Intelligent Reflecting Surface-Assisted Passive Covert Wireless Detection

In this work, we consider an intelligent reflecting surface (IRS)-assisted covert wireless detection system, where a transmitter aims to improve the detection accuracy of its transmission at a legitimate receiver, but avoid this transmission being detected by a warden. Considering Rician fading channels and a bounded noise uncertainty model, the detection performance at the legitimate receiver and warden is thoroughly analyzed. Then, the optimal design of the transmit power and IRS reflection matrix is proposed for maximizing the detection performance at the legitimate transceiver subject to a covertness constraint (i.e., the detection accuracy at the warden is less than a specific tolerable level). Our design is solely based on the line-of-sight (LoS) channel components and thus partially eliminates the need for acquiring the IRS's instantaneous channels, which takes a step towards the practical deployment of IRS. Our examination shows that IRS can still improve the covert detection performance even when only the LoS channel components are known and the improvement increases as the LoS components become more dominant.