Interference suppression for distributed CPU deployments in Cell-Free massive MIMO

We propose an interference suppression method that reduces the amount of data transmitted between sites and the computational load of radio signal processing as well as maintaining radio quality in wide-area deployment for cell-free massive MIMO (CF-mMIMO). In a large-scale CF-mMIMO, the central processing units (CPUs) that process CF-mMIMO signals are assumed to be deployed at multiple sites. There is the problem of inter-site interference between UEs connected to CPUs at different sites. Conventional methods which deal with this problem require the exchange of radio signals between CPUs, which significantly increases the transmission load between the sites. In this paper, we consider the application of coordinated beamforming (CoBF) techniques to CF-mMIMO to reduce the inter-site transmission load. For the CoBF across the sites, the channel estimation processing function for inter-site interference is operated at each site by exchanging the pilot assignment information instead of relying on radio signals. Inter-site interference is suppressed by signal processing using the estimated channel. However, since this approach requires extensive channel estimation across sites, the computational load is high. Thus, we propose a method to reduce the computational load by adjusting the range of channel estimation for each user according to the required radio quality and interference power. Simulation results show that the proposed method can significantly reduce the transmission and computational load for the same area throughput compared to existing methods. Furthermore, we show the evaluation results of experiments using the CF-mMIMO testbed and the feasibility of the proposed method in suppressing inter-site interference.