Supporting Data for State-Aware Upload Gating for Personalized Federated Learning in Mobile Edge Environments

This dataset contains the client-partition files, per-seed raw experimental logs, processed summary data, and analysis scripts supporting the results reported in “State-Aware Upload Gating for Personalized Federated Learning in Mobile Edge Environments.” The package covers the CIFAR-10 and CIFAR-100 main experiments, Random-Gate comparisons, threshold experiments, and upload-budget-matched state-factor ablations. The corresponding source code, environment specifications, and verified execution commands are archived separately in SAUG-PFLlib, Version 1.0.0(https://doi.org/10.5281/zenodo.21807089). The files are restricted during peer review. Access is provided to editors and reviewers through a private link. The files will be made publicly available upon publication of the associated article.