VREM-FL datasets: a collection of datasets for vehicular federated learning

This dataset collection includes three files that were used for the paper

L. Ballotta, N. D. Fabbro, G. Perin, L. Schenato, M. Rossi and G. Piro, "VREM-FL: mobility-aware computation-scheduling co-design for vehicular federated learning," in IEEE Transactions on Vehicular Technology, doi: 10.1109/TVT.2024.3479780.

Specifically, each file contains 6 columns: {timestep, vehicle ID, x coordinate in the map, y coordinate in the map, real bitrate, estimated bitrate}. The datasets, obtained from REMs with Gaussian estimation and real (https://ieee-dataport.org/open-access/crawdad-romataxi) or simulated (https://eclipse.dev/sumo/) vehicular mobility, are used in the original paper for optimizing the task of federated learning (client scheduling and resource allocation).