Privacy-Preserving Range Query Indexing With Oblivious Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing range query methods on privacy data lack sufficient confidentiality and privacy protection, as they may reveal the user's query behavior and data access mode, making them vulnerable to inference attacks.
Innovation Solution
Implement a dual index system using address and range indexes in a tree-structured database, where data blocks are stored with oblivious random access, and an n-dimensional query vector is used to determine results, masking the access mode and query behavior.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data are stored in an encrypted manner on a cloud server to protect privacy, then data security is improved, but the ability to perform range queries while maintaining confidentiality deteriorates
Solution Approach 1:
The patent segments the index structure into a tree-structured database with multiple nodes, where each node contains encrypted data blocks. This segmentation allows the system to perform range queries on encrypted data by traversing the tree structure and comparing encrypted values, thereby maintaining both data security and query functionality without requiring decryption of the entire dataset.
Solution Approach 2:
The patent introduces an intermediary encrypted index structure that mediates between the encrypted data storage and the range query operation. The index nodes contain encrypted data blocks that serve as intermediaries, allowing the system to perform comparisons and range queries on encrypted values through the index structure without exposing the actual plaintext data, thus preserving confidentiality while enabling efficient querying.
2Ease of operation
If traditional range query methods are used on encrypted data, then query functionality is maintained, but query behavior and access mode are revealed to attackers
Solution Approach 1:
The patent performs preliminary actions by pre-building an encrypted tree-structured index database before any queries are executed. The index nodes are pre-computed and stored in encrypted form, allowing the system to perform range queries by traversing this pre-established structure. This preliminary preparation enables the system to maintain query functionality while preventing attackers from inferring query behavior patterns, as the encrypted index structure masks the actual data access patterns.
3Speed
If data are accessed using conventional methods, then access speed is maintained, but access mode patterns are exposed enabling inference attacks
Solution Approach 1:
The patent introduces dynamics by implementing a configurable tree-structured index where nodes can be dynamically traversed based on query parameters. The system dynamically adjusts the query path through the encrypted index tree, allowing efficient range queries to be performed by navigating only the necessary nodes. This dynamic traversal maintains access speed by avoiding full-scan operations while preventing inference attacks by masking the actual access patterns through the encrypted structure, making it difficult for attackers to determine which data blocks are being accessed.
Data Source
AI summary
Implementations of this specification provide a numerical range query method and apparatus for privacy protection and an index construction method and apparatus for privacy protection, and relate to the field of computer technologies. During index construction, a correspondence between m range indexes and n address indexes is established by using a mapping relationship between n data and m numerical range intervals, and an address index and a range index corresponding to the address index are used as a data block, which is stored in a tree-structured database in an oblivious random access manner. When range query is performed on data, a first range index corresponding to a numerical range to be queried is determined by using the foregoing correspondence, a corresponding first leaf node is computed based on the first range index, several corresponding data blocks are read from the tree-structured database in an oblivious random access manner, a target address index is determined from the data blocks, an n-dimensional query vector is generated based on the target address index and n address indexes, and a query result is determined from n pieces of data by using the n-dimensional query vector.