Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Liu, Ya, Yang, Kai, Zhu, Yu, Yang, Keying, Zhao, Haibo
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2505.09106
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916736894238720
author Liu, Ya
Yang, Kai
Zhu, Yu
Yang, Keying
Zhao, Haibo
author_facet Liu, Ya
Yang, Kai
Zhu, Yu
Yang, Keying
Zhao, Haibo
contents The space-air-ground integrated network (SAGIN) has recently emerged as a core element in the 6G networks. However, traditional centralized and synchronous optimization algorithms are unsuitable for SAGIN due to infrastructureless and time-varying environments. This paper aims to develop a novel Asynchronous algorithm a.k.a. Argus for tackling non-convex and non-smooth decentralized federated bilevel learning over SAGIN. The proposed algorithm allows networked agents (e.g. autonomous aerial vehicles) to tackle bilevel learning problems in time-varying networks asynchronously, thereby averting stragglers from impeding the overall training speed. We provide a theoretical analysis of the iteration complexity, communication complexity, and computational complexity of Argus. Its effectiveness is further demonstrated through numerical experiments.
format Preprint
id arxiv_https___arxiv_org_abs_2505_09106
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Argus: Federated Non-convex Bilevel Learning over 6G Space-Air-Ground Integrated Network
Liu, Ya
Yang, Kai
Zhu, Yu
Yang, Keying
Zhao, Haibo
Machine Learning
68T07
I.2
The space-air-ground integrated network (SAGIN) has recently emerged as a core element in the 6G networks. However, traditional centralized and synchronous optimization algorithms are unsuitable for SAGIN due to infrastructureless and time-varying environments. This paper aims to develop a novel Asynchronous algorithm a.k.a. Argus for tackling non-convex and non-smooth decentralized federated bilevel learning over SAGIN. The proposed algorithm allows networked agents (e.g. autonomous aerial vehicles) to tackle bilevel learning problems in time-varying networks asynchronously, thereby averting stragglers from impeding the overall training speed. We provide a theoretical analysis of the iteration complexity, communication complexity, and computational complexity of Argus. Its effectiveness is further demonstrated through numerical experiments.
title Argus: Federated Non-convex Bilevel Learning over 6G Space-Air-Ground Integrated Network
topic Machine Learning
68T07
I.2
url https://arxiv.org/abs/2505.09106