Statistical channel estimation for large-scale fading processing in cell-free massive MIMO

Cell-free massive MIMO (mMIMO) emerges as a crucial physical layer technology, offering wireless transmission with high spectral efficiency. Large scale fading (LSF) processing is suggested as a promising technique to address the performance gap between distributed and centralized processing solutions in cell-free mMIMO. However, both LSF detection (LSFD) for uplink and LSF precoding (LSFP) in the downlink require the access points (APs) to transmit their locally estimated statistical channel state information (CSI) to the central processing unit (CPU), resulting in additional CSI overhead in the fronthaul links. In this paper, we propose a novel method to estimate the necessary statistical CSI for LSFD and LSFP at the CPU based on the uplink data signal. Our analytical and simulation results demonstrate that the proposed method can significantly reduce the CSI overhead in the fronthaul links while achieving similar or even higher spectral efficiency than existing solutions based on statistical CSI forwarding via fronthaul.