Power Allocation in 1-Bit Massive MIMO Downlink with Zero-Forcing Precoding

Abstract

The use of 1-bit digital-to-analog converters (DACs) is a promising way to improve the energy efficiency of massive multiple-input-multiple-output (MIMO) systems. Many linear and nonlinear precoding techniques mitigating the severe quantization distortions have been introduced in order to enable the use of 1-bit DACs in the massive MIMO downlink. However, the vast majority of the proposed designs do not account for the channel strength imbalances between the users, which are highly likely to occur in the downlink scenarios. To achieve high sum rates in such scenarios, a user-specific power allocation scheme (USPA) is needed. In this paper, we propose a quantization aware USPA technique for the linear zero-forcing (ZF) precoder. Numerical results illustrate that higher achievable sum rates than the existing linear methods can be achieved with the proposed method, especially in the systems with low order modulated inputs. In the systems with higher modulation, the proposed method achieves similar performance to the existing methods, yet with a lower computational complexity.