LSTM-Based Traffic Prediction for 5G Mobile Network Operator

The exponential increase in mobile data traffic in 5G networks requires advanced predictive tools for efficient resource management. In this paper, we present a comparative analysis of a predictive methodology based on Long Short Term Memory (LSTM), using input lags for traffic modeling within the mobile core network, using real Mobile Network Operator (MNO) data collected at 15-minute intervals over several days. The performance of the LSTM model is compared with a traditional Recurrent Neural Network (RNN). The results demonstrate that the LSTM achieves higher prediction accuracy and more stable computational performance across different configurations, with inference times ranging from 2.2 to 2.8 seconds.