Information Security in Intelligent Reflecting Surface-Aided Two-way Network

Abstract

In this paper, intelligent reflecting surface (IRS), containing a large number of low-cost passive reflection elements, has assisted to improve secure and efficient two-way transmission. In the considered network, two transceivers simultaneously transmit and receive information with the aid of one IRS. The artificial noises (AN) are injected to protect information from the attack of an unknown eavesdropper. The aim is to minimize the signal transmission power, i.e., maximize the AN signal power, under the signal-to-interference-plus-noise (SINR) threshold constraints and the total transmission power of the transceivers. The alternating optimization and semidefinite relaxation methods are combined to find the two variable groups of beamformers, AN covariance matrices, and IRS phase shifts. The numerical simulation presents the power gain of the proposed IRS-aided algorithm compared to the no-IRS baseline scheme.