LAUSR

In this letter, we investigate the passive beamforming for intelligent reflecting surface (IRS)-assisted multi-user multiple-input multiple-output systems with one-bit analog-to-digital converters aiming at minimizing the symbol error rate (SER). By leveraging the concept of signal space, we formulate an optimization problem of maximizing the minimum Euclidean distance between different signal vectors in such a space. Two algorithms are proposed to tackle the highly non-convex problem, where one is to improve the distribution of the distances so that the minimum one is increased, and the other is to derive a lower bound of the minimum distance and maximize the bound using alternating optimization and manifold optimization. Simulation results reveal that with the proposed two algorithms, the optimized IRS can significantly improve the SER performance over the conventional designs.