Microcircuit Anomaly Tracking via Dedicated Memory Write
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for protecting microcircuits against fault injection attacks are vulnerable to detection and can be circumvented by attackers, particularly due to differences in current consumption during normal and abnormal operations, which may lead to slowed operations and potential data loss.
Innovation Solution
A method and device that verify normal operation in microcircuits, writing a first datum to a first address when operation is normal and a second datum to a second address when abnormal, with the second address pointing to a dedicated memory area and including an identifier of abnormal operation, ensuring a write of the same number of bytes as in normal operation to maintain anomaly tracking without detectable changes in current consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a trace of abnormal operation is written to memory to identify the origin of the anomaly, then the security anomaly can be tracked and protection functions can be triggered, but the current consumption changes making the write detectable by attackers
Solution Approach 1:
The patent creates a copy of the normal write operation pattern. When an anomaly is detected, instead of writing a small trace byte that changes consumption signature, the system performs a copy of the full normal write operation (same number of bytes, same timing characteristics) but stores it in a dedicated memory area. This copying approach maintains the current consumption signature while still recording the anomaly for later detection.
Solution Approach 2:
The patent designates a specific memory area with special properties for storing anomaly traces. This dedicated memory zone has the quality of being writable under both normal and abnormal conditions, and its content is consulted during card startup to detect anomalies. By localizing the trace storage to this specific area with specialized handling, the system can track anomalies without affecting the overall current consumption profile visible to attackers.
2Object-affected harmful factors
If decoy writes are implemented to simulate writing in a secure zone during normal operation, then the current consumption signature is maintained, but additional processing time slows down the overall operation
Solution Approach 1:
The patent prepares the system in advance by establishing a dedicated memory area and setting up the anomaly detection logic before any attack occurs. The verification module is pre-configured to check for abnormal operations, and the memory structure is pre-arranged to accommodate both normal data and anomaly traces. This preliminary preparation eliminates the need for time-consuming decoy writes during normal operation, as the system is already structured to handle both cases efficiently.
Solution Approach 2:
The patent implements a dynamic write operation that adapts based on the operational state. The write module verifies whether the current operation is normal or abnormal and dynamically adjusts its behavior: during normal operation, it performs standard writes without additional overhead; when an anomaly is detected, it automatically redirects the write to the dedicated memory area. This dynamic adaptation eliminates the need for continuous decoy writes, maintaining high productivity during normal operation while still providing security during abnormal conditions.
3Productivity
If the number of bytes written during anomaly tracking is reduced to minimize impact, then the processing time is reduced, but the current consumption changes making detection easier
Solution Approach 1:
The patent changes the parameters of the write operation based on the operational state. During normal operation, the write operation uses standard parameters (full byte count, standard timing). When an anomaly is detected, the system changes the parameter of write destination (to dedicated memory area) while maintaining the byte count and timing parameters identical to normal operations. This parameter adaptation allows the system to reduce processing overhead while maintaining the current consumption signature that prevents detection by fault injection attacks.
Data Source
Figure 1~2
Figure 3
Figure 4a~4d
AI summary
The invention relates to a method implemented in a microcircuit and comprising a verification step (20) of the normal operation of the microcircuit, followed by a writing step (22; 24) of data (An, A'n) to an address (@A; @A'). When the verification step detects that the microcircuit is not functioning normally, part of the data written during normal operation is replaced by an identifier of the abnormal operation, and the address points to a dedicated memory area of the microcircuit.