Reconfigurable Intelligent Surface Relying on Low-complexity Joint Sector Non-diagonal Structure

In recent years, research on reconfigurable intelligent surface (RIS) has received extensive attention due to its capability to manipulate the propagation of incident electromagnetic waves in a programmable manner to smartly configure the channel environment, thereby optimizing the overall performance of the system. Several RIS architectures have been proposed, including simultaneous transmitting and reflecting RIS (STAR-RIS) and beyond diagonal RIS (BD RIS) architectures. Compared to the conventional RIS architecture, these structures offer broader service ranges and enhanced performance, albeit accompanied by more complex circuit design and higher computational overhead. In this paper, we design a multi-sector RIS joint service model based on the BD RIS architecture and we compare the corresponding system rates, circuit complexity, and gains provided by different architectures. Additionally, we derive the theoretical receive power for the proposed model based on non-diagonal and diagonal phase shift optimization methods, demonstrating that the total rate of the non-diagonal group connected architecture approaches the theoretical values of a fully connected architecture. Theoretical analysis based on the gains between different models under various user-RIS positions confirm that the multi-sector RIS joint service model can achieve
$30\%-100\%$
gains as the users' positions change, while also saving on the overall hardware costs of the RIS system design. Furthermore, we explore the optimal trade-off between the performance and circuit complexity among different architectures. Simulation results show that performance versus complexity trade-off of the different considered architectures.