Intelligent Reflecting Surfaces at Terahertz Bands: Channel Modeling and Analysis

TL;DR

This paper investigates the use of intelligent reflecting surfaces at terahertz frequencies, modeling spherical wave channels to analyze power gain and energy efficiency improvements in near-field MIMO systems, revealing the limitations of beamforming.

Contribution

It introduces a spherical wave channel model for IRS at THz bands and provides a comprehensive analysis of power gain and energy efficiency, highlighting the benefits of beamfocusing over beamforming.

Findings

Beamfocusing significantly improves energy efficiency in near-field IRS-assisted MIMO.

Beamforming is suboptimal for power gain when the IRS is several meters away.

Derived closed-form expression accurately predicts power gain loss under beamforming.

Abstract

An intelligent reflecting surface (IRS) at terahertz (THz) bands is expected to have a massive number of reflecting elements to compensate for the severe propagation losses. However, as the IRS size grows, the conventional far-field assumption starts becoming invalid and the spherical wavefront of the radiated waves should be taken into account. In this work, we consider a spherical wave channel model and pursue a comprehensive study of IRS-aided multiple-input multiple-output (MIMO) in terms of power gain and energy efficiency (EE). Specifically, we first analyze the power gain under beamfocusing and beamforming, and show that the latter is suboptimal even for multiple meters away from the IRS. To this end, we derive an approximate, yet accurate, closed-form expression for the loss in the power gain under beamforming. Building on the derived model, we next show that an IRS can significantly improve the EE of MIMO when it operates in the radiating near-field and performs beamfocusing. Numerical results corroborate our analysis and provide novel insights into the design and performance of IRS-assisted THz communication.