Quantifying the Operational Benefits of Deep Learning Based Dynamic Traffic Prediction Using Real-World Dataset (W2A.46)
Presenter: Dimitris Uzunidis, University of Patras
A convolutional neural network is trained using real-world data, for dynamic prediction of the required transceivers supporting 6G Χ-haul, leading to 20% and 16% lower average transceiver utilization over static and semi-static cases, respectively
Authors:Dimitris Uzunidis, University of Patras / Christos Christofidis, University of Patras / Ivan De Francesca, Telefónica / Jose Manuel Rivas Moscoso, Telefónica / David Larrabeiti, Universidad Carlos III de Madrid / Josep Maria Fàbrega, Centre Tecnologic de Telecomunicacions de Catalunya / Dan Marom, Hebrew University of Jerusalem / Ioannis Tomkos, University of Patras