IRS-Assisted Crowd Spectrum Sensing in B5G Cellular IoT Networks
Xiaohui Li, Qi Zhu, Ying Wang
Abstract
Crowd spectrum sensing (CSS) is a promising paradigm for discovering real-time radio resource availability with higher flexibility. Conventional CSS schemes encounter difficulties in the best trade-off between high accuracy of the detection result and the less consumptions of energy and network resources. To solve this problem, this paper utilizes the advanced intelligent reflecting surface (IRS) to improve the spectrum detection performance meanwhile reduce the energy consumption. IRS is envisioned to be equipped with intelligent internet of things (IoT) devices in the B5G cellular IoT networks. Therefore, it is applicable into the CSS scheme for timely and accurately detecting spectrum opportunities with crowd collaborations. In this paper, a crowd of IRS units reflect incident primary signal to the requestor, who then fuses all the reflected signals and makes final spectrum decision. The proposed IRS-assisted CSS scheme is in essence a soft-fusion scheme due to the direct reflection of the original primary signal, which thus guarantees the objectivity of spectrum information and the accuracy of detection result. Furthermore, compared with conventional soft-fusion based CSS, the IRS-assisted scheme enables neither noises nor energy consumptions for data reporting. This paper illustrates the system mole and derives the closed expressions of the crowd detection performance metrics. Simulation results verify the theoretical correctness and demonstrate the superior performance of the newly proposed IRS-assisted CSS scheme.