Over the past few years, federated learning (FL) has become a popular
distributed machine learning paradigm. FL involves a group of clients with
decentralized data who collaborate to learn a common model under the
coordination of a centralized server, with the goal of protecting client