MACR: Machine-Assisted Coverage Estimation for Radio Planning in Mobile Networks Using a Pre-Trained Approach

Radio propagation modeling is an essential process in network planning and optimization, particularly in urban environments where obstacles highly affect signal propagation. Recently, deep learning advancements have enabled precise predictions. This paper introduces a novel transfer learning-based approach with a pre-trained FCN-ResNet50 model for radio coverage estimation. Our method demonstrates enhanced accuracy and computational efficiency. To the best of our knowledge, this is the first application of transfer learning in coverage estimation that combines efficiency, scalability, and high accuracy.