Hierarchical over-the-air federated edge learning

buir.contributor.authorAygün, Ozan
buir.contributor.authorKazemi, Mohammad
buir.contributor.authorDuman, Tolga M.
buir.contributor.orcidKazemi, Mohammad|0000-0001-5177-1874
buir.contributor.orcidDuman, Tolga M.|0000-0002-5187-8660
dc.citation.epage3381en_US
dc.citation.spage3376en_US
dc.contributor.authorAygün, Ozan
dc.contributor.authorKazemi, Mohammad
dc.contributor.authorGündüz, D.
dc.contributor.authorDuman, Tolga M.
dc.coverage.spatialSeoul, Republic of Koreaen_US
dc.date.accessioned2023-02-24T06:29:31Z
dc.date.available2023-02-24T06:29:31Z
dc.date.issued2022
dc.departmentDepartment of Electrical and Electronics Engineeringen_US
dc.descriptionConference Name: ICC 2022 - IEEE International Conference on Communicationsen_US
dc.descriptionDate of Conference: 16-20 May 2022en_US
dc.description.abstractFederated learning (FL) over wireless communication channels, specifically, over-the-air (OTA) model aggregation framework is considered. In OTA wireless setups, the adverse channel effects can be alleviated by increasing the number of receive antennas at the parameter server (PS), which performs model aggregation. However, the performance of OTA FL is severely limited by the presence of mobile users (MUs) located far away from the PS. In this paper, to mitigate this limitation, we propose hierarchical over-the-air federated learning (HOTAFL), which utilizes intermediary servers (IS) to form clusters near MUs. We provide a convergence analysis for the proposed setup, and demonstrate through experimental results that local aggregation in each cluster before global aggregation leads to a better performance and faster convergence than OTA FL.en_US
dc.identifier.doi10.1109/ICC45855.2022.9839230en_US
dc.identifier.eisbn978-1-5386-8347-7
dc.identifier.eissn1938-1883
dc.identifier.isbn978-1-5386-8348-4
dc.identifier.issn1550-3607
dc.identifier.urihttp://hdl.handle.net/11693/111670
dc.language.isoEnglishen_US
dc.publisherIEEEen_US
dc.relation.isversionofhttps://dx.doi.org/10.1109/ICC45855.2022.9839230en_US
dc.source.titleIEEE International Conference on Communications (ICC)en_US
dc.titleHierarchical over-the-air federated edge learningen_US
dc.typeConference Paperen_US

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Hierarchical_over-the-air_federated_edge_learning.pdf
Size:
2.1 MB
Format:
Adobe Portable Document Format
Description:

License bundle

Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.69 KB
Format:
Item-specific license agreed upon to submission
Description: