Blind Channel Estimation in Ambient Backscatter Communication Systems with Multiple-Antenna Reader

Abstract

Ambient backscatter exploits radio frequency (RF) signals to enable passive devices such as tags or sensors to communicate with readers, which has fascinating application for Internet of Things (IoT). The majority of the existing studies assume that channel state information (CSI) is perfectly acquired. Nevertheless, in ambient backscatter systems, the environmental RF signals are unknown at reader and thus may not serve as pilots. Therefore, the traditional channel estimators that routinely require pilots are not acceptable. In this paper, we focus on the problem of channel estimation with no pilots in ambient backscatter systems with multiple-antenna reader. Specifically, a blind channel estimator based on the eigenvalue decomposition (EVD) of the covariance matrix of the received signals is proposed, and the corresponding Cramér-Rao lower bounds (CRLBs) are derived. Simulation results are also provided to corroborate theoretical analysis.