Tamper-Resistant Element Event Counting for Fault Injection Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault injection mitigation techniques for tamper resistant elements fail to account for the frequency of sensor events, leading to unnecessary disabling or resetting of the elements, even when events are not the result of a security attack, due to spurious or benign conditions.
Innovation Solution
Implement an erasable event counter and an always-on register to track sensor events over time, determining when to update a permanent event registry based on event frequency, thereby preventing unnecessary fault mitigation actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fault mitigation actions are performed based on sensor event quantity alone, then security protection is improved, but false positive rate increases causing unnecessary disabling of tamper resistant elements
Solution Approach 1:
The system performs preliminary actions by maintaining an erasable event counter that accumulates sensor events before committing them to the permanent event registry. This preliminary accumulation phase allows the system to filter out spurious events that occur infrequently, while still detecting genuine attack patterns that persist over time. The fault mitigation action is only triggered after the event quantity in the permanent registry exceeds the threshold, ensuring that transient noise does not cause false positives.
Solution Approach 2:
The system uses a dual-counter approach where the erasable event counter tracks all sensor events temporarily, but only a subset of these events (those that persist and exceed the threshold) are committed to the permanent event registry. This partial action approach allows the system to monitor all events for security analysis while only registering significant persistent events that indicate genuine threats, thereby avoiding false positive disabling of the tamper resistant element.
2Reliability
If all sensor events are registered in permanent memory, then security monitoring is improved, but device complexity and false disabling increase
Solution Approach 1:
The event registration system is segmented into two distinct components: an erasable event counter for temporary event accumulation and a permanent event registry for persistent threat recording. This segmentation allows the system to handle events differently based on their significance - transient events are tracked temporarily in the erasable counter and discarded if they don't persist, while genuine threats are committed to the permanent registry. This reduces the burden on permanent memory and simplifies the decision logic for fault mitigation.
Solution Approach 2:
The erasable event counter serves as an intermediary between the sensor events and the permanent event registry. Instead of directly registering all sensor events in permanent memory, the system uses this intermediate buffer to filter and evaluate events. The intermediary accumulates events temporarily, allows for threshold comparison, and only commits significant events to the permanent registry, thereby reducing device complexity and preventing false positives.
3Speed
If fault mitigation threshold is set low, then security response speed is improved, but false positive rate increases
Solution Approach 1:
The system performs preliminary accumulation of sensor events in the erasable event counter before triggering fault mitigation actions. By setting the threshold on the permanent event registry rather than requiring immediate response to every sensor event, the system maintains fast response to genuine persistent threats while filtering out transient noise. The preliminary counting phase enables the system to respond quickly to sustained attacks without false positives from momentary anomalies.
Data Source
AI summary
Systems and techniques are provided for a method for fault injection mitigation. For example, a process for fault injection mitigation may include: receiving a sensor event notification: incrementing, based on the sensor event notification, an erasable event counter: making a first determination that an event quantity of the erasable event counter is below an event threshold; and making a second determination, based on the first determination, not to update a permanent event registry.