Side-channel attacks leverage correlations between power consumption and intermediate encryption results to infer encryption keys. Recent studies show that deep learning offers promising results in the context of side-channel attacks. However, neural networks utilized in deep-learning side-channel attacks are complex with a substantial number of parameters and consume significant memory. As a result, it is challenging to perform deep-learning side-channel attacks on resource-constrained devices. In this paper, we propose a framework, TinyPower, which leverages pruning to reduce the number of neural network parameters for side-channel attacks. Pruned neural networks obtained from our framework can successfully run side-channel attacks with significantly fewer parameters and less memory. Specifically, we focus on structured pruning over filters of Convolutional Neural Networks (CNNs). We demonstrate the effectiveness of structured pruning over power and EM traces of AES-128 running on microcontrollers (AVR XMEGA and ARM STM32) and FPGAs (Xilinx Artix-7). Our experimental results show that we can achieve a reduction rate of 98.8% (e.g., reducing the number of parameters from 53.1 million to 0.59 million) on a CNN and still recover keys on XMEGA. For STM32 and Artix-7, we achieve a reduction rate of 92.9% and 87.3% on a CNN respectively. We also demonstrate that our pruned CNNs can effectively perform the attack phase of side-channel attacks on a Raspberry Pi 4 with less than 2.5 millisecond inference time per trace and less than 41 MB memory usage per CNN.
more »
« less
Breaking AES-128: Machine Learning-Based SCA Under Different Scenarios and Devices
Machine learning-based side-channel attacks (MLSCAs) have demonstrated the capability to extract secret keys from AES by learning the correlation between leakages from power traces or timing of AES execution. Previous work has focused on unmasked AES, the captured power traces for profiling and testing have been collected from the same device, and they are primarily implemented on microcontrollers. In this paper, we present a comprehensive MLSCA that considers both masked and unmasked AES running on software and hardware with a side-channel leakage model under four scenarios involving two target boards (Artix-7 XC7AI00T FPGAs and STM32F415 microcontrollers) and different keys for training and testing the model. Our implementation results indicate that support vector machines outperformed other machine learning techniques on masked software and unmasked software AES with only 4 traces. Long short-term memory networks were found to outperform other techniques on unmasked hardware AES (FPGA) with only 283 power traces.
more »
« less
- Award ID(s):
- 2301940
- NSF-PAR ID:
- 10451860
- Date Published:
- Journal Name:
- 2023 IEEE International Conference on Cyber Security and Resilience (CSR)
- Page Range / eLocation ID:
- 564 to 571
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Machine learning (ML) models can be trade secrets due to their development cost. Hence, they need protection against malicious forms of reverse engineering (e.g., in IP piracy). With a growing shift of ML to the edge devices, in part for performance and in part for privacy benefits, the models have become susceptible to the so-called physical side-channel attacks. ML being a relatively new target compared to cryptography poses the problem of side-channel analysis in a context that lacks published literature. The gap between the burgeoning edge-based ML devices and the research on adequate defenses to provide side-channel security for them thus motivates our study. Our work develops and combines different flavors of side-channel defenses for ML models in the hardware blocks. We propose and optimize the first defense based on Boolean masking . We first implement all the masked hardware blocks. We then present an adder optimization to reduce the area and latency overheads. Finally, we couple it with a shuffle-based defense. We quantify that the area-delay overhead of masking ranges from 5.4× to 4.7× depending on the adder topology used and demonstrate a first-order side-channel security of millions of power traces. Additionally, the shuffle countermeasure impedes a straightforward second-order attack on our first-order masked implementation.more » « less
-
This paper investigates countermeasures to side-channel attacks. A dynamic partial reconfiguration (DPR) method is proposed for field programmable gate arrays (FPGAs)s to make techniques such as differential power analysis (DPA) and correlation power analysis (CPA) difficult and ineffective. We call the technique side-channel power resistance for encryption algorithms using DPR, or SPREAD. SPREAD is designed to reduce cryptographic key related signal correlations in power supply transients by changing components of the hardware implementation on-the-fly using DPR. Replicated primitives within the advanced encryption standard (AES) algorithm, in particular, the substitution-box (SBOX)s, are synthesized to multiple and distinct gate-level implementations. The different implementations change the delay characteristics of the SBOXs, reducing correlations in the power traces, which, in turn, increases the difficulty of side-channel attacks. The effectiveness of the proposed countermeasures depends greatly on this principle; therefore, the focus of this paper is on the evaluation of implementation diversity techniques.more » « less
-
This paper investigates countermeasures to side-channel attacks. A dynamic partial reconfiguration (DPR) method is proposed for field programmable gate arrays (FPGAs)s to make techniques such as differential power analysis (DPA) and correlation power analysis (CPA) difficult and ineffective. We call the technique side-channel power resistance for encryption algorithms using DPR, or SPREAD. SPREAD is designed to reduce cryptographic key related signal correlations in power supply transients by changing components of the hardware implementation on-the-fly using DPR. Replicated primitives within the advanced encryption standard (AES) algorithm, in particular, the substitution-box (SBOX)s, are synthesized to multiple and distinct gate-level implementations. The different implementations change the delay characteristics of the SBOXs, reducing correlations in the power traces, which, in turn, increases the difficulty of side-channel attacks. The effectiveness of the proposed countermeasures depends greatly on this principle; therefore, the focus of this paper is on the evaluation of implementation diversity techniques.more » « less
-
This paper investigates countermeasures to side-channel attacks. A dynamic partial reconfiguration (DPR) method is proposed for field programmable gate arrays (FPGAs)s to make techniques such as differential power analysis (DPA) and correlation power analysis (CPA) difficult and ineffective. We call the technique side-channel power resistance for encryption algorithms using DPR, or SPREAD. SPREAD is designed to reduce cryptographic key related signal correlations in power supply transients by changing components of the hardware implementation on-the-fly using DPR. Replicated primitives within the advanced encryption standard (AES) algorithm, in particular, the substitution-box (SBOX)s, are synthesized to multiple and distinct gate-level implementations. The different implementations change the delay characteristics of the SBOXs, reducing correlations in the power traces, which, in turn, increases the difficulty of side-channel attacks. The effectiveness of the proposed countermeasures depends greatly on this principle; therefore, the focus of this paper is on the evaluation of implementation diversity techniques.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/CSR57506.2023.10225009
