Repository logo
 
Publication

Poisoning federated learning with graph neural networks in Internet of Drones

2024-05-06Conference object Restricted access
Simple item page
dc.contributor.authorLi, Kai
dc.contributor.authorNOOR, ALAM
dc.contributor.authorNi, Wei
dc.contributor.authorTovar, Eduardo
dc.contributor.authorFu, Xiaoming
dc.contributor.authorAkan, Ozgur B.
dc.date.accessioned2024-05-09T15:51:13Z
dc.date.available2024-05-09T15:51:13Z
dc.date.issued2024-05-06
dc.descriptionThis work was supported by the CISTER Research Unit (UIDP/UIDB/04234/2020) and project ADANET (PTDC/EEICOM/3362/2021), financed by National Funds through FCT/MCTES (Portuguese Foundation for Science and Technology); and supported in part by the AXA Research Fund (AXA Chair for Internet of Everything at Koc¸ University), as well as the EU Horizon Europe project COVER (Grant Agreement ID: 101086228).pt_PT
dc.description.abstractInternet of Drones (IoD) is an innovative technology that integrates mobile computing capabilities with drones, enabling them to process data at or near the location where it is collected. Federated learning can significantly enhance the efficiency and effectiveness of data processing and decision-making in IoD. Since federated learning relies on aggregating updates from multiple drones, a malicious drone can generate poisoning local model updates that involves erroneous information, leading to incorrect decisions or even dangerous situations. In this paper, a new data-independent model poisoning attack is developed to manipulate federated learning accuracy, which does not rely on training data at drones. The proposed attack leverages an adversarial graph neural network (A-GNN) to generate poisoning local model updates based on the benign local models overheard. Particularly, the A-GNN discerns the graph structural correlations between the benign local models and the features of the training data that underpin these models. The graph structural correlations are reconstructively manipulated at the malicious drone to crafts poisoning local model updates, where the training loss of the federated learning is maximized.pt_PT
dc.description.versionN/Apt_PT
dc.identifier.citationLi, K., Noor, A., Ni, W., Tovar, E., Fu, X., Akan, O. (2024, July 29-31). Poisoning federated learning with graph neural networks in Internet of Drones [Paper presentation] 33rd International Conference on Computer Communications and Networks, Big Island, Hawaii, USApt_PT
dc.identifier.urihttp://hdl.handle.net/10400.22/25489
dc.language.isoengpt_PT
dc.peerreviewedyespt_PT
dc.subjectInternet of Drones (IoD)pt_PT
dc.subjectFederated learningpt_PT
dc.subjectAdversarial Graph Neural Networks (A-GNN)pt_PT
dc.subjectPoisoning attackpt_PT
dc.subjectMobile computingpt_PT
dc.titlePoisoning federated learning with graph neural networks in Internet of Dronespt_PT
dc.title.alternative240501pt_PT
dc.typeconference object
dspace.entity.typePublication
oaire.citation.conferencePlaceBig Island, Hawaii, USApt_PT
oaire.citation.titleICCCN 2024 - 33rd International Conference on Computer Communications and Networkspt_PT
person.familyNameLi
person.familyNameNOOR
person.familyNameTovar
person.givenNameKai
person.givenNameALAM
person.givenNameEduardo
person.identifier.ciencia-idEE10-B822-16ED
person.identifier.ciencia-idF919-244E-A2A5
person.identifier.ciencia-id6017-8881-11E8
person.identifier.orcid0000-0002-0517-2392
person.identifier.orcid0000-0002-0077-6509
person.identifier.orcid0000-0001-8979-3876
person.identifier.scopus-author-id7006312557
rcaap.rightsclosedAccesspt_PT
rcaap.typeconferenceObjectpt_PT
relation.isAuthorOfPublication21f3fb85-19c2-4c89-afcd-3acb27cedc5e
relation.isAuthorOfPublicationd9f59cbb-6fee-45c2-ada0-77ef35475525
relation.isAuthorOfPublication80b63d8a-2e6d-484e-af3c-55849d0cb65e
relation.isAuthorOfPublication.latestForDiscovery21f3fb85-19c2-4c89-afcd-3acb27cedc5e

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
CISTER-TR-240501.pdf
Size:
804.79 KB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

ISEP – CISTER – Artigos