A Reliable and Energy-Efficient Federated Learning in Computing Force Network

Shizhan Lan, Shizhan Lan, Zhenyu Wang, Yuxuan Long, Weichao Kong

2023

Abstract

A Computing Force Network (CFN) is a distributed computing network that utilizes distributed computing power to solve complex computational problems. Unlike traditional computing networks that centralize resources in a single location, CFNs distribute processing power across an interconnected network of devices. The intelligent devices within CFNs possess computational capabilities and carry varying quantities of private training samples. Due to CFNs being provided by different service providers, private data between devices cannot be directly shared, which makes it challenging to train models directly using the private data. Federated Learning (FL) emerges as a novel distributed training paradigm that enables distributed model training while preserving user privacy. This paper presents a K-Means-based communication-assured federated learning algorithm for CFNs. It allows for federated training tasks under non-iid conditions and selects reliable clients for communication in each round to ensure algorithm convergence. Experimental results demonstrate the superior performance of our algorithm compared to the-state-of-arts.

Paper Citation

in Harvard Style

Lan S., Wang Z., Long Y. and Kong W. (2023). A Reliable and Energy-Efficient Federated Learning in Computing Force Network. In Proceedings of the 2nd International Seminar on Artificial Intelligence, Networking and Information Technology - Volume 1: ANIT; ISBN 978-989-758-677-4, SciTePress, pages 117-124. DOI: 10.5220/0012275700003807

in Bibtex Style

@conference{anit23,
author={Shizhan Lan and Zhenyu Wang and Yuxuan Long and Weichao Kong},
title={A Reliable and Energy-Efficient Federated Learning in Computing Force Network},
booktitle={Proceedings of the 2nd International Seminar on Artificial Intelligence, Networking and Information Technology - Volume 1: ANIT},
year={2023},
pages={117-124},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012275700003807},
isbn={978-989-758-677-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Seminar on Artificial Intelligence, Networking and Information Technology - Volume 1: ANIT
TI - A Reliable and Energy-Efficient Federated Learning in Computing Force Network
SN - 978-989-758-677-4
AU - Lan S.
AU - Wang Z.
AU - Long Y.
AU - Kong W.
PY - 2023
SP - 117
EP - 124
DO - 10.5220/0012275700003807
PB - SciTePress