Mobile Terminal Session SIR Prediction Method Based on Clustering and Classification Algorithms

In this paper, a novel method for mobile terminal session signal to interference ratio (SIR) prediction is presented. The method is based on clustering and classification algorithms configuring a fully-automatic process, highly adaptable to the deployment scenario. It is focused on the prediction of the session SIR probability density function (pdf) of each user whose knowledge will allow for the application of advanced radio resource management (RRM) techniques. The proposed method has been applied in simulated and real scenarios showing its validity in both cases.