Accelerating Federated Learning via Momentum Gradient Descent | IEEE Journals & Magazine | IEEE Xplore