Homomorphic Encryption Blind Rotation with Preprocessed Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Homomorphic encryption technologies face challenges with large public key sizes and increased computational requirements, particularly in blind rotation operations, which can strain memory and processing resources.
Innovation Solution
A method and apparatus that preprocess vector components of an operand ciphertext based on the order of a polynomial and modulus, using a modified vector to perform blind rotation operations with reduced public keys, incorporating techniques like key-switching and automorphism to optimize computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If blind rotation operation is performed with standard public key size, then cryptographic security is maintained, but memory usage and computational load increase significantly
Solution Approach 1:
The patent segments the blind rotation operation into multiple sub-operations with smaller public keys. Instead of using one large public key for the entire operation, the system divides the rotation into stages, each requiring a smaller public key, thereby reducing peak memory usage while maintaining the overall cryptographic security through the composition of these segmented operations.
Solution Approach 2:
The patent transforms the problem from a single-dimension approach (one large public key) to a multi-dimensional approach (multiple smaller public keys used in sequence). By changing the operational dimension from spatial (memory) to temporal (sequential operations), the system reduces memory requirements while achieving the same cryptographic function through time-multiplexed operations.
2Quantity of substance
If public key size is reduced to decrease memory usage, then memory efficiency improves, but computational complexity increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-processing the ciphertext and organizing data structures before the actual blind rotation operation. This pre-processing step prepares the input data in a format that reduces the computational burden during the rotation operation itself, allowing the use of smaller public keys without excessively increasing overall computational complexity.
Solution Approach 2:
The patent changes parameters by adjusting the decomposition level of the blind rotation operation and optimizing the size of intermediate public keys. By carefully selecting parameter values such as the number of decomposition stages and the size of individual public keys, the system achieves an optimal balance between memory efficiency and computational complexity.
3Measurement precision
If blind rotation operation uses larger public keys, then operation accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The patent applies periodic action by breaking down the blind rotation into periodic sub-operations, each using a smaller public key. The system performs multiple passes with smaller keys rather than one pass with a large key, maintaining accuracy through the cumulative effect of these periodic operations while reducing the processing time required for each individual operation.
Data Source
AI summary
An apparatus with a homomorphic encryption operation includes: one or more processors configured to: generate a modified vector by preprocessing vector components of an operand ciphertext of a blind rotation operation based on an order of a polynomial of an output ciphertext of the blind rotation operation and a modulus of the operand ciphertext; and generate a homomorphic encryption operation result by performing the blind rotation operation based on a public key for performing the blind rotation operation and the modified vector.