MIMO systems For this reason, this study considers a Laplacian physical channel model to design the channel covariance matrices with distinct array geometries. The results demonstrate that the SE is significantly degraded using the URA topology even when the RZF precoder is used This is because the level of interference is significantly increased in the highly correlated scenario. CN (μ, G) denotes the Gaussian distribution with mean μ and covariance G
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
