Lattice-Reduction-Aided One-Bit Precoding for Massive MU-MIMO Systems

Abstract

One-bit precoding is a recently proposed scheme that facilitates the use of one-bit digital-to-analog converters to reduce the cost and power consumption in a massive antenna system. In line with this, we consider the massive multiuser-multiple-input multiple-output scenario and propose a lattice-reduction-aided one-bit precoding algorithm. The proposed algorithm solves the mean square error-based one-bit precoder design problem by augmenting the Babai's nearest plane procedure with a newly proposed list generation algorithm. The proposed precoder contains no tuning parameters and hence avoids additional overheads for parameter tuning when applied to a practical system. Furthermore, the symbol precoding stage in our precoder is noniterative and can be implemented via distributed processing. These desirable features along with an excellent performance-complexity tradeoff demonstrated that the proposed precoder stands out as a competitive alternative in addition to the existing one-bit precoders.