Federated Model Training With TEE Secure Channels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current federated learning methods face computational inefficiencies and network load issues due to heavy computational loads and long latencies, especially when scaling with an increasing number of participants, and they fail to address the computational and network security concerns, making deployment in large-scale production environments difficult.
Innovation Solution
A federated learning-based model training method utilizing a Trusted Execution Environment (TEE) for secure data processing, enabling plaintext computations and reducing computational load by establishing secure channels between central and edge node devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional federated learning methods are used for model training, then data privacy and security are protected, but computational efficiency deteriorates due to heavy computational loads and long latencies
Solution Approach 1:
The patent introduces a Trusted Execution Environment (TEE) as an intermediary hardware component between the central node and edge nodes. The TEE acts as a secure mediator that enables efficient computational operations while maintaining data privacy, resolving the contradiction between security and computational efficiency by providing a trusted platform for secure computations without requiring heavy encryption overhead
2Reliability
If the number of participants in federated learning is increased, then model training robustness is improved, but network load increases and computational efficiency deteriorates
Solution Approach 1:
The patent segments the computational workload by dividing model training into local operations at edge nodes and centralized aggregation at the central node. Each edge node performs local model updates independently using its own data, then transmits only the updated parameters to the central node. This segmentation reduces network load by transmitting only essential parameters rather than raw data, while still benefiting from increased participant robustness
Solution Approach 2:
The patent changes the nature of data transmission from raw data exchange to parameter exchange. Instead of transmitting complete datasets between participants, the system transmits only the learned model parameters and gradients. This parameter-based communication approach significantly reduces network bandwidth requirements and computational overhead while maintaining training robustness across multiple participants
Data Source
AI summary
A method in one embodiment comprises: communicating, in response to receiving configuration information, with a remote attestation device to attest a trusted execution environment (TEE) of a central node device, the configuration information including a division manner for training sample data for model training; establishing a secure channel with one or more edge node devices in response to passing the attestation for the TEE, wherein the one or more edge node devices each have a TEE; and performing the following operations iteratively until a predetermined condition is satisfied: selecting, from the one or more edge node devices, at least one edge node device for training a model; updating a global model parameter based on a training result received over the secure channel from the selected at least one edge node device; and sending, to the one or more edge node devices, the updated global model parameter over the secure channel.