KEM Decapsulation Count Mechanism for Fault Attack Protection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Key Encapsulation Mechanisms (KEMs) based on lattice-based post-quantum cryptography schemes are vulnerable to Fault-Assisted Chosen Ciphertext Attacks, which can cause unnecessary blocking or disabling of processing entities due to decapsulation errors, impairing operability and allowing statistical attacks.
Innovation Solution
Implementing a decapsulation count mechanism that detects and counts errors, allowing a predefined number of decapsulation errors before initiating an error measure, thereby distinguishing between harmless mistakes and malicious attacks, and combining this with a fault detection counter for enhanced security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a fault detection counter is implemented to detect decapsulation errors, then security against fault attacks is improved, but false activation by memory malfunctions causes unnecessary blocking and worsens operability
Solution Approach 1:
The system dynamically adjusts the response to decapsulation errors by implementing a threshold-based counting mechanism. Instead of immediately blocking on the first error, the system allows a predefined number of errors before triggering the error measure, adapting the security response to the actual error pattern observed during operation.
Solution Approach 2:
The decapsulation count mechanism acts as an intermediary between the fault detection counter and the error measure initiation. It intermediates by counting errors and comparing against a threshold, preventing direct activation of error measures by transient faults while still detecting persistent attack patterns.
2Ease of operation
If the fault detection counter is temporarily disabled to prevent false activation, then operability is improved, but security risks increase due to lack of protection during vulnerable periods
Solution Approach 1:
The system maintains continuous monitoring through the decapsulation count mechanism without disabling the fault detection capability. The dynamic threshold-based approach ensures security remains active while accommodating legitimate error variations, eliminating the need to disable protection temporarily.
3Reliability
If immediate error measures are taken upon detecting decapsulation errors, then security response is improved, but repetitive harmless errors cause unnecessary blocking and worsen operability
Solution Approach 1:
The error response is made dynamic through the threshold-based counting mechanism. The system responds immediately to persistent error patterns that exceed the threshold while tolerating transient errors within the allowed count, creating an adaptive security response that maintains productivity.
Solution Approach 2:
The system allows a predefined number of errors (excessive action threshold) before initiating error measures. This partial tolerance of errors enables legitimate operational variations while still providing security against sustained attacks, balancing security response with operational continuity.
Data Source
Figure 1
AI summary
A first processing entity, comprising an implementation covering at least a Key Decapsulation Phase of a Key Encapsulation Mechanism, KEM, constructed to share a key to be shared between a second processing entity and said first processing entity, wherein the first processing entity, when executing the implementation of the Key Decapsulation Phase, executes a decapsulation method, the decapsulation method is characterized in that the decapsulation method comprises: a decapsulation count mechanism coupled to or integrated into the Key Decapsulation Phase of the KEM, comprising steps to: (i) detect and count and store in the processing entity decapsulation errors occurring when the Key Decapsulation Phase is executed, and (ii) when the count of decapsulation errors reaches or exceeds a predefined count limit, initiate an error measure.