Passive Beamforming for 3-D Coverage in IRS-Assisted Communications

In this letter, we study the passive beamforming design for optimizing the three-dimensional (3-D) coverage performance in an intelligent reflecting surface (IRS)-assisted communication system. In particular, we maximize the minimum (i.e., max-min) beamforming gain over all locations in the target area by jointly optimizing the passive reflection of all IRS elements. The resulting problem has a non-convex and semi-infinite beamforming gain constraint and a non-convex constant-modulus constraint. Both of them are reformulated to be finite and convex in the lifted matrix space with an additional rank-one constraint, and a difference-of-convex technique is developed to solve the reformulated problem efficiently. Finally, we demonstrate that the proposed passive beamforming solution outperforms the existing methods and can approach the upper bound on the max-min beamforming gain.