Fenics is a comprehensive simulator designed for decentralized federated learning environments. It allows researchers and practitioners to experiment with various network topologies, participant selection strategies, and attack scenarios to evaluate the robustness and efficiency of federated learning algorithms.
This is the historical branch of the Fenics framework.
As a true phoenix, Fenics has now evolved into https://github.com/annalithell/blixtbird.
For the version 1 of our framework, see README.v1.md.
For documentation, please refer to the following publication:
- Saha, S., Nova, S. N., Duvignau, R., & Chiasserini, C. F. (2025, June). Fenics: A Modular Framework for Security Evaluation in Decentralized Federated Learning. In Proceedings of the 19th ACM International Conference on Distributed and Event-based Systems DEBS2025 (pp. 146-151). https://dl.acm.org/doi/pdf/10.1145/3701717.3730550
This project is licensed under the MIT License.
- This project was part of a course project (DAT-300: Data-driven support for Cyberphysical Systems) at Chalmers University of Technology, under the supervision of Romaric Duvignau and Carla Fabiana Chiasserini.
For any questions or feedback, please contact Shubham Saha or Sifat Nawrin Nova.