AI Model Data Scrambling for Secure Device-to-Device Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data transmission methods in artificial intelligence models lack sufficient security measures to protect sensitive data, such as AI model parameters and user privacy data, which are vulnerable to leakage during transmission.
Innovation Solution
A data transmission method involving scrambling policies to secure AI model data by using scrambling policy indication information, ensuring that data is scrambled based on the capabilities and requirements of the receiving device, thereby enhancing transmission security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is transmitted without scrambling in AI models, then transmission speed and processing efficiency are improved, but data security and confidentiality deteriorate
Solution Approach 1:
The patent applies preliminary action by performing data scrambling before transmission. The sending device scrambles the data in advance using a scrambling policy, so that when data is transmitted, it is already protected. This resolves the contradiction by ensuring security is established before the transmission process begins, without affecting the transmission speed.
Solution Approach 2:
The patent introduces scrambling as an intermediary process between the original data and transmission. The scrambling policy acts as a mediator that transforms readable data into scrambled form, providing a security layer without blocking the transmission channel. This allows both security and transmission efficiency to coexist.
2Reliability
If complex scrambling policies are applied to AI model data, then data security is improved, but computational overhead and processing time increase
Solution Approach 1:
The patent applies dynamics by making the scrambling policy adaptable and configurable. Different scrambling policies can be selected based on the specific data type, security requirements, and computational resources available. This allows the system to balance security needs against processing time constraints dynamically, rather than using a fixed complex scrambling method for all cases.
Solution Approach 2:
The patent uses parameter changes by allowing the scrambling policy to be adjusted according to different scenarios. The scrambling strength, algorithm selection, and processing depth can be modified as parameters to optimize the balance between security and processing time. This resolves the contradiction by enabling flexible adaptation rather than applying uniform complex scrambling.
3Reliability
If data is scrambled during transmission in AI models, then confidentiality is improved, but compatibility and ease of processing deteriorate
Solution Approach 1:
The patent applies feedback by implementing a complete scrambling and unscrambling process. The receiving device receives the scrambled data and uses the corresponding unscrambling policy to restore it to its original form for processing. This feedback mechanism ensures that while confidentiality is maintained during transmission, the ease of processing is preserved at the receiving end through automatic unscrambling.
Solution Approach 2:
The patent uses copying by creating a scrambled version of the data for transmission while maintaining the ability to restore the original. The scrambling process creates a transformed copy that protects confidentiality, but the unscrambling capability ensures the original data can be recovered for processing, thus maintaining ease of operation.
Data Source
AI summary
Embodiments of this application relate to the field of artificial intelligence (AI). Embodiments of this application provide example data transmission methods and related apparatuses. One example method includes obtaining, by a first communication device, scrambling policy indication information, where the scrambling policy indication information indicates a scrambling policy used by the first communication device. The first communication device scrambles data for an AI model based on the scrambling policy indication information to obtain scrambled data. The first communication device sends the scrambled data to a second communication device.