RIS Deployment Optimization with Iterative Detection and Decoding in Multiuser Multiple-Antenna Systems

TL;DR

This work addresses optimizing RIS-assisted uplink IDD in a multiuser MU-MISO setting by first deriving analytical insights for active and passive RIS in a SISO baseline and then validating them with simulations in a MU-MIMO context. It adopts an MMSE-based reflection parameter design and an iterative detection and decoding framework that exchanges extrinsic information between a soft detector and an LDPC decoder, yielding per-user SINR $\gamma_k^{sic}$ and sum-rate $R_{sum}$. The key contributions include MMSE-based reflection optimization with relaxation and truncation rules for both RIS types, a SIC-enhanced detection scheme, and evidence that SISO-derived insights extend to MU-MIMO configurations (up to $K=12$) with notable BER and sum-rate gains. The results offer practical deployment guidance on RIS placement (active vs passive) and parameter tuning (numbers of users, AP antennas, and RIS elements) to achieve robust performance in real-world 6G scenarios.

Abstract

This work investigates a Reconfigurable Intelligent Surface (RIS)-assisted uplink system employing iterative detection and decoding (IDD) techniques. We analyze the impact of tuning system parameter tuning for several deployment configurations, including the number of users, access point (AP) antennas, and RIS elements on the IDD performance. Analytical results for both active and passive RIS in a single-input single-output (SISO) scenario demonstrate how deployment choices affect system performance. Numerical simulations confirm the robustness of the RIS-assisted IDD system to variations in these parameters, showing performance gains in certain configurations. Moreover, the findings indicate that the insights derived from SISO analysis extend to multiuser MIMO IDD systems.

Figures (8)

• Figure 1: System model of an IDD multiuser multiple-antenna system.
• Figure 2: Block Diagram of RIS Deployment in SISO Systems
• Figure 3: Performance for $K = 12$, $M = 32$, $N = 64$, $\sigma_s^2 = -100$ dBm, $\sigma_v^2 = 0$ dBm and $P_T/K = 6$ dBm.
• Figure 4: Uplink performance for $K = 12$, $M = 32$, $N = 64$, $\sigma_s^2 = -95$ dBm, $\sigma_v^2 = -95$ dBm and $P_T/K = 0$ dBm.
• Figure 5: Performance for $M = 32$, $N = 64$, $\sigma_s^2 = -100$ dBm, $\sigma_v^2 = 0$ dBm and $P_T/K = 6$ dBm.