Alternating Minimization Algorithms for One-Bit Precoding in Massive Multiuser MIMO Systems

The use of low-cost and power-efficient hardware components has elicited considerable interest in the context of massive multiuser multiple-input and multiple-output (MIMO) systems because such hardware architecture enables significant power and cost savings at the base station (BS). In this study, we focus on the downlink precoding problem in massive multiuser MIMO systems, where each BS antenna element is equipped with 1-bit digital-to-analog converters (DACs). We aim to find the optimal precoding vector and the associated precoding factor, such that multiuser interference can be minimized for all users. However, such precoding design problem with 1-bit DACs at the BS is non-convex and challenging because of the coupled precoding vector and associated precoding factor. To deal with this problem, we propose a two-step iterative procedure that applies the alternating minimization (AltMin) framework to alternately solve for the precoding vector and the associated precoding factor while keeping the other constant. The primary difficulty is the precoding vector design, which we address by proposing a low-complexity algorithm developed from the coordinate descent framework to find the optimal precoding vector. Simulation results show that the proposed AltMin algorithm not only exhibits low complexity but can also achieve a bit error rate performance that is comparable with or higher than those of state-of-the-art algorithms under various test scenarios.