­

Mobile traffic modeling

  • Traffic prediction
    • RNN and CNN based models
    • transfer learning
  • Traffic classification
    • service and app identification
    • based on active measurements
  • Joint traffic prediction and classification
    • using multi-task learning
  • Traffic anomaly detection
    • semi-supervised learning
    • identification of urban anomalies
  • Usage of federated learning to build a general model from multiple cell sites (ITU AI4Good 5G Data Challenge)
­