Processing Device Active Attack Protection via Data Infection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cryptographic systems are vulnerable to both passive and active attacks, as they fail to effectively prevent the extraction of sensitive data, such as encryption keys, through physical attacks that analyze system characteristics or disrupt operations.
Innovation Solution
A method that divides data processing into multiple functions, where each function application is executed simultaneously with an attack detection step, and if an attack is detected, the intermediate data is infected with random or null values, making it difficult for attackers to isolate processing steps and correlate output data with device behavior.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If encryption duplication is carried out in parallel to protect against active attacks, then security against active attacks is improved, but the noise-to-signal ratio becomes observable and usable by attackers through passive attacks
Solution Approach 1:
The patent segments the encryption operation into multiple intermediate processing steps (at least two intermediate results) that are computed sequentially. Each intermediate result is verified against expected values before proceeding to the next step. This segmentation allows the system to detect active attacks at intermediate stages while avoiding the parallel execution that creates observable noise patterns for passive attacks.
Solution Approach 2:
The patent performs preliminary verification of intermediate results during the encryption process. Before completing the full encryption operation, the system checks intermediate values against pre-computed expected values. This preliminary action enables early detection of active attacks without requiring parallel execution, thus avoiding passive attack vulnerabilities while maintaining active attack protection.
2Object-affected harmful factors
If encryption duplication is carried out in series to avoid doubling power consumption, then passive attack vulnerability is reduced, but the attacker can make active attacks at multiple possible times
Solution Approach 1:
The patent divides the encryption process into multiple segmented steps with verification points in between. Each segment produces an intermediate result that is verified against expected values. This segmentation transforms a single vulnerable series execution into multiple controlled stages, maintaining the sequential power consumption profile while limiting active attack effectiveness at each stage through verification.
Solution Approach 2:
The patent implements feedback mechanisms where intermediate results are verified against expected values before proceeding. This feedback loop provides continuous monitoring throughout the sequential encryption process, enabling detection of active attacks at any intermediate stage while maintaining the power-efficient sequential execution structure.
3Measurement precision
If attack detection is performed after each processing function to improve detection speed, then active attack detection capability is improved, but the processing time increases due to multiple verification steps
Solution Approach 1:
The patent performs verification only at intermediate stages rather than continuously monitoring every operation. By selecting specific intermediate results for verification against expected values, the system achieves effective attack detection without the overhead of continuous verification, thus reducing the time penalty while maintaining detection accuracy.
Data Source
Figure 1
Figure 2
AI summary
Processing device implementing a method for protecting the processing (AG) of sensitive data, intended for processing sensitive input data (DE) in order to obtain sensitive output data (DS), wherein data processing is divided into at least two processing functions (f, g). The device comprises: - a processing entity (ET) for successively applying the processing functions from sensitive input data, and obtaining the sensitive output data after applying the last processing function, and - a protection entity (EP) for detecting an attack simultaneously with each application of a processing function, and infecting the data resulting from the application of each processing function if an attack is detected.