Over-the-Air Federated Learning in MIMO Cloud Radio Access Networks

To address the limited server coverage of traditional over-the-air federated learning (OA-FL), we propose a new OA-FL framework for MIMO-based cloud radio access network (Cloud-RAN), called MIMO Cloud-RAN OA-FL (MIMOCROF). The proposed MIMOCROF consists of three stages in each training round. The fi...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE transactions on wireless communications Ročník 24; číslo 7; s. 5825 - 5839
Hlavní autoři: Ma, Haoming, Yuan, Xiaojun, Ding, Zhi
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York IEEE 01.07.2025
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
Broadcasting
cloud radio access network
Computational modeling
Correlation
Federated learning
lossy distributed source coding
MIMO
Model updating
multiple-input multiple-output multiple access channel
Optimization
over-the-air computation
Parameters
Servers
Source coding
Training
Uplink
Vectors
Wireless networks
ISSN:1536-1276, 1558-2248
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:To address the limited server coverage of traditional over-the-air federated learning (OA-FL), we propose a new OA-FL framework for MIMO-based cloud radio access network (Cloud-RAN), called MIMO Cloud-RAN OA-FL (MIMOCROF). The proposed MIMOCROF consists of three stages in each training round. The first stage of edge aggregation allows each access point (AP) to collect local updates from edge devices and construct an edge update using MIMO multiple access. In the second stage of global aggregation, the cloud server (CS) aggregates edge updates received from the APs to form a global update through a fronthaul network. In the third stage of model updating and broadcasting, the CS sends the updated global model parameters to the APs, and the latter then broadcast the parameters to their served devices. To effectively exploit inter-AP correlation, we model the global aggregation stage as a lossy distributed source coding (L-DSC) problem. Based on the rate-distortion theory, we further analyze the performance of the MIMOCROF framework. We formulate a communication-learning optimization problem to improve the system performance by considering the inter-AP correlation. To solve this problem, we develop an algorithm by using alternating optimization (AO) and majorization-minimization (MM). Furthermore, we propose a practical L-DSC that exploits inter-AP correlation. Numerical results show that the proposed practical L-DSC effectively utilizes inter-AP correlation and is superior to other baseline schemes in performance.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1536-1276
1558-2248
DOI:10.1109/TWC.2025.3549499