Federated Learning for Resource-Constrained IoT Devices: Panoramas and State of the Art

Nowadays, devices are equipped with advanced sensors with higher processing and computing capabilities. Besides, widespread Internet availability enables communication among sensing devices that results the generation of vast amounts of data on edge devices to drive Internet-of-Things (IoT), crowdsourcing, and other emerging technologies. The extensive amount of collected data can be preprocessed, scaled, classified, and finally, used for predicting future events with machine learning (ML) methods. In traditional ML approaches, data is sent to and processed in a central server, which encounters communication overhead, processing delay, privacy leakage, and security issues. To overcome these challenges, each client can be trained locally based on its available data and by learning from the global model. This decentralized learning approach is referred to as federated learning (FL). However, in large-scale networks, there may be clients with varying computational resource capabilities. This may lead to implementation and scalability challenges for FL techniques. In this paper, we first introduce some recently implemented real-life applications of FL underlying the applications that are suitable for FL-based resource-constrained IoT environments. We then emphasize the core challenges of implementing the FL algorithms from the perspective of resource limitations (e.g., memory, bandwidth, and energy budget) of client devices. We finally discuss open issues associated with FL for resource-constrained environments and highlight future directions in the FL domain concerning resource-constrained devices.