IRS-Assisted Interference Mitigation for Full-Duplex IAB Network

This letter studies intelligent reflecting surface (IRS)-aided full-duplex (FD) integrated access and backhaul (IAB) network, where IRSs are deployed near small-cell base stations (SBSs) to mitigate several types of interference including self-interference in FD system. An optimization problem is formulated to maximize network throughput by jointly optimizing precoder at base stations (BSs) for establishing wireless backhaul and access links, and passive beamforming at IRSs. To solve this non-convex problem, we divide the problem into two subproblems by using the relationship between data rate and weighted rate-minimum mean square error (WMMSE). Then, we propose the block coordinate descent (BCD) algorithm, where the BS precoder is optimized using quadratically constrained quadratic programming (QCQP) and the IRS passive beamforming is obtained via semi-definite relaxation (SDR) method. Simulation results illustrate that our proposed algorithm can effectively mitigate interference with the help of IRSs and achieve a better sum rate performance compared to conventional schemes.