Cryptographic Data Deformation for Fault Attack Protection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cryptographic processing devices face challenges in efficiently generating dummy data to protect against fault attacks and side-channel analysis, leading to performance bottlenecks and security vulnerabilities, particularly in CPU-based systems without hardware support for dummy data generation.
Innovation Solution
A data processing device that uses a pseudo-random number generator (PRNG) to deform and generate dummy data, allowing for efficient CPU-based dummy computation by selecting a random initial seed, generating a sequence of seeds, and using these seeds to create both real and dummy data blocks, which are processed cryptographically, enabling protection against attacks without the need for full data transfer between the CPU and crypto module.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dummy calculations are introduced to protect against fault attacks and side-channel analysis, then security is improved, but processing power is consumed due to generation and execution of dummy data
Solution Approach 1:
The patent uses a pseudo-random number generator to create dummy data that copies the structure and format of real cryptographic data without requiring actual cryptographic processing. This allows the system to generate convincing dummy data that consumes minimal processing power while still providing security against fault attacks and side-channel analysis.
Solution Approach 2:
The patent employs disposable, easily generated pseudo-random values as dummy data that are used once and discarded. These cheap, short-lived dummy data blocks provide security protection without the overhead of generating and managing complex, long-lived dummy datasets, thereby reducing processing power consumption.
2Reliability
If dummy data is generated and processed to protect against attacks, then security is improved, but execution time increases
Solution Approach 1:
The patent performs preliminary generation of pseudo-random dummy data blocks before cryptographic processing begins. By preparing dummy data in advance using fast pseudo-random number generation, the system avoids time-consuming dummy data generation during critical cryptographic operations, thereby reducing overall execution time while maintaining security.
Solution Approach 2:
The patent changes the parameters of dummy data generation by using pseudo-random number generators with configurable seed values and data block sizes. This allows optimization of dummy data characteristics to match specific security requirements while minimizing processing overhead and execution time.
3Manufacturing precision
If full data transfer between CPU and crypto module is used for dummy computation, then dummy data generation is accurate, but data transfer overhead increases
Solution Approach 1:
The patent extracts the dummy data generation function from the main cryptographic processing path by implementing it within the crypto module itself using a hardware or firmware-based pseudo-random number generator. This eliminates the need for repeated data transfer between CPU and crypto module for dummy data generation, reducing data transfer overhead while maintaining generation accuracy.
Solution Approach 2:
The patent introduces an intermediary pseudo-random number generator within the crypto module that serves as a local source of dummy data. This intermediary eliminates the need for external CPU intervention in dummy data generation, reducing data transfer time while ensuring the dummy data meets cryptographic requirements through proper seeding and generation algorithms.
Data Source
AI summary
According to an embodiment, a data processing device is described comprising a deformer configured to deform a first data block in accordance with a first seed, a seed generator configured to generate a sequence of second seeds, wherein the sequence of second seeds comprises the first seed, a cryptographic processor configured to receive the deformed first data block and, for each second seed, to deform the deformed first data block in accordance with the second seed, to generate a sequence of second data blocks and to cryptographically process each second data block of the sequence of second data blocks to generate a sequence of processed data blocks and an extractor configured to extract a result of the cryptographic processing of the first data block from the sequence of processed data blocks.