Method and device for model aggregation in federated learning

By performing local model aggregation between computing nodes, the flexibility, scalability, security and cost problems existing in the model aggregation mechanism that relies on central servers in the prior art are solved, decentralized federated learning is realized, and the flexibility and security of the system are improved.

CN120218272APending Publication Date: 2025-06-27ROBERT BOSCH GMBH
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202311826669.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-12-27
Publication Date
2025-06-27

AI Technical Summary

Technical Problem

The model aggregation mechanism in existing federated learning relies on central servers, with potential issues in flexibility, scalability, security, and cost.

Method used

When another computing node is encountered between each computing node, local model aggregation operations are performed to avoid participation from the central server. By compute data communication and metadata exchange between nodes, model aggregation decisions are generated, and model aggregation is performed locally to obtain the aggregated model.

Benefits of technology

Decentralized federated learning is realized, improving flexibility, scalability and security, reducing costs, and ensuring data privacy protection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120218272A_ABST
    Figure CN120218272A_ABST
Patent Text Reader

Abstract

The present disclosure provides a method for model aggregation in federated learning, where the method is executed by a source computing node and the source computing node is communicatively coupled with a target computing node, the method comprising: receiving metadata of the target computing node; generating a model aggregation decision based on the metadata and an aggregation history of the source computing node; in response to generating a decision indicating execution of model aggregation, performing model aggregation on the local model of the source computing node and the local model of the target computing node to obtain an aggregated model; and updating a local model of the source computing node to the aggregated model.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Cited By

  • Agricultural frequency conversion monitoring method and system fusing LSTM-GRU time sequence prediction and fuzzy logic

    CN120893916A