On the Ergodic Capacity of Reconfigurable Intelligent Surface (RIS)-Aided MIMO Channels

TL;DR

This paper analyzes the ergodic capacity of RIS-assisted MIMO channels, deriving bounds and asymptotic behaviors to understand how RIS deployment and the number of reflecting elements influence spectral efficiency.

Contribution

It provides novel analytical expressions for the bounds of ergodic channel capacity in RIS-aided MIMO systems, including asymptotic analysis for large SNR and REs.

Findings

ECC can be improved by increasing the number of reflecting elements.

RIS deployment influences the high-SNR power offset of ECC.

Asymptotic analysis reveals capacity behavior in large SNR and RE regimes.

Abstract

Reconfigurable intelligent surfaces (RISs) have emerged as a promising technique to enhance the system spectral efficiency. This paper investigates the ergodic channel capacity (ECC) of an RIS-aided multiple-input multiple-output channel under the assumption that the transmitter-RIS, RIS-receiver, and transmitter-receiver channels contain deterministic line-of-sight paths. Novel expressions are derived to characterize the upper and lower bounds of the ECC. To unveil more system insights, asymptotic analyses are performed to the system ECC in the limit of large signal-to-noise ratio (SNR) and number of reflecting elements (REs). Theoretical analyses suggest that the RIS's deployment can shape the ECC curve by influencing its high-SNR power offset and the ECC can get improved by increasing the number of REs.