Energy-Efficient Beam Shaping in MIMO System Using Machine Learning
摘要
The work presents an algorithm for energy-efficient beamforming in a MU-MIMO (multiuser multiple input multiple output) system using machine learning methods. The task of clustering subscribers by service level and distance was formulated. The k-means algorithm was chosen as a clustering method for solving the beamforming problem. It is shown that the suggested algorithm may increase the 5G network energy efficiency up to 38%.