Resource Allocation for IRS-assisted Cognitive Radio Network with Sensing-enhanced Spectrum Sharing

‍ ‍Over the past few decades， most of the available spectrum has been used for high-speed communication services. This has led to a scarcity of spectrum resources for wireless communication systems. By contrast， a large portion of the available spectrum has a low utilization rate. In order to improve the spectrum utilization of communication systems， the academic community has proposed sharing-based cognitive radio technology. Cognitive radio allows secondary users to share spectrum resources with primary users， which can significantly improve the spectrum utilization efficiency of wireless communication networks. However， the current communication networks based on cognitive radio have some shortcomings such as a low perception performance under a low signal-to-noise ratio and limited performance improvement under severe interference conditions. Recently， an intelligent reflecting surface （IRS） has been widely used in various wireless-communication fields. It can intelligently regulate the wireless transmission environment to meet the high spectrum and energy efficiency requirements. Specifically， an IRS is a planar array consisting of a large number of reconfigurable passive component units， each of which can independently modulate the amplitude and phase of the incident signal （controlled by a central intelligent controller）. This improves the signal transmission environment between the transmitting and receiving ends， assists in signal transmission， and improves the signal-to-noise ratio on the receiving end. In order to further improve the sensing performance and spectral efficiency of cognitive radio， a novel IRS-assisted sensing enhanced spectrum sharing cognitive radio network is proposed. For this network， an optimization problem that involved maximizing the secondary users and rates was established， with the goal of jointly optimizing the beamforming vector of the secondary base stations， intelligent reflection surface phase shift matrix， and perception time. Because of variable coupling， the constructed optimization problem was non-convex. An iterative algorithm based on block coordinate descent was proposed to alternately optimize the beamforming， intelligent reflector phase， and perception time. The non-convex objectives and constraint functions were solved using successive convex approximation， and rank-one conditional constraints were solved using semidefinite relaxation. A one-dimensional search was used to determine the optimal perception time. The simulation results showed that the proposed method could quickly converge and greatly improve the sensing performance and spectral efficiency of a cognitive radio network.