IRS‐assisted cooperative power transfer and spectrum sharing for cognitive radio networks

Summary We propose an intelligent reflecting surface (IRS)‐assisted three‐phase protocol based on cooperative wireless power transfer for a primary system to enable spectrum sharing with a secondary system in a cognitive radio network. The primary system comprises an access point (AP) and an energy‐harvesting primary user (PU), whereas the secondary system consists of a secondary transmitter (ST) and a secondary receiver (SR). Within each time block, the protocol operates in three phases. In Phase I, AP and ST transmit energy to PU. In Phase II, PU uses the harvested energy to transmit data to AP. In Phase III, ST accesses the spectrum and transmits data to SR. Each phase is assisted by an IRS. We formulate an optimization problem that maximizes the rate of the secondary system under a primary‐rate constraint by optimizing the IRS phase shifts and time fractions. To tackle the nonconvex problem, we propose two joint solutions: one based on semidefinite relaxation and the other on deep reinforcement learning. Numerical results show that the proposed protocol outperforms benchmarks in terms of the secondary system's rate while guaranteeing the performance of the primary system.