AP selection in Cell-Free Massive MIMO system using Machine Learning Algorithm

Current cellular networks work by creating autonomous cells that fail to serve many users due to uneven coverage area. With the world becoming more reliant on the wireless communication system, a cellular network with multiple wireless access points, high spectral and energy efficiency is a must. These requirements are fulfilled by a cell-free Massive multiple-input multiple-output (MIMO) network. Apart from being ideal for the current wireless communication's requirements, it can also resolve a lot of currently faced interference issues. With this work, we aim to reduce the complexity so this system becomes more practical to deploy in the real world and reduce one of the limitations of a cell free massive MIMO system. We propose a new access point selection algorithm called CAPS (Cluster based AP selection) for the cell-free Massive MIMO which is aimed at reducing the computations workload and pilot contamination by introducing machine learning algorithm for clustering which in our work is the K-means++ clustering algorithm. The simulations and calculations show that using the proposed algorithm have performed better compared to prevailing approaches.