부산대학교 도서관

We investigate the use of passive and active intelligent reflecting surface (IRS) to detect the presence of a jammer in a wireless communication system. We use a generalized likelihood ratio test and exploit pilots not assigned to any user to decide on the presence of a jammer. We deduce analytical expressions for the probability of correct detection and the probability of false alarm. We observe that the detection performance improves for a given false alarm probability as the number of IRS elements increases or the number of training samples increases. We show that by deploying more IRS elements, the target probability of correct detection can be achieved with fewer training samples. We further show that a passive IRS performs well when the direct channel between jammer/user and the access point is blocked, but an active IRS is required when the direct channel is present to achieve significant performance gains.