mlsanityData to 14 Oct 2026

FedRP: A Communication-Efficient Approach for Differentially Private Federated Learning Using Random Projection

arXiv 2509.10041 ↗First posted 12 Sep 2025cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Federated learning (FL) enables collaborative model training without centralizing data, but exchanging high-dimensional updates can expose sensitive information and incur substantial communication costs. We present FedRP, a communication-efficient method combining Gaussian random projection with consensus optimization based on the alternating direction method of multipliers (ADMM). In each round, clients project their model parameters into an m-dimensional space using a shared random matrix hidden from the server, which aggregates only compressed representations. We establish a high-probability guarantee linking projected-space consensus to proximity among client models and derive an (ε,δ)-differential privacy guarantee for each release under bounded ℓ2-sensitivity and a positive lower bound on parameter norms. With randomized parameter vectors and the projection matrix also limits information available to common reconstruction attacks. Experiments on MNIST and CIFAR-10 with LeNet-5 and a custom convolutional network show that FedRP achieves accuracy comparable to FedAvg and consistently exposes noise-perturbed privacy via FedAvg. Because clients transmit m rather than n values per round, FedRP reduces communication by orders of magnitude when m ll n. The results demonstrate a favorable trade-off among accuracy, privacy, and communication efficiency. Code is available at https://github.com/mhnarimani/FedRP

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 41 of its 80 references that are in the map. First posted on arXiv 12 Sep 2025.

Builds on

41 of 80 references.

References · 41

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  3. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
  4. 2019Deep Leakage from Gradientscited by 1,021
  5. 2017Renyi Differential Privacycited by 650
  6. 2018Federated Learning for Mobile Keyboard Predictioncited by 572
  7. 2019Federated Learning with Differential Privacy: Algorithms and Performance Analysiscited by 536
  8. 2017Differentially Private Federated Learning: A Client Level Perspectivecited by 459
  9. 2021See through Gradients: Image Batch Recovery via GradInversioncited by 326
  10. 2013The Composition Theorem for Differential Privacycited by 297
  11. 2018A Hybrid Approach to Privacy-Preserving Federated Learningcited by 210
  12. 2020FetchSGD: Communication-Efficient Federated Learning with Sketchingcited by 179
29 more
  1. 2018Protection Against Reconstruction and Its Applications in Private Federated Learningcited by 130
  2. 2021Local Learning Matters: Rethinking Data Heterogeneity in Federated Learningcited by 114
  3. 2015Global Convergence of ADMM in Nonconvex Nonsmooth Optimizationcited by 98
  4. 2018A General Approach to Adding Differential Privacy to Iterative Training Procedurescited by 84
  5. 2023Federated Learning for Medical Image Analysis: A Surveycited by 78
  6. 2019Differentially Private Meta-Learningcited by 41
  7. 2020Local Differential Privacy and Its Applications: A Comprehensive Surveycited by 41
  8. 2022The Fundamental Price of Secure Aggregation in Differentially Private Federated Learningcited by 39
  9. 2022Cross-Silo Federated Learning: Challenges and Opportunitiescited by 38
  10. 2023FS-Real: Towards Real-World Cross-Device Federated Learningcited by 22
  11. 2023Privacy Amplification via Compression: Achieving the Optimal Privacy-Accuracy-Communication Trade-off in Distributed Mean Estimationcited by 22
  12. 2020FedSKETCH: Communication-Efficient and Private Federated Learning via Sketchingcited by 20
  13. 2022Adap DP-FL: Differentially Private Federated Learning with Adaptive Noisecited by 13
  14. 2021An Introduction to Johnson-Lindenstrauss Transformscited by 11
  15. 2020Characterizing Impacts of Heterogeneity in Federated Learning upon Large-Scale Smartphone Datacited by 9
  16. 2022Deep Leakage from Model in Federated Learningcited by 9
  17. 2014On the Convergence of Alternating Direction Lagrangian Methods for Nonconvex Structured Optimization Problemscited by 8
  18. 2017Learning of Gaussian Processes in Distributed and Communication Limited Systemscited by 5
  19. 2021Federated Learning with Heterogeneous Differential Privacycited by 5
  20. 2018Learning of Tree-Structured Gaussian Graphical Models on Distributed Data under Communication Constraintscited by 4
  21. 2024FedADMM-InSa: An Inexact and Self-Adaptive ADMM for Federated Learningcited by 4
  22. 2022Proximal ADMM for Nonconvex and Nonsmooth Optimizationcited by 3
  23. 2024Linkage on Security, Privacy and Fairness in Federated Learning: New Balances and New Perspectivescited by 3
  24. 2017Random Projection and Its Applicationscited by 2
  25. 2018Structure Learning of Sparse GGMs over Multiple Access Networkscited by 2
  26. 2024Privacy Threats and Countermeasures in Federated Learning for Internet of Things: A Systematic Reviewcited by 2
  27. 2024On-device Federated Learning in Smartphones for Detecting Depression from Reddit Postscited by 2
  28. 2024Bayes' capacity as a measure for reconstruction attacks in federated learningcited by 1
  29. 2025FedCAPrivacy: Privacy-Preserving Heterogeneous Federated Learning with Anonymous Adaptive Clusteringcited by 1

Built on by

Not cited yet.