Performance and Optimization of Reconfigurable Intelligent Surface Aided THz Communications

TL;DR

This paper investigates the use of reconfigurable intelligent surfaces (RIS) to enhance THz communications, modeling signal fading accurately, proposing a swarm intelligence optimization method, and deriving key performance metrics showing capacity improvements.

Contribution

It introduces a novel RIS phase-shift optimization method without full channel estimation and provides analytical performance characterizations for THz systems.

Findings

Small-scale fading modeled by fluctuating two-ray distribution.

Increasing RIS elements significantly boosts ergodic capacity.

Analytical bounds for outage probability and capacity derived.

Abstract

TeraHertz (THz) communications can satisfy the high data rate demand with massive bandwidth. However, severe path attenuation and hardware imperfection greatly alleviate its performance. Therefore, we utilize the reconfigurable intelligent surface (RIS) technology and investigate the RIS-aided THz communications. We first prove that the small-scale amplitude fading of THz signals can be accurately modeled by the fluctuating two-ray distribution based on two THz signal measurement experiments conducted in a variety of different scenarios. To optimize the phase-shifts at the RIS elements, we propose a novel swarm intelligence-based method that does not require full channel estimation. We then derive exact statistical characterizations of end-to-end signal-to-noise plus distortion ratio (SNDR) and signal-to-noise ratio (SNR). Moreover, we present asymptotic analysis to obtain more insights when the SNDR or the number of RIS's elements is high. Finally, we derive analytical expressions for the outage probability and ergodic capacity. The tight upper bounds of ergodic capacity for both ideal and nonideal radio frequency chains are obtained. It is interesting to find that increasing the number of RIS's elements can significantly improve the THz communications system performance. For example, the ergodic capacity can increase up to 25% when the number of elements increases from 40 to 80, which incurs only insignificant costs to the system.