Encryption Device Segmentation for DPA Resistance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing encryption devices are vulnerable to secondary DPA attacks, which can extract secret keys by analyzing power consumption patterns, despite the use of mask methods for countermeasures.
Innovation Solution
The encryption device employs a segmentation method where masked intermediate data is processed in segments, using multiple pieces of data to perform nonlinear transforms, thereby eliminating the correlation between intermediate data and power consumption, making it resistant to secondary DPA attacks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If mask method is applied to eliminate correlation between power consumption and data, then resistance to DPA is improved, but vulnerability to secondary DPA increases
Solution Approach 1:
The patent applies segmentation by dividing the masked intermediate data into multiple segments before performing nonlinear transformation. Each segment is processed separately through the nonlinear transformation unit, which computes transformation results based on multiple input segments. This segmented processing approach prevents secondary DPA attacks by ensuring that power consumption patterns do not reveal information about the original masked data, as the transformation is distributed across multiple segments rather than concentrated on single data points.
2Reliability
If duplication method is used to segment data for eliminating correlation, then resistance to DPA is improved, but the method remains vulnerable to secondary DPA and increases device complexity
Solution Approach 1:
The patent implements segmentation by dividing masked intermediate data into multiple segments that are then processed through nonlinear transformation. This is different from simple duplication as each segment undergoes distinct transformation operations.
Solution Approach 2:
The patent introduces an intermediary nonlinear transformation unit that operates on segmented masked data. This transformation unit serves as a mediator between the masking operation and the final encryption output, performing nonlinear operations that prevent direct correlation analysis while maintaining encryption integrity.
3Object-affected harmful factors
If nonlinear transform processing is performed on segmented masked data, then resistance to secondary DPA is improved, but processing time increases
Solution Approach 1:
The patent segments masked intermediate data into multiple parts and processes each segment through nonlinear transformation. The nonlinear transformation unit is designed to efficiently handle multiple segments in parallel or sequential manner, computing transformation results based on combinations of input segments.
Solution Approach 2:
The patent performs preliminary masking of intermediate data before segmentation and nonlinear transformation. This preliminary action prepares the data in a form that is more amenable to efficient segmented processing, as the masking operation is completed upfront before the more computationally intensive nonlinear transformation steps.
Data Source
AI summary
According to one embodiment, in an encryption device, a segmentation unit segments masked plain data into pieces of first segmented data. A first processing unit generates pieces of second segmented data from the pieces of first segmented data. A nonlinear transform unit generates pieces of third segmented data transformed from the pieces of second segmented data. A data integration unit integrates fourth segmented data to generate masked encrypted data. An unmask processing unit generates encrypted data from the masked encrypted data. The exclusive OR of the pieces of second segmented data matches the exclusive OR of input data, subjected to nonlinear transform processing and calculated from the plain data, and the first mask. The exclusive OR of the pieces of third segmented data matches the exclusive OR of transform data, obtained when the nonlinear transform processing is performed on the input data, and the second mask.