Downlink Performance Analysis of CF-MMIMO With RCEs and LO Phase Drift: A Stochastic Geometry Approach

In this article, we study the system-level performance of cell-free massive multiple-input multiple-output (CF-mMIMO) systems with reciprocity calibration errors (RCEs) and imperfect phase synchronization due to the local oscillator (LO) phase drift. Considering the practical implementation of a remote radio unit (RRU) and the nonreciprocity of downlink–uplink channels due to RCEs and LO phase drift, we model the downlink channel with a random phase rotation. Then, we study the system-level performance of both centralized and distributed CF-mMIMO architectures with maximum ratio transmission (MRT) and zero-forcing (ZF) precoding schemes. We derive the closed-form expressions for the signal-to-interference-plus-noise ratio of the downlink precoding schemes, and obtain the closed-form expressions for the average spectral efficiency (SE) per user with locations of both RRU and user equipment (UE) following the Poisson point process. Simulation results show that the closed-form expressions are accurate when compared with Monte Carlo results. The results demonstrate that the impact of RCEs and LO phase drift on system performance is more severe for centralized ZF precoding than for MRT and distributed ZF precoding. In addition, we reveal the SE performance concerning antenna configuration of RRU, RRU density, UE density, and LO phase drift range.