Advancing Asynchronous Federated Learning for High-Latency Environments

This research explores a new asynchronous federated learning framework, addressing limitations in high-latency scenarios. By simulating Earth-Mars collaboration, we validate performance against delays, data heterogeneity, and resource constraints, paving the way for innovative solutions in federated learning applications.