[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to main content

Data Poisoning Attacks in Gossip Learning

  • Conference paper
  • First Online:
Advanced Information Networking and Applications (AINA 2024)

Abstract

Traditional machine learning systems were designed in a centralized manner. In such designs, the central entity maintains both the machine learning model and the data used to adjust the model’s parameters. As data centralization yields privacy issues, Federated Learning was introduced to reduce data sharing and have a central server coordinate the learning of multiple devices. While Federated Learning is more decentralized, it still relies on a central entity that may fail or be subject to attacks, provoking the failure of the whole system. Then, Decentralized Federated Learning removes the need for a central server entirely, letting participating processes handle the coordination of the model construction. This distributed control urges studying the possibility of malicious attacks by the participants themselves. While poisoning attacks on Federated Learning have been extensively studied, their effects in Decentralized Federated Learning did not get the same level of attention. Our work is the first to propose a methodology to assess poisoning attacks in Decentralized Federated Learning in both churn free and churn prone scenarios. Furthermore, in orde r to evaluate our methodology on a case study representative for gossip learning we extended the gossipy simulator with an attack injector module.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Subscribe and save

Springer+ Basic
£29.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Chapter
GBP 19.95
Price includes VAT (United Kingdom)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
GBP 139.99
Price includes VAT (United Kingdom)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
GBP 199.99
Price includes VAT (United Kingdom)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Similar content being viewed by others

Notes

  1. 1.

    https://github.com/makgyver/gossipy/.

  2. 2.

    https://gitlab.lip6.fr/apham/data-poisoning-attacks-in-gossip-learning.

  3. 3.

    This is done to evaluate the model against unseen data, but close to data that were used for adjusting model’s parameters. This allows us to see whether the model generalize well.

  4. 4.

    This means that data is equally distributed among nodes, every node has approximately 25 images of each number.

  5. 5.

    We borrow the idea behind these strategies from Magnien et al. [12], where they use these strategies in order to select nodes to be removed from a graph to study the graph connectivity.

  6. 6.

    https://github.com/makgyver/gossipy/tree/3d655829805fc0dc2f01f5b0862240fca08ffe1c.

  7. 7.

    https://gitlab.lip6.fr/apham/data-poisoning-attacks-in-gossip-learning.

References

  1. Belal, Y., Bellet, A., Mokhtar, S.B., Nitu, V.: PEPPER: empowering user-centric recommender systems over gossip learning. In: Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 6(3), 101:1–101:27, September 2022

    Google Scholar 

  2. Beltrán, E.T.M., Pérez, M.Q., Sánchez, P.M.S., Bernal, S.L., Bovet, G., Pérez, M.G., Pérez, G.M., Celdrán, A.H.: Decentralized federated learning: fundamentals, state of the art, frameworks, trends, and challenges. IEEE Commun. Surv. Tutor., 1 (2023)

    Google Scholar 

  3. Bernstein, J., Wang, Y.-X., Azizzadenesheli, K., Anandkumar, A.: signSGD: compressed optimisation for non-convex problems. In: Proceedings of the 35th International Conference on Machine Learning, pp. 560–569. PMLR, July 2018

    Google Scholar 

  4. Bernstein, J., Zhao, J., Azizzadenesheli, K., Anandkumar, A.: signSGD with majority vote is communication efficient and fault tolerant. In: International Conference on Learning Representations, September 2018

    Google Scholar 

  5. Danner, G., Hegedűs, I., Jelasity, M.: Improving gossip learning via limited model merging. In: Advances in Computational Collective Intelligence, Communications in Computer and Information Science, pp. 351–363. Springer, Cham (2023). https://doi.org/10.1007/978-3-031-41774-0_28

  6. Erdős, P., Rényi, A.: On random graphs I. Publicationes Mathematicae Debrecen 6, 290–297 (1959)

    Article  MathSciNet  Google Scholar 

  7. Giaretta, L., Girdzijauskas, Š.: Gossip Learning: Off the Beaten Path. In: 2019 IEEE International Conference on Big Data (Big Data), pp. 1117–1124, December 2019

    Google Scholar 

  8. Guo, J., Zuo, Y., Wen, C.-K., Jin, S.: User-centric online gossip training for autoencoder-based CSI feedback. IEEE J. Sel. Top. Signal Process. 16(3), 559–572 (2022)

    Article  Google Scholar 

  9. Hegedűs, I., Danner, G., Jelasity, M.: Decentralized learning works: an empirical comparison of gossip learning and federated learning. J. Parallel Distributed Comput. 148, 109–124 (2021)

    Article  Google Scholar 

  10. LeCun, Y., Cortes, C., Burges, C.J.: MNIST handwritten digit database. ATT Labs [Online], 2 (2010)

    Google Scholar 

  11. Liu, P., Xiangrui, X., Wang, W.: Threats, attacks and defenses to federated learning: issues, taxonomy and perspectives. Cybersecurity 5(1), 1–19 (2022)

    Article  Google Scholar 

  12. Magnien, C., Latapy, M., Guillaume, J.-L.: Impact of random failures and attacks on Poisson and power-law random networks. ACM Comput. Surv. 43(3), 13:1–13:31 (2011)

    Google Scholar 

  13. Brendan McMahan, H., Moore, E., Ramage, D., Hampson, S., Arcas, B.J.: Communication-Efficient Learning of Deep Networks from Decentralized Data, February 2016

    Google Scholar 

  14. Ormándi, R., Hegedűs, I., Jelasity, M.: Gossip learning with linear models on fully distributed data. Concurrency Comput. Pract. Exp. 25(4), 556–571 (2013)

    Article  Google Scholar 

  15. Paszke, A., et al.: PyTorch: an imperative style, high-performance deep learning library. In: Advances in Neural Information Processing Systems, vol. 32. Curran Associates, Inc. (2019)

    Google Scholar 

  16. Youyang, Q., Chenhao, X., Gao, L., Xiang, Y., Shui, Yu.: FL-SEC: privacy-preserving decentralized federated learning using SignSGD for the internet of artificially intelligent things. IEEE Internet Things Mag. 5(1), 85–90 (2022)

    Article  Google Scholar 

  17. Watts, D.J., Strogatz, S.H.: Collective dynamics of ‘small-world’ networks. Nature 393(6684), 440–442 (1998)

    Article  Google Scholar 

  18. Wu, C., Yang, X., Zhu, S., Mitra, P.: Mitigating Backdoor Attacks in Federated Learning, January 2021

    Google Scholar 

  19. Xia, G., Chen, J., Chaodong, Yu., Ma, J.: Poisoning attacks in federated learning: a survey. IEEE Access 11, 10708–10722 (2023)

    Article  Google Scholar 

Download references

Acknowledgements

The work presented in this document has received funding from the EU Horizon Europe research and innovation Programme under Grant Agreement No. 101070118.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Alexandre Pham .

Editor information

Editors and Affiliations

Rights and permissions

Reprints and permissions

Copyright information

© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Pham, A., Potop-Butucaru, M., Tixeuil, S., Fdida, S. (2024). Data Poisoning Attacks in Gossip Learning. In: Barolli, L. (eds) Advanced Information Networking and Applications. AINA 2024. Lecture Notes on Data Engineering and Communications Technologies, vol 200. Springer, Cham. https://doi.org/10.1007/978-3-031-57853-3_18

Download citation

Publish with us

Policies and ethics