IRS-Assisted Spectrum Sensing and Primary-Secondary Transmission for Cognitive Radio Networks
Xiaopeng Liang, Liangji Huang, Qian Deng, Feng Shu, Guangcheng Yu, Jiangzhou Wang · IEEE Transactions on Cognitive Communications and Networking · 2024
A novel intelligent reflecting surface (IRS)-assisted sensing and communication model is proposed to simultaneously improve the performance of both secondary network (SN) and primary network (PN) in the opportunistic spectrum access cognitive radio networks (OSA-CRNs), where IRS assists not only the sensing but also the transmission of both SN and PN according to the spectrum sensing results. Our goal is to maximize the SN’s sum rate by jointly optimizing secondary transmitter’s (ST’s) beamforming, sensing duration and three-stage IRS phase shifts matrices, while fully satisfying PN’s average achievable rate requirement. Considering that the formulated problem is non-convex, which can be decomposed into five sub-problems. For the IRS-assisted sensing-stage and primary transmission-stage phase shifts matrices sub-problems, the closed-form solutions are deduced. For the IRS-assisted secondary transmission-stage phase shift matrix and the ST’s beamforming sub-problems, the approximately optimal solutions can be obtained by applying the low complexity penalty dual decomposition based gradient projection (PDDGP) algorithm. For the sensing duration sub-problem, the optimal solution is derived via a golden section search method. Finally, the original problem is efficiently solved via an alternate iterative framework. Simulation results demonstrate that the spectral efficiencies of both primary transmission and secondary transmission are significantly enhanced in the proposed IRS-assisted OSA-CRNs.