Angle-Domain Channel and DOA Estimation in Airborne Massive MIMO System with Eavesdropper

Abstract

In this paper, we consider a hybrid air-terrestrial network secure transmission scenario, where a low-altitude air platform (AP) equipped with a two-dimensional (2D) rectangular antenna array serves a set of legitimate users (LUs). A single-antenna vehicle eavesdropper is within the courage of the AP, and the eavesdropper sends jamming signals in the uplink training and transmission phase and eavesdrops on LUs signals in the downlink transmission phase. During the training phase, a angle-domain (AD) channel estimation scheme is proposed. The active beam sets (ABS) of the interference channel and the LU channels can be separately extracted by comparing the beam gain vectors from different LUs. The pilot contamination can be eliminated and the channel state information (CSI) of LUs can be purified. By using spatial basis expansion model (BEM), full-dimension CSI of LUs and ED can be obtained, based on which the digital domain angle alignment method is designed to locate the ED. The superiority of the proposed scheme is evaluated by simulations.