Secure and Privacy-Preserving Machine Learning for Federated Learning in Wireless Networks

EasyChair Preprint no. 14087

Abstract

This research investigates advanced techniques for ensuring security and privacy in federated learning environments within wireless networks. Federated learning, which allows decentralized devices to collaboratively train machine learning models without sharing raw data, presents unique challenges related to data privacy, security, and efficiency. This study explores cryptographic methods, differential privacy, and secure multiparty computation to protect sensitive data during the training process. It also examines efficient communication protocols to reduce the overhead and latency associated with federated learning in wireless settings. By addressing these challenges, the research aims to develop robust frameworks that ensure the confidentiality and integrity of data while maintaining high model accuracy and performance. The findings will contribute to the deployment of secure and privacy-preserving federated learning systems in various wireless applications, from smart cities to autonomous vehicles.

Keyphrases: communication protocols, cryptographic methods, data privacy, differential privacy, Federated Learning, privacy preserving, Secure Machine Learning, Secure Multiparty Computation, wireless networks

@Booklet{EasyChair:14087,
  author = {Dylan Stilinki and Hubert Klaus},
  title = {Secure and Privacy-Preserving Machine Learning for Federated Learning in Wireless Networks},
  howpublished = {EasyChair Preprint no. 14087},

  year = {EasyChair, 2024}}