Spatially Correlated Channels Investigation: Estimation and Hardening in Millimeter-Wave Massive MIMO Systems

This study delves into the channel estimation (CE) process within Millimeter-wave M-MIMO networks, acknowledging the real-world scenario of spatially correlated channels. The presence of spatial correlation (SC) significantly impacts the outcomes of Millimeter-wave M-MIMO systems, with SC being influenced by various factors, notably the array arrangement of base stations (BSs). Consequently, we systematically explore the effects of SC on CE, considering different array configurations. To characterize SC, we employ a Gaussian Local Multi-Scatterer (GLMS) model. We also investigate the outcomes of the MMSE estimator in the context of correlated and uncorrelated channels. In this endeavor, we extend the application of the GLMS model beyond the uniform linear configuration (ULC) to propose a tailored GLMS model for the uniform circular configuration (UCC). Furthermore, we introduce a GLMS model customized for the UCC, underpinned by a theoretical exposition to substantiate its validity. Our investigation encompasses an evaluation of channel hardening for both ULC and UCC configurations. To assess system outcomes, we employ the Normalized Mean Square Error metric, substantiating our theoretical analyses with simulation results. This comprehensive approach provides insights into the impact of SC on CE within Millimeter-wave M-MIMO networks across various array arrangements.