Data Encoding for Side-Channel-Resistant Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data processing devices are vulnerable to side channel attacks, which exploit measurable physical variables to infer processed data, such as cryptographic keys, due to the distinctiveness of measured values associated with data processing.
Innovation Solution
Incorporating an encoding unit that transforms input data into encoded data words, ensuring that a significant portion of measured values differ minimally from a default value, thereby reducing the correlation with processed data and making side channel attacks more difficult, using a deterministic encoding rule based on encoding parameters without random numbers or entropy sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data words are processed directly without encoding, then processing speed is maintained, but security against side channel attacks deteriorates due to distinctive measured values
Solution Approach 1:
The encoding unit applies encoding rules to data words before they are processed by the processing unit. This preliminary transformation ensures that measured values during processing have reduced information content about the original data, thereby preventing side channel attacks while maintaining normal processing speeds throughout the system.
2Reliability
If encoding is applied to reduce measured value distinctiveness, then security against side channel attacks is improved, but device complexity increases due to additional encoding unit
Solution Approach 1:
The encoding unit serves as an intermediary component between the input interface and the processing unit. It transforms data words into encoded representations that maintain functional equivalence while eliminating security vulnerabilities, thereby protecting the system without requiring fundamental redesign of the core processing architecture.
Solution Approach 2:
The encoding unit changes the parameter representation of data words by applying encoding rules that transform the data into a different format. This parameter transformation ensures that physical measurements during processing no longer directly reveal information about the original data values, thereby securing the system against side channel attacks.
3Ease of manufacture
If encoding rules are made deterministic without random numbers, then ease of manufacture is improved, but adaptability to different attack scenarios deteriorates
Solution Approach 1:
The encoding rules are designed to be dynamic in their application rather than static. The encoding unit can adapt the encoding process based on the specific data being processed and the threat model, allowing the same deterministic hardware implementation to provide security against various side channel attack scenarios without requiring physical redesign.
Data Source
AI summary
A device for processing data includes: an input interface receiving input data; a processing unit processing data; and an encoding unit encoding data words which are obtained as input data at the input interface data in order to obtain encoded data words, the data words being encoded in such a way that a predefined portion of measured values which characterize the encoded data words and/or their processing by the device and which are ascertainable as a function of at least one physical variable of the device has a difference from a default value, the difference being less than or equal to a predefinable threshold value. The encoding unit executes an encoding rule for encoding the data words as a function of at least one encoding parameter, and the processing unit processes the encoded data words.