URLLC CF Massive MIMO Systems With Dual Polarized Antennas and Hardware Impairments

Abstract

We consider the downlink of an ultra-reliable low-latency communication (URLLC) cell-free (CF) massive multi-input-multi-output (mMIMO) system, with dual-polarized antennas at access point (AP), and user equipments. We assume that APs are equipped with low-quality radio-frequency (RF) chains and low-resolution analog-to-digital converters/digital-to-analog converters (ADC/DACs), which distort their transmit and receive signals. To mitigate these impairments, we propose a novel distortion-aware minimum-mean-square-error (DA-MMSE) precoder by exploiting the knowledge of RF and ADC/DAC impairment correlation matrices. We apply large dimensional random matrix theory to derive a closed-form spectral efficiency expression. We use it to analytically show that RF and ADC/DAC impairment levels increase as the system becomes more distributed, which necessitates the use of DA-MMSE precoder to achieve a desirable URLLC performance. We also show the effectiveness of DA-MMSE precoder and dual-polarized antennas in meeting the strict URLLC constraints.