RNN-Based Detection of Fault Attacks on RSA

Abstract

Physical fault injection attacks are becoming an important threat to computer systems, as fault injection equipment becomes more and more accessible. In this work, we propose a new strategy to detect fault attacks in cryptosystems. We use a recurrent neural network (RNN) to detect problems in the program flow caused by injected faults. Our neural network is trained using the instructions of non-faulty operations and therefore, it can protect against both current and future attacks. As a case study, we use two implementations of software RSA. To test the effectiveness of our detector, we propose a collection of fault injection models, where each model represents different types of faults in the instructions. Evaluation results show that we obtain a high detection accuracy in case injected faults lead to changes in the instruction flow and hence, making it difficult to steal secrete keys. Finally, we propose an efficient hardware implementation with only a 6% area overhead compared to a RISC-V processor.

Using Hopfield Networks to Correct Instruction Faults

“…Many fault attacks on instructions can be used to exploit these vulnerabilities. To evaluate the strength of the detection and correction scheme, we use the following fault attack models [19]: bit-level, byte-level, branch-to-opposite, and instruction-to-instruction I/II fault models.…”

Exploiting PUF Variation to Detect Fault Injection Attacks

“…The first group of countermeasures uses redundancy mechanisms to validate a fault-free operation. These mechanisms can be added in time (e.g., algorithm-wise [5], instructionwise [6], and via delay mechanisms [7]) or in space (i.e., extra hardware [8]) [9]. However, these mechanisms result in a high-performance penalty or overhead, which limit their application to the IoT devices with strict resource budgets.…”

Deterministic and Statistical Strategies to Protect ANNs against Fault Injection Attacks

“…Today, this is especially of concern due to the prominence of fault attacks. Although classical fault injection techniques like underfeeding, heating, or shooting EM waves and lasers [8] might not represent a huge threat due to their physical proximity requirements, logical approaches are still possible. Moreover, it also important to protect against faults caused by radiation [9].…”