[Paper Review] Joint Training of the Superimposed Direct and Reflected Links in Reconfigurable Intelligent Surface Assisted Multiuser Communications

Jiancheng An, Chao Xu+4 more|arXiv (Cornell University)|May 30, 2021

Advanced Wireless Communication Technologies4 citations

TL;DR

This paper proposes a novel joint training framework for Reconfigurable Intelligent Surface (RIS)-assisted multiuser communications that estimates the superimposed direct and reflected end-to-end channels in multiple training periods, enabling optimized RIS phase shifts without separate channel estimation. The method reduces pilot and signaling overhead while achieving near-optimal performance, especially at low SNR, as validated by theoretical analysis and simulations in SISO and MIMO scenarios.

ABSTRACT

In Reconfigurable intelligent surface (RIS)-assisted systems the acquisition of CSI and the optimization of the reflecting coefficients constitute a pair of salient design issues. In this paper, a novel channel training protocol is proposed, which is capable of achieving a flexible performance vs. signalling and pilot overhead as well as implementation complexity trade-off. More specifically, first of all, we conceive a holistic channel estimation protocol, which integrates the existing channel estimation techniques and passive beamforming design. Secondly, we propose a new channel training framework. In contrast to the conventional channel estimation arrangements, our new framework divides the training phase into several periods, where the superimposed end-to-end channel is estimated instead of separately estimating the direct BS-user channel and cascaded reflected BS-RIS-user channels. As a result, the reflecting coefficients of the RIS are optimized by comparing the objective function values over multiple training periods. Moreover, the theoretical performance of our channel training protocol is analyzed and compared to that under the optimal reflecting coefficients. In addition, the potential benefits of our channel training protocol in reducing the complexity, pilot overhead as well as signalling overhead are also detailed. Thirdly, we derive the theoretical performance of channel estimation protocols and our channel training protocol in the presence of noise for a SISO scenario, which provides useful insights into the impact of the noise on the overall RIS performance. Finally, our numerical simulations characterize the performance of the proposed protocols and verify our theoretical analysis. In particular, the simulation results demonstrate that our channel training protocol is more competitive than the channel estimation protocol at low signal-to-noise ratios.

Motivation & Objective

• To address the challenge of acquiring accurate channel state information (CSI) in RIS-aided systems where conventional pilot-based estimation is infeasible due to passive RIS elements.
• To reduce pilot and signaling overhead in RIS-assisted multiuser downlink communications by avoiding separate estimation of direct and cascaded channels.
• To enable joint optimization of RIS reflecting coefficients through a new training protocol that estimates the superimposed end-to-end channel across multiple periods.
• To theoretically analyze the performance of the proposed training protocol under noise and compare it with the optimal reflecting coefficient design.
• To demonstrate the superiority of the proposed method over conventional channel estimation in low SNR regimes through simulations and analytical derivations.

Proposed method

• Introduces a holistic channel estimation protocol that integrates passive beamforming design with channel estimation, avoiding separate estimation of direct and cascaded channels.
• Designs a new training framework that divides the training phase into multiple periods, during which the superimposed end-to-end channel (direct + reflected) is estimated collectively.
• Optimizes RIS reflecting coefficients by comparing objective function values (e.g., signal power or rate) across multiple training periods, enabling adaptive beamforming.
• Derives theoretical expressions for the average end-to-end channel gain under channel estimation errors, incorporating complex Gaussian noise models for direct and reflected links.
• Uses complex Gaussian distributions to model channel estimation errors, enabling derivation of closed-form expectations for key performance metrics.
• Applies analytical tools such as expectation of real and imaginary parts of normalized channel estimates to derive the final performance expression in Equation (66).

Experimental results

Research questions

• RQ1 Can a joint training protocol that estimates the superimposed direct and reflected channels reduce pilot and signaling overhead compared to conventional separate estimation?
• RQ2 How does the performance of the proposed joint training protocol compare to the optimal reflecting coefficient design under realistic channel estimation errors?
• RQ3 What is the theoretical impact of noise on the average end-to-end channel gain in the proposed framework?
• RQ4 Does the proposed method achieve better performance than conventional channel estimation at low signal-to-noise ratios (SNR)?
• RQ5 To what extent does the joint training framework maintain performance while reducing implementation complexity in RIS-aided multiuser systems?

Key findings

• The proposed joint training protocol achieves near-optimal performance with significantly reduced pilot and signaling overhead compared to conventional methods.
• Theoretical analysis shows that the average end-to-end channel gain under estimation errors is given by Equation (66), which accounts for both direct and reflected link power and interference terms.
• The protocol demonstrates superior performance at low SNR, outperforming conventional channel estimation techniques in numerical simulations.
• The derived expression in Equation (66) reveals that the performance gain is proportional to the number of RIS elements N and depends on the ratio of channel power to estimation error variance.
• The real part of the normalized channel estimate contributes to the average channel gain, while the imaginary part averages to zero, simplifying performance analysis.
• Simulation results confirm that the proposed method maintains robust performance across various SNR regimes and offers a flexible trade-off between performance, overhead, and complexity.