The revolution of technologies such as Internet-of-Things, cloud computing, and 5G, has facilitated the increased adoption of various mobile devices, such as smart phones, sensors, and Internet-of-Vehicles, etc. Moreover, the heterogeneous mobile sources generate a considerable and ever increasing scale of data with abundant information, commonly referred as mobile big data, making it possible to gain business insights and better decision making from the large volume of data information.

In practice, the substantial volume of collected mobile data is difficult to be efficiently analyzed in mobile devices due to their limited resources. Typically, a centralized server can be adopted to collect the large volume of data generated by heterogeneous mobile devices; and the mobile data will be further processed and analyzed in a centralized manner. However, such centralized solutions suffer from limitations like large network communication cost, data privacy leakage, to name a few. Federated learning is an emerging machine learning paradigm that empowers IoT applications in a manner that preserves data heterogeneity and privacy, by training the model without transferring data from local mobile devices to the central server. In each round of federated learning, multiple devices are selected to train models locally to produce a global model under the coordination of a central server. The data of each mobile device is generated and stored locally, without being transferred to the central server or other devices. Instead, only model updates are sent from mobile devices to the central server for global model formulation.

Recently, federated learning for IoT has attracted great interest from academia and industry. However, the topic is quite new and has not been investigated under its different profiles until now. There is a lack of literature from both a theoretical and an empirical point of view. In this Special Issue, we are looking for cutting edge technologies and novel studies and reviews, which can realize and evaluate the effectiveness and advantages of federated learning for advancing IoT.

Potential topics include but are not limited to the following:

• Privacy-preserving mechanisms in federated learning driven IoT applications
• Aggregation algorithms for federated learning driven IoT applications
• Security mechanisms for federated learning driven IoT applications
• Evaluation framework for federated learning driven IoT applications
• Resource management for federated learning driven IoT applications
• Communication mechanisms for federated learning driven IoT applications
• Data provenance and auditing for federated learning driven IoT applications
• Incentive mechanisms for federated learning driven IoT applications