Precipitated Handover Decision Detection in LTE Networks Through Recurrent Neural Networks

The handover decision algorithms based on the Reference Signal Received Power (RSRP) are sensitive to measurement errors and oscillation of the user's geographic position, causing signaling overload and interruption of services provided by the mobile network. However, the use of intelligent algorithms based on Recurrent Neural Networks (RNN) offer a solution to mitigate handover failures. This work validates the implementation of two RNN models, Long Short-Term Memory (LSTM) and Gate Recurrent Unit (GRU), for detecting precipitated handover decisions in the Long-Term Evolution (LTE) network. The implemented neural Is estimate the probability of unnecessary handover decisions, using as input the LTE Key Performance Indicators (KPI) and mobile device location. The results of the experiments show that the classification method did not obtain optimal results for detection of precipitated handover decision, however, according to the literature, they are adequate to what is expected from a classification scenario using RNN networks in a small and unbalanced database.