Cell-Free Massive MIMO for URLLC With Imperfect Channel Reciprocity.

This paper considers the cell-free massive multiple-input multiple-output (MIMO) system for ultra reliable and low-latency communication (URLLC) under non-reciprocal channels and pilot contamination. To investigate the impact of the channel non-reciprocity, we model it as a random variable using the more realistic truncated Gaussian distribution. In this spirit, we derive the closed-form expression of the downlink achievable rate, which depends only on the statistics of channel non-reciprocity (NRC) and channel state information. To improve the system performance and fairness among users, we formulate the max-min power control problem to maximize the minimum user downlink achievable rate. We propose a path-following method to reformulate the non-convex original problem as a series of geometric programmings. Simulation results show new findings that both amplitude and phase reciprocity errors at access points can significantly degrade the achievable rate when only the channel NRC variable statistics are available. Moreover, the proposed power control algorithm converges rapidly and greatly improves the system performance.