Privacy-Preserving Wireless Federated Learning Exploiting Inherent Hardware Impairments
Paper i proceeding, 2021

We consider a wireless federated learning system where multiple data holder edge devices collaborate to train a global model via sharing their parameter updates with an honest-but-curious parameter server. We demonstrate that the inherent hardware-induced distortion perturbing the model updates of the edge devices can be exploited as a privacy-preserving mechanism. In particular, we model the distortion as power-dependent additive Gaussian noise and present a power allocation strategy that provides privacy guarantees within the framework of differential privacy. We conduct numerical experiments to evaluate the performance of the proposed power allocation scheme under different levels of hardware impairments.

Författare

Sina Rezaei Aghdam

Chalmers, Elektroteknik, Kommunikation, Antenner och Optiska Nätverk

Ehsan Amid

Google Inc.

Marija Furdek Prekratic

Chalmers, Elektroteknik, Kommunikation, Antenner och Optiska Nätverk

Alexandre Graell I Amat

Chalmers, Elektroteknik, Kommunikation, Antenner och Optiska Nätverk

2021 IEEE 26th International Workshop on Computer Aided Modeling and Design of Communication Links and Networks (CAMAD)

2378-4865 (ISSN) 2378-4873 (eISSN)

9617764
978-1-6654-1779-2 (ISBN)

2021 IEEE 26th International Workshop on Computer Aided Modeling and Design of Communication Links and Networks (CAMAD)
Porto, Portugal,

Förbättrad säkerhet och integritet för trådlöst federerat lärande

Chalmers AI-forskningscentrum (CHAIR) (2020-013 CHAIR CO), 2020-08-01 -- 2021-01-31.

Ämneskategorier

Datorteknik

Telekommunikation

Annan elektroteknik och elektronik

Styrkeområden

Informations- och kommunikationsteknik

DOI

10.1109/CAMAD52502.2021.9617764

ISBN

9781665417792

Mer information

Senast uppdaterat

2022-01-31