Lightweight AP selection algorithm for uniform quality everywhere in Cell-Free massive MIMO

We propose an AP selection algorithm that is computationally lightweight and feasible for uniform radio quality in wide-area deployment for cell-free massive MIMO (CF-mMIMO). To address the scalability problem in a large-scale CF-mMIMO, recent studies involved the use of distributed CPU architecture and an optimized method of AP selection for each user based on the received signal power, channel state, and user mobility. An AP selection algorithm that achieves uniform quality everywhere with lightweight computation is needed to adapt user mobility and maintain high radio quality. In addition, the algorithm needs to overcome the radio quality degradation of area boundary UEs due to inter-site interference in distributed CPUs. As an algorithm for AP selection, optimization-based methods have been proposed to maximize the throughput sum rate or fairness index. However, these methods require inverse matrix calculations per UE for the objective function, resulting in computational complexity becoming an issue. To address this problem, we propose a list-processing AP selection algorithm that selects APs that contribute to quality improvement to uniform signal power and interference effects among users without inverse matrices calculation for estimating the radio quality and a metaheuristic solution search. Simulation results show that the proposed method can provide uniform and high radio quality in urban environments with a large number of UEs.