Abstract: Federated Learning (FL) has emerged as a promising paradigm for decentralized machine learning, where a central server coordinates distributed clients to collaboratively train a global model ...