[Paper Review] MIMO Detection for Reconfigurable Intelligent Surface-Assisted Millimeter Wave Systems

Xi Yang, Chao-Kai Wen+1 more|arXiv (Cornell University)|Apr 13, 2020

Advanced Wireless Communication Technologies46 references4 citations

TL;DR

This paper proposes a reconfigurable intelligent surface (RIS)-assisted millimeter wave (mmWave) MIMO system using low-precision ADCs and discrete phase-shift RIS arrays to enhance spatial diversity and spectral efficiency. By leveraging linear spatial processing and a tailored MIMO detector, the system achieves robust performance even with 3-bit ADCs and 2-bit RIS phase shifts, demonstrating only moderate performance degradation compared to ideal hardware.

ABSTRACT

Millimeter wave (mmWave) band, or high frequencies such as THz, has large undeveloped band of spectrum. However, wireless channels over the mmWave band usually have one or two paths only due to the severe attenuation. The channel property restricts its development in the multiple-input multiple-output (MIMO) system, which can improve throughput by increasing the spectral efficiency. Recent development in reconfigurable intelligent surface (RIS) provides new opportunities to mmWave communications. In this study, we propose a mmWave system, which used low-precision analog-to-digital converters (ADCs), with the aid of several RIS arrays. Moreover, each RIS array has many reflectors with discrete phase shift. By employing the linear spatial processing, these arrays form a synthetic channel with increased spatial diversity and power gain, which can support MIMO transmission. We develop a MIMO detector according to the characteristics of the synthetic channel. RIS arrays can provide spatial diversity to support MIMO transmission, however, different number, antenna configuration, and deployment of RIS arrays affect the bit error rate (BER) performance. We present state evolution (SE) equations to evaluate the BER of the proposed MIMO detector in the different cases. The BER performance of indoor system is studied extensively through leveraging by the SE equations. We reveal numerous insights about the RIS effects and discuss the appropriate system settings. In addition, our results demonstrate that the low-cost hardware, such as the 3-bit ADCs of the receiver side and the 2-bit uniform discrete phase shift of the RIS arrays, only moderately degenerate the system performance.

Motivation & Objective

• To address the limited spatial diversity and high path loss in mmWave MIMO systems due to sparse scattering environments.
• To explore the feasibility of integrating low-precision hardware (3-bit ADCs, 2-bit RIS phase shifts) into RIS-assisted mmWave systems for cost and energy efficiency.
• To design a MIMO detection scheme tailored to the synthetic channel formed by RIS arrays, optimizing bit error rate (BER) performance.
• To analyze the impact of RIS array count, configuration, and deployment on system BER using state evolution (SE) equations.

Proposed method

• Proposes a mmWave MIMO system enhanced by multiple RIS arrays with discrete phase shift reflectors to create a synthetic channel with increased spatial diversity and path gain.
• Employs linear spatial processing at the receiver to combine signals from multiple RIS-assisted paths, forming a virtual MIMO channel.
• Develops a MIMO detector specifically designed for the synthetic channel structure, accounting for RIS-induced phase shifts and path gains.
• Derives state evolution (SE) equations to analytically evaluate BER performance across different RIS configurations and hardware precision levels.
• Uses Gaussian approximation of log-likelihood ratio (LLR) distributions to enable tractable MMSE estimation and BER prediction.
• Employs BER matching to calibrate noise precision in Gaussian approximations, ensuring accuracy in SE-based performance evaluation.

Experimental results

Research questions

• RQ1 How does the number and configuration of RIS arrays affect the BER performance in mmWave MIMO systems?
• RQ2 To what extent do low-precision ADCs (e.g., 3-bit) and low-resolution RIS phase shifts (e.g., 2-bit) degrade system performance?
• RQ3 Can RIS arrays effectively provide spatial diversity to support MIMO transmission in mmWave bands with sparse scattering?
• RQ4 How accurately can state evolution (SE) equations predict BER performance under varying RIS deployment and hardware constraints?
• RQ5 What are the optimal system settings (e.g., RIS array count, phase shift resolution) for balancing performance and hardware cost?

Key findings

• The proposed RIS-assisted mmWave MIMO system achieves significant spatial diversity and path gain through synthetic channel formation via multiple RIS arrays.
• With 3-bit ADCs and 2-bit RIS phase shifts, the system experiences only moderate performance degradation compared to full-precision counterparts.
• State evolution (SE) equations accurately predict BER performance across various RIS configurations, enabling system design optimization.
• The BER performance is highly sensitive to RIS array count and deployment geometry, with higher array counts improving diversity gain.
• Gaussian approximation of LLR distributions enables accurate MMSE estimation and BER prediction, validated through BER matching with simulation results.
• The system achieves robust performance even under low-resolution hardware, demonstrating strong potential for cost-effective mmWave deployment.