Network Computing Laboratory (NCLAB) conducts research on a variety of networking and computing technologies, including edge computing, container orchestration, Internet of Things, and edge AI. To realize IoT Edge service infrastructure, we’ve recently focused on federated learning technologies combined with edge computing. We have extensive experience with open-source projects like Kubernetes, KubeEdge, Flower, and SUMO.

Open-Source Release Update: FedEasy & FedProC

[1] FedEasy: We are excited to introduce FedEasy, an intuitive and powerful open-source federated learning framework designed to streamline federated learning experiments. It makes advanced machine learning more accessible to researchers and developers. With its intuitive design, robust foundation, and comprehensive documentation, getting started has never been easier.
Docs: https://fedeasy.readthedocs.io
GitHub: https://github.com/nclabteam/FedEasy

[2] FedProC: FedProC is a benchmarking framework specifically designed for federated time-series forecasting. It provides standardized evaluation pipelines for time-series FL, supports diverse real-world datasets and configurable settings, and enables fair and reproducible benchmarking across methods.
Docs: https://fedproc.readthedocs.io/
GitHub: https://github.com/nclabteam/FedProC