Mobility-Aware Joint User Scheduling and Resource Allocation for Low Latency Federated Learning