A Proactive Defense Against Model Poisoning Attacks in Federated Learning | IEEE Journals & Magazine | IEEE Xplore