Achievable Rate Analysis for Extra-Large RIS-Aided Massive MIMO with Visibility Regions

We investigate the achievable rate of extra-large reconfigurable intelligent surface (XL RIS) aided downlink massive multiple-input multiple-output (MIMO) in the presence of wholly/partial visibility regions, spatial correlation, and imperfectly estimated channel state information (CSI). To reduce high pilot overhead associated with XL RIS-aided massive MIMO, a statistical CSI-based phase-shift optimization technique is adopted to maximize the minimum sum of eigenvalues of user covariance matrices. To ensure user-fairness in the presence of wholly/partial visibility regions, a transmit power allocation technique is designed based on a max-min optimization criterion. We present numerical/simulation results to validate our analysis/optimization solutions, to investigate the effects of wholly/partial visibility regions, and to reveal the performance gain of the XL RIS-aided massive MIMO systems.