Capacity Maximization Based Multi-Resource Assignment for IRS-Assisted Cognitive SWIPT with Nonlinear Energy Harvesting

In this paper, capacity maximization based multi-resource assignment strategy is proposed for intelligent reflecting surface (IRS)-assisted cognitive simultaneous wireless information and power transfer (SWIPT) with nonlinear energy harvesting in overlay spectrum sharing scenario. The maximum cognitive SWIPT sum capacity is set as the optimization target, which is subject to secondary transmitter (ST) power control, cognitive SWIPT user nonlinear energy harvesting and the modulus of IRS reflection coefficient element constraints. The multi-resource assignment strategy is investigated via the resolution of this nonlinear non-convex optimization problem with multi-variable coupling. Alternating optimization (AO) is implemented to obtain optimal subcarrier power and optimum IRS phase shift vector. Firstly, IRS phase shift vector is fixed, then the Lagrange duality is applied to search optimal subcarrier and power. Once the optimal subcarrier and power are obtained, IRS phase shift vector is derived via successive convex approximation (SCA) approach. Simulation results indicated that, the proposed scheme effectively allocates multi-resource such as subcarrier power and IRS reflective beamforming vector. The maximum sum capacity can be achieved in IRS-assisted cognitive SWIPT with practical nonlinear energy harvesting case.