Data processing method and apparatus
By establishing a negotiation key between vehicles and roadside units, utilizing hash values and encryption protocols for privacy protection, and designing chain aggregation and cross-validation mechanisms, the privacy and security issues in vehicle-to-everything (V2X) federated learning are resolved, improving the stability and security of V2X cooperative federated learning.
CN117932655BActive Publication Date: 2026-07-21NEUSOFT CORP +1
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NEUSOFT CORP
- Filing Date
- 2023-12-14
- Publication Date
- 2026-07-21
Smart Images
Figure CN117932655B_ABST
Abstract
The present disclosure relates to a data processing method and device; wherein the data processing method applied to a road side unit comprises: in response to receiving a plurality of vehicle model gradients respectively sent by a plurality of vehicles, determining a first hash value according to a negotiation key between the road side unit and the plurality of vehicles, the vehicle model gradient being obtained by encrypting an original model gradient; determining a target road side unit associated with the plurality of vehicle model gradients; obtaining a second hash value corresponding to the target road side unit; in response to obtaining an aggregated model gradient, aggregating the vehicle model gradient according to the aggregated model gradient, the first hash value and the second hash value to obtain a new aggregated model gradient, and sending the new aggregated model gradient to a first road side unit, the first road side unit being a node for aggregating the vehicle model gradient next time. The data processing method and device can realize privacy protection of collaborative federated learning of vehicles and road side units, and improve the stability and security of the collaborative federated learning.
Need to check novelty before this filing date? Find Prior Art