HE-Floc: Federated Learning Fingerprint Indoor Localization Method Based on Homomorphic Encryption

Indoor positioning WiFi fingerprint signals and federated learning positioning models can easily lead to user privacy breaches. This article proposes a federated learning fingerprint indoor localization method based on homomorphic encryption. Firstly, federated learning technology breaks down data barriers, realizes the monetization of isolated data value, and utilizes differential privacy to reduce user privacy concerns. Furthermore, by integrating homomorphic encryption into federated learning, the privacy of both data and model parameters is ensured through the homomorphic encryption of the model’s weight parameters. The experimental results show that the scheme can effectively protect the privacy of RSS data and positioning model while ensuring the accuracy of the model.