Spatially Correlated Cell-Free Massive MIMO Network with Centralized Operation and Low-Resolution ADCs

This paper investigates the effect of low-resolution analog-to-digital converters (ADCs) on the centralized cell-free massive MIMO network over spatially correlated fading channels. The minimum mean squared error (MMSE) estimator is employed to acquire the channel state information through the established ADC quantization model. Our results show that both imperfect ADCs and pilot contamination will cause error floors. Then, the optimal receive combining is established to maximize the uplink ergodic channel capacity. Furthermore, a closed-form expression for the achievable uplink spectral efficiency (SE) is obtained with the help of maximum ratio (MR) combining, which provides important insights into how finite-resolution ADCs affect the centralized cell-free massive MIMO. Simulation results show that the SE of MR combining rapidly saturates with the increase of ADC quantization bits, while MMSE combining requires more quantization bits to further improve the SE compared with ideal ADCs.