KiFramework — A Federated Learning Platform with an Innovative Communication Protocol
Abstract
This paper presents a newly developed framework for federated learning adapted to crisis management applications on the Android platform. This framework focuses on solving client communication problems during the federated learning process. KiFramework introduces an innovative communication protocol and framework that enhances data exchange reliability and supports efficient collaboration in multi-server environments, addressing key challenges in traditional centralized server architectures. Our research provides a comprehensive solution to optimize communication and ensure seamless collaboration among multiple servers, improving applications’ efficiency using federated learning.
Keywords
Android, communication protocol, federated learning, framework, Machine Learning
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
J. Michalek, V. Oujezsky and V. Skorpil, "KiFramework — A Federated Learning Platform with an Innovative Communication Protocol," in Journal of Communications Software and Systems, vol. 20, no. 4, pp. 329-339, December 2024, doi: https://doi.org/10.24138/jcomss-2024-0091
@article{michalek2024kiframeworkfederated,
author = {Jakub Michalek and Vaclav Oujezsky and Vladislav Skorpil},
title = {KiFramework — A Federated Learning Platform with an Innovative Communication Protocol},
journal = {Journal of Communications Software and Systems},
month = {12},
year = {2024},
volume = {20},
number = {4},
pages = {329--339},
doi = {https://doi.org/10.24138/jcomss-2024-0091},
url = {https://doi.org/https://doi.org/10.24138/jcomss-2024-0091}
}