User Mobility-Based Dynamic Clustering Algorithm for Cell-Free Massive MIMO Systems

The Cell-Free massive MIMO system (CF-mMIMO) serves all users in a scenario simultaneously through a large number of distributed access points (APs), and studies have shown that each user achieves a nearly 5 -fold improvement in throughput compared to small cell scenarios with the CF-mMIMO system. Although the CF-mMIMO system shows advantages under static network conditions, the effects of increased delay and reduced throughput caused by switching APs for mobile users with different speeds have not been addressed. In this paper, we analyze the user-centric AP dynamic cluster selection algorithm UC-kmeans++ under the scalable CF-mMIMO system for different user mobile speeds. The results show that the scalable CF-mMIMO system can reduce the CPU cooperative cluster switching rate, decrease the number of inter-AP switching times by about $40 \%$ and improve the throughput by $12 \%$ under high mobility conditions, which can provide a more effective user mobility management method than the existing dynamic clustering to provide more stable and efficient communication services.