Performance Analysis of Non-Profiled Side Channel Attacks Based on Convolutional Neural Networks

Do, Ngoc-Tuan; Hoang, Van‐Phuc; Doan, Van‐Sang

doi:10.1109/apccas50809.2020.9301673

Cited by 8 publications

(9 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, one drawback of DDLA is that it is necessary to perform a DL training for each key guess (i.e., 256 training times in the case of AES-128) [6], the higher the data dimension, the more complex the network architecture. Therefore, high dimension data is a serious problem for DDLA, which motivated us to extend the previous work [7] for a comprehensive analysis to assess the efficiency of DDLA techniques in a more complicated context. It is expected that a new DL architecture combining with the HW labelling technique can tackle the high dimension data issue.…”

Section: Motivationmentioning

confidence: 99%

“…In the DL-based SCA context, the popular activation functions used in hidden layers are ELU and RELU, which are computed as formula ( 6) and (7), respectively. Our proposed model used ELU instead of ReLU to avoid the vanishing problem and produce negative outputs for each node in the hidden layer.…”

Section: Multilayer Perceptronmentioning

confidence: 99%

“…Both works have indicated that adding noise is beneficial to implement the DL‐based profiled SCA. In the non‐profiling scenario, we have performed the preliminary investigation of Gaussian noise on the DL‐based SCA attack [7]. In contrast to DL‐based profiled SCA, the results in Ref.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

On the performance of non‐profiled side channel attacks based on deep learning techniques

Hoang

Doan³

et al. 2022

IET Information Security

View full text Add to dashboard Cite

In modern embedded systems, security issues including side-channel attacks (SCAs) are becoming of paramount importance since the embedded devices are ubiquitous in many categories of consumer electronics. Recently, deep learning (DL) has been introduced as a new promising approach for profiled and non-profiled SCAs. This paper proposes and evaluates the applications of different DL techniques including the Convolutional Neural Network and the multilayer perceptron models for non-profiled attacks on the AES-128 encryption implementation. Especially, the proposed network is fine-tuned with different number of hidden layers, labelling techniques and activation functions. Along with the designed models, a dataset reconstruction and labelling technique for the proposed model has also been performed for solving the high dimension data and imbalanced dataset problem. As a result, the DL based SCA with our reconstructed dataset for different targets of ASCAD, RISC-V microcontroller, and ChipWhisperer boards has achieved a higher performance of non-profiled attacks. Specifically, necessary investigations to evaluate the efficiency of the proposed techniques against different SCA countermeasures, such as masking and hiding, have been performed. In addition, the effect of the activation function on the proposed DL models was investigated. The experimental results have clarified that the exponential linear unit function is better than the rectified linear unit in fighting against noise generation-based hiding countermeasure.

show abstract

Section: Motivationmentioning

confidence: 99%

Section: Multilayer Perceptronmentioning

confidence: 99%

See 1 more Smart Citation

On the performance of non‐profiled side channel attacks based on deep learning techniques

Hoang

Doan³

et al. 2022

IET Information Security

View full text Add to dashboard Cite

show abstract

“…Since the changes of the IC design will also lead to corresponding changes in the physical parameters, SCA techniques can detect many types of HTs with different sizes and structures. Moreover, recently, the research on the application of different artificial intelligence (AI) techniques including ML and deep learning (DL) in hardware security has also shown promising results [6][7]. Hence, in this work, we aim to propose a new HT detection technique with power traces and ML.…”

Section: Existing Techniques For Hardware Trojan Detectionmentioning

confidence: 99%

Hardware Trojan Detection Based on Side-Channel Analysis Using Power Traces and Machine Learning

Hoang¹

2022

Proceedings of the Sixth International Conference on Research in Intelligent and Computing

Self Cite

View full text Add to dashboard Cite

With the continuous development of the Integrated Circuit (IC) manufacturing where international outsourcing is one of the main trends, hardware Trojan (HT) has been considered as a serious problem for hardware security in modern electronic systems. This paper presents a novel HT detection method based on the side-channel analysis with power traces and the machine learning (ML) technique. Side-channel information of the AES encryption core was acquired by the power consumption measurement equipment and then classified with Softmax regression. The ML technique was applied to classify and detect the HT. The experimental results have clarified the efficiency of the proposed method.

show abstract

“…One of them is the state-ofthe-art non-profiled based technique called differential deep learning analysis (DDLA). However, we have already shown that it is sensitive with the additional Gaussian noise [23]. Despite attacking successfully the masking countermeasure, this technique is less effective than the conventional secondorder DPA which can overcome the hiding countermeasure by using a large enough number of power traces.…”

Section: Introductionmentioning

confidence: 98%

Low Complexity Correlation Power Analysis by Combining Power Trace Biasing and Correlation Distribution Techniques

Hoang

Pham

2022

IEEE Access

Self Cite

View full text Add to dashboard Cite

Side channel attack (SCA) is a class of crypt-analytic attacks for security evaluation of cryptographic and embedded microprocessor implementations. Among several SCA approaches, the correlation power analysis (CPA) is an efficient way to recover the secret key of the specific cryptographic algorithms running on the target devices such as embedded microprocessors. However, the evaluation process is time-consuming since a large number of traces are required to overcome the impact of noise. Hence, this paper proposes new methods to reduce the computation time by using Point of Interest (POI) extractor with the power trace biasing technique and the correlation distribution for the low complexity correlation power analysis (CPA). The theoretical explanations are provided and the experiments on different platforms such as ASCAD and RISC-V processor based databases are conducted to justify the proposed techniques. Especially, our experiments are performed with different protected schemes such as masking, hiding and combined hiding-masking techniques. The experimental results indicate that our proposed methods provide reliable results in comparison with the standard CPA. By using only a half of the power traces for taking the POIs, our first proposal not only decreases the execution time approximately by half but also enhances the success rate of the attack. Moreover, the second method based on power trace biasing technique is proposed in order to achieve better results and reduce the number of traces needed for selecting the POIs. With only 28.9% of given power traces needed, our second proposed technique reduces the execution time to only 2.6 times of the standard CPA.INDEX TERMS Correlation power analysis, side channel attack, embedded microprocessor security, power trace biasing, correlation distribution.

show abstract

Performance Analysis of Non-Profiled Side Channel Attacks Based on Convolutional Neural Networks

Cited by 8 publications

References 6 publications

On the performance of non‐profiled side channel attacks based on deep learning techniques

On the performance of non‐profiled side channel attacks based on deep learning techniques

Hardware Trojan Detection Based on Side-Channel Analysis Using Power Traces and Machine Learning

Low Complexity Correlation Power Analysis by Combining Power Trace Biasing and Correlation Distribution Techniques

Contact Info

Product

Resources

About