Encoder-Multi-Decoder Data Obfuscation With Plausible Decoy Outputs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data encryption methods fail to provide robust protection against malicious actors by allowing plausible decoy data, increasing the risk of sensitive information exposure.
Innovation Solution
Employing an encoder-multi-decoder architecture with neural networks trained using stochastic gradient descent and loss functions to generate embeddings and keys, enabling multiple plausible data representations based on decoder identifiers, ensuring only the correct key reveals the original data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional encryption methods are used, then data security is provided, but malicious actors can identify and attack the encryption system more easily
Solution Approach 1:
The patent creates multiple copies of data in different formats (encrypted data, decoy data, placeholder data) that all appear valid but only one contains the actual information. This confuses attackers by providing multiple targets that look equally valuable, thereby reducing the effectiveness of targeted attacks while maintaining security.
Solution Approach 2:
The patent combines multiple data types and encryption layers into a composite data structure that includes real data, decoy data, and placeholder data. This composite approach makes the encryption system more resilient because attackers must penetrate through multiple layers and distinguish between real and fake data, increasing the complexity and cost of attacks.
2Object-affected harmful factors
If multiple data representations are provided, then the cost and complexity for attackers increases, but the system complexity also increases
Solution Approach 1:
The patent segments data into distinct components (real data, decoy data, placeholder data) that can be independently managed and processed. This segmentation allows the system to handle complexity in a structured way, where each segment serves a specific purpose in increasing attack complexity while the overall system remains organized and manageable.
Solution Approach 2:
The patent changes the parameters of data representation by creating multiple versions of data with different characteristics (encrypted form, decoy form, placeholder form). Each version has different parameters that make it appear valid to attackers, thereby increasing attack complexity. The system manages this through automated parameter transformation rather than manual complexity management.
3Reliability
If decoy data is used to protect sensitive information, then security is enhanced, but data integrity and authenticity become more difficult to verify
Solution Approach 1:
The patent implements feedback mechanisms through metadata and authentication protocols that allow the system to verify which data representations are authentic. The metadata provides feedback about the true nature of each data segment, enabling verification of authenticity while maintaining the security benefits of having multiple data representations. This feedback loop resolves the contradiction by providing a way to distinguish real from fake data.
Solution Approach 2:
The patent introduces metadata and authentication protocols as intermediaries between the multiple data representations and the verification process. These intermediaries carry authentication information that allows the system to verify authenticity without exposing the actual data, thereby maintaining both security and verifiability.
Data Source
AI summary
An encoder receives first data encapsulating second data in a hidden compartment along with a decoder identifier corresponding to either of a first decoder or a second decoder. The encoder then generates an embedding corresponding to the first data. The first decoder decodes the embedding to result in a representation of the first data when the decoder identifier corresponds to the first decoder. The second decoder decodes the embedding to result in a representation of the second data when the decoder identifier corresponds to the second decoder. The decoded embedding can be provided to a consuming application or process. Related apparatus, systems, techniques and articles are also described.