Relational Database Table Index Encryption via Initialization Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Relational databases face security vulnerabilities as third parties can access and exploit relationship information between tables, even with encryption, due to the inability to effectively obscure key field entries across different tables.
Innovation Solution
The method involves replacing key field entries in relational database tables with adjusted entries generated using different initialization vectors and encoding them based on index keys, ensuring that common entries are encrypted differently across tables, thus obscuring the relationship structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional encryption is applied to database entries, then data confidentiality is improved, but the relationship structure between tables remains discernible
Solution Approach 1:
The patent applies different initialization vectors to different tables, creating local variations in encryption. Each table's key field entries are encrypted with table-specific IVs, so the same plaintext value produces different ciphertext in different tables. This local differentiation obscures the relationship structure while maintaining data confidentiality.
Solution Approach 2:
The patent introduces asymmetry in the encryption process by using multiple initialization vectors instead of a single uniform encryption key. The encryption transformation becomes asymmetric across tables, where the same data value is transformed differently in each table context, preventing relationship inference.
2Reliability
If encryption keys are used to protect database entries, then access security is improved, but computation time increases
Solution Approach 1:
The patent performs encryption preprocessing on key field entries before database operations. By encrypting the key fields with table-specific initialization vectors in advance, the system secures the relationship structure without requiring complex decryption operations during query execution, thus reducing computation time for routine operations.
3Loss of information
If standard encryption schemes are applied to large databases, then data protection is improved, but the encryption scheme may be discerned from encrypted data
Solution Approach 1:
The patent applies local quality by using table-specific initialization vectors that create unique encryption patterns for each table. This local variation prevents attackers from analyzing encrypted data across tables to discern the encryption scheme, as the same plaintext produces different ciphertexts in different tables, masking the underlying encryption structure.
Data Source
AI summary
A system (108) for securing a database including a first and second tables is provided, where the first table includes a first key field and a first data field and the second table includes a second key field and a second data field. The first and second key fields specifying first and second key field entries having one common entry. The system (108) can replace the first key field entries with first adjusted key field entries based on first initialization vectors. The system (108) also replaces the second key field entries with second adjusted key field entries based on second initialization vectors different from the first initialization vectors. The system further encodes the first and the second adjusted key field entries based on index keys to generate an encrypted database. In the encrypted database, a value of the common entries in the first and second tables are different.