Privacy-Preserving Data Aggregation Techniques for Enhanced Security in Wireless Sensor Networks

Abstract

In the SecureSense system, the authors propose a machine learning methodology for data aggregation that prioritizes privacy preservation while maintaining data utility. This approach utilizes a combination of techniques such as federated learning and differential privacy. Federated learning allows individual sensor nodes to train a local machine learning model using their own data while keeping it on-device, thus minimizing the need to transmit raw data across the network. This decentralized training process helps preserve privacy by avoiding centralized data aggregation points where sensitive information could be compromised. Additionally, the authors incorporate differential privacy mechanisms to further protect the privacy of individual data points during the aggregation process.