[Paper Review] Hybrid Relay-Reflecting Intelligent Surface-Aided Wireless Communications: Opportunities, Challenges, and Future Perspectives

Nhan T. Nguyen, Jiguang He+6 more|arXiv (Cornell University)|Apr 5, 2021

Advanced Wireless Communication TechnologiesEngineering15 references19 citations

TL;DR

This paper proposes a hybrid relay-reflecting intelligent surface (HR-RIS) architecture that integrates a few active elements with passive reflecting elements to enhance beamforming, enable efficient channel estimation, and improve spectral and energy efficiency. Numerical results show HR-RISs with just four active elements achieve up to 42.8% higher spectral efficiency and 41.8% higher energy efficiency than conventional RISs.

ABSTRACT

Reconfigurable intelligent surfaces (RISs) have emerged as a cost- and energy-efficient technology that can customize and program the physical propagation environment by reflecting radio waves in preferred directions. However, the purely passive reflection of RISs not only limits the end-to-end channel beamforming gains, but also hinders the acquisition of accurate channel state information for the phase control at RISs. In this paper, we provide an overview of a hybrid relay-reflecting intelligent surface (HR-RIS) architecture, in which only a few elements are active and connected to power amplifiers and radio frequency chains. The introduction of a small number of active elements enables a remarkable system performance improvement which can also compensate for losses due to hardware impairments such as the deployment of limited-resolution phase shifters. Particularly, the active processing facilitates efficient channel estimation and localization at HR-RISs. We present two practical architectures for HR-RISs, namely, fixed and dynamic HR-RISs, and discuss their applications to beamforming, channel estimation, and localization. The benefits, key challenges, and future research directions for HR-RIS-aided communications are also highlighted. Numerical results for an exemplary deployment scenario show that HR-RISs with only four active elements can attain up to 42.8 percent and 41.8 percent improvement in spectral efficiency and energy efficiency, respectively, compared with conventional RISs.

Motivation & Objective

• To address the performance limitations of conventional reconfigurable intelligent surfaces (RISs), such as finite-resolution phase shifters and purely passive reflection, which restrict beamforming gains and channel estimation accuracy.
• To propose a hybrid relay-reflecting intelligent surface (HR-RIS) architecture that incorporates a small number of active elements with power amplifiers and RF chains to enhance system performance.
• To enable efficient channel estimation, localization, and beamforming through active processing while maintaining low hardware cost and energy consumption.
• To explore practical architectures—fixed and dynamic HR-RISs—for real-world deployment and system integration.
• To identify key challenges and future research directions for deploying HR-RIS-aided wireless networks at scale.

Proposed method

• The HR-RIS architecture integrates a few active reflecting elements connected to power amplifiers and RF chains, enabling both signal reflection and amplification, unlike purely passive RISs.
• The signal model represents the HR-RIS as a linear mapping that combines passive reflection and active relaying, with the active elements functioning as full-duplex amplify-and-forward (AF) relays.
• Two architectures are proposed: fixed HR-RIS, where active elements are pre-configured, and dynamic HR-RIS, which allows adaptive reconfiguration based on channel conditions.
• Joint optimization of beamforming at the base station, mobile station, and HR-RIS is considered to maximize spectral and energy efficiency, though it introduces high computational complexity.
• Artificial intelligence (AI)-based methods are proposed to address non-convex, high-dimensional optimization problems with low latency and complexity.
• Data-driven AI techniques are explored for pilot-free channel estimation and localization, though robustness to dynamic channels remains a challenge.

Experimental results

Research questions

• RQ1 How can a hybrid relay-reflecting intelligent surface (HR-RIS) architecture overcome the limitations of purely passive RISs in terms of beamforming gain and channel estimation accuracy?
• RQ2 What are the performance gains of HR-RISs compared to conventional RISs in terms of spectral efficiency (SE) and energy efficiency (EE), especially with minimal active elements?
• RQ3 How do fixed and dynamic HR-RIS architectures differ in their design, implementation, and application to beamforming and localization?
• RQ4 What are the key challenges in deploying HR-RIS-aided systems, particularly regarding computational complexity, synchronization, and control signaling?
• RQ5 Can AI-based methods enable robust, low-complexity, and pilot-free channel estimation and localization in dynamic HR-RIS environments?

Key findings

• HR-RISs with only four active elements achieve up to 42.8% higher spectral efficiency and 41.8% higher energy efficiency compared to conventional RISs in an exemplary deployment scenario.
• The integration of active elements enables effective beamforming and significantly improves end-to-end channel beamforming gains, overcoming limitations of passive RISs.
• Active processing at HR-RISs facilitates accurate and efficient channel estimation and localization, which are challenging in purely passive RIS systems.
• The proposed fixed and dynamic HR-RIS architectures offer practical trade-offs between performance, complexity, and adaptability for real-time applications.
• Joint optimization of beamforming across base stations, mobile stations, and HR-RISs is essential to fully exploit active beamforming gains, though it introduces high computational complexity.
• AI-based solutions show promise for low-latency, sub-optimal solutions to complex HR-RIS optimization problems, though robustness to dynamic channels remains a key challenge.