Robust Beamforming Design for RIS-Assisted Cognitive Radio Systems With Hardware Impairments

In this paper, we consider an reconfigurable intelligent surface (RIS)-assisted downlink cognitive radio system in the presence of hardware impairments (HWIs) and propose an RIS-assisted robust beamforming design (RIS-Robust) scheme. We aim to maximize the average achievable sum rate (AASR) for cognitive users while ensuring the transmit power budget of the cognitive base station, the quality-of-service requirement of the primary user, and the unit-modulus constraint on the reflection matrix of the RIS. Considering that our formulated AASR maximization (AASRM) problem is non-convex, we first utilize a fractional programming technique to reformulate the original objective function, and then use a block coordinate descent (BCD) method to convert the reformulated problem into several subproblems. For each subproblem, we customize a low-complexity iterative algorithm with each step in closed form satisfying the Karush-Kuhn-Tucker (KKT) conditions. Numerical results illustrate that the proposed RIS-Robust scheme performs better than the conventional benchmarks in terms of the AASR.