Channel Estimation for MIMO System Assisted by Intelligent Reflective Surface

Gilderlan Tavares de Araújo, Lucas Campos de Paula Pessoa, André de Almeida

DOI: 10.14209/SBRT.2020.1570661259

Evento: XXXVIII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2020)

Keywords: Intelligent Reflective Surface Channel Estimation MIMO Khatri-Rao Factorization

Abstract

Intelligent reflective surface (IRS) has being envisioned to be the key technology for beyond 5G or 6G systems. Due to the passive nature of the IRS, channel estimation is one of the main challenges in IRS-based communications. In addition, due to hardware constraints, a perfect reflection cannot always be achieved by the IRS. In this paper, we face the channel estimation problem in a multiple-input and multiple-output (MIMO) communication system assisted by an IRS, where a base station (BS) communicates with an user terminal (UT) via an IRS panel. We discuss two channel estimation schemes. The first is based on the least squares (LS) estimator, while the second adds an extra step based on the Khatri-Rao factorization (KRF) algorithm to achieve separate estimates of the BS-IRS and IRS-UT channels via rank-1 approximation steps. By using simplified models to capture a non-perfect IRS reflection, we numerically evaluate the performance of the two channel estimation schemes and discuss their normalized mean square error (NMSE) performance for some scenarios, including the effect of quantized IRS phase shifts and non-constant reflection amplitudes.

Download