Abstract: Federated learning (FL) has the characteristic of implementing model training without data sharing and operating effective resource management while protecting data privacy. Therefore, it has become one of the research hotspots in the field of mobile communication resource management. In this survey, the algorithms, progress and future trends of FL in mobile communication resource management were summarized and analyzed. First, the basic concept of FL was introduced. Then, the performance of FL resource management methods in distributed wireless network, mobile edge network, Internet of vehicles, fog radio access network, and ultra dense network scenarios were discussed, and their advantages and disadvantages were analyzed. Based on the progress of FL, the open issues of FL were analyzed, and the possible solutions were proposed. Finally, the potential development trends of FL in the field of mobile communication resource management were prospected.