Federal learning security aggregation method and system based on privacy protection reinforcement learning

By employing a hierarchical secure aggregation architecture and a reinforcement learning-driven intelligent group selection mechanism, the computational efficiency and privacy protection issues of federated learning in scenarios with non-independent and identically distributed data are addressed. This enables high-precision model training and security optimization, thereby improving the robustness and efficiency of the system.

CN121770871APending Publication Date: 2026-03-31GUANGZHOU UNIVERSITY +1
2 Cites 2 Cited by

Patent Information

Application Number
CN202512030741.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-12-30
Publication Date
2026-03-31

AI Technical Summary

Technical Problem

Existing federated learning techniques have limitations in terms of computational efficiency, model accuracy, and privacy protection, especially in scenarios with non-independent and identically distributed data, where model convergence is slow, accuracy decreases, and the model is vulnerable to attacks.

Method used

It adopts a hierarchical security aggregation architecture and a reinforcement learning-driven intelligent group selection mechanism. It protects client data privacy through double mask encryption technology, optimizes model training through multi-dimensional state awareness and deep reinforcement learning dynamic selection strategy, and identifies and suppresses malicious attacks by combining an adaptive reputation mechanism.

Benefits of technology

It significantly improves model convergence speed and global accuracy, optimizes communication overhead and computational efficiency, overcomes the high computational complexity of traditional homomorphic encryption and the performance degradation caused by differential privacy protection, and provides a robust solution in untrusted environments.

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

Abstract

The invention discloses a federated learning security aggregation method and system based on privacy protection reinforcement learning, belongs to the technical field of information security, and solves the technical problems of low calculation efficiency and poor model precision in the prior art. The system comprises a central server, G group servers and N clients, the N clients are divided into G groups, each group server manages one client group, and a hierarchical management architecture is formed among the central server, the group servers and the clients; the central server distributes a global model to the client, the client carries out training to obtain a classification model, encrypts the classification model and uploads the classification model to the group server, the group server carries out aggregation and constructs a group state vector, and the aggregation result and the group state vector are uploaded to the central server; and the central server dynamically screens out the client groups participating in global aggregation in the round by using the reinforcement learning agent, aggregates to obtain the global model, synchronously executes the reinforcement learning training process, and sends the updated global model to the client, so that the calculation complexity is low, and the model precision is good.
Need to check novelty before this filing date? Find Prior Art