A Method and Device for Constructing a Cross-Island Heterogeneous Federated Learning System Based on Homomorphic Encryption

By employing homomorphic encryption, sparsity, contribution-aware weight aggregation, and client selection in a cross-island heterogeneous federated learning system, the problems of low training efficiency and privacy leakage caused by heterogeneity are solved, achieving efficient and secure model training.

CN117892322BActive Publication Date: 2026-05-26WUHAN UNIV
0 Cites 0 Cited by

Patent Information

Application Number
CN202410018128.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-04
Publication Date
2026-05-26
Estimated Expiration
2044-01-04

AI Technical Summary

Technical Problem

Existing cross-island heterogeneous federated learning systems suffer from low training efficiency, privacy leaks, and high computational and communication overhead in addressing the heterogeneity problem. In particular, the lag caused by client computing power and data heterogeneity has not been effectively resolved.

Method used

A cross-island heterogeneous federated learning system based on homomorphic encryption is adopted. By packaging homomorphic encryption, sparsity, contribution-aware weight aggregation and client selection mechanism, combined with locality-sensitive hashing dimensionality reduction and K-means clustering, client selection and encrypted weight aggregation are realized, optimizing data transmission and model training process.

Benefits of technology

It effectively solves the problems of low training efficiency and lag caused by client heterogeneity without compromising privacy, while taking into account the efficiency of computation and communication, and improving the security and accuracy of model training.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

This invention discloses a method and device for constructing a cross-island heterogeneous federated learning system based on homomorphic encryption. After local training on the client, the local model is packaged and sparsified at the packet granularity. The sparsed packet data is homomorphically encrypted, and the ciphertext and mask are sent to a parameter server. The parameter server aggregates the data according to the client's ciphertext weights and sends the aggregated ciphertext global model and mask to the client. The client decrypts the global ciphertext and unpacks it to obtain the global model, and updates its local model. The client calculates and returns a sketch of its local model. By clustering the sketch, the fastest client in each cluster is selected as a participant in the next round of training, and the contribution is calculated based on the sketch to obtain the aggregation weights for the next round. The trained local model is then applied to the cross-island heterogeneous federated learning system. This invention can improve the efficiency of homomorphic encryption while protecting privacy, and simultaneously resist the negative impact of lagging on model synchronization.
Need to check novelty before this filing date? Find Prior Art