Traffic Prediction in Satellite Optical Networks Based on Asynchronous Federated Learning

An asynchronous federated learning method predicts satellite optical network traffic, improving convergence and reducing communication overhead while preserving privacy.

Xueqiang Zhao, Zihao Qin, Shaobo Qin, Haicheng Li, Xiaoke Sun, Pengyu Wang, Longchao Wei, Yajun Li, Ruijie Zhu

Published Nov 2, 2024Paper ↗

68%

OverallHighly rated

OverallHighly ratedVote to see the score

Readers

–

Only vote on papers you've read. Sign in to vote.

AI panel4/20reviewers recommend it

lenient 4/5

medium 0/10

strict 0/5

AI panel?Vote to see what the 20 AI reviewers said

Abstract

This paper presents a traffic prediction method for satellite optical networks using asynchronous federated learning, enhancing model convergence, reducing communication burden, and addressing traffic distribution and overload challenges while ensuring privacy.