Leveraging Channel Noise for Sampling and Privacy via Quantized Federated Langevin Monte Carlo

Yunchuan Zhang, Dongzhu Liu, Osvaldo Simeone

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

For engineering applications of artificial intelligence, Bayesian learning holds significant advantages over standard frequentist learning, including the capacity to quantify uncertainty. Langevin Monte Carlo (LMC) is an efficient gradient-based approximate Bayesian learning strategy that aims at producing samples drawn from the posterior distribution of the model parameters. Prior work focused on a distributed implementation of LMC over a multi-access wireless channel via analog modulation. In contrast, this paper proposes quantized federated LMC (FLMC), which integrates one-bit stochastic quantization of the local gradients with channel-driven sampling. Channel-driven sampling leverages channel noise for the purpose of contributing to Monte Carlo sampling, while also serving the role of privacy mechanism. Analog and digital implementations of wireless LMC are compared as a function of differential privacy (DP) requirements, revealing the advantages of the latter at sufficiently high signal-to-noise ratio.

Original languageEnglish (US)
Title of host publication2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication, SPAWC 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781665494557
DOIs
StatePublished - 2022
Externally publishedYes
Event23rd IEEE International Workshop on Signal Processing Advances in Wireless Communication, SPAWC 2022 - Oulu, Finland
Duration: Jul 4 2022Jul 6 2022

Publication series

NameIEEE Workshop on Signal Processing Advances in Wireless Communications, SPAWC
Volume2022-July

Conference

Conference23rd IEEE International Workshop on Signal Processing Advances in Wireless Communication, SPAWC 2022
Country/TerritoryFinland
CityOulu
Period7/4/227/6/22

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering
  • Computer Science Applications
  • Information Systems

Keywords

  • Differential privacy
  • Federated learning
  • Langevin Monte Carlo
  • Power allocation

Fingerprint

Dive into the research topics of 'Leveraging Channel Noise for Sampling and Privacy via Quantized Federated Langevin Monte Carlo'. Together they form a unique fingerprint.

Cite this