Incentive-Based Delay Minimization for 6G-Enabled Wireless Federated Learning