基于流形學(xué)習(xí)能量數(shù)據(jù)預(yù)處理的模板攻擊優(yōu)化方法

袁慶軍; 王安; 王永娟; 王濤

doi:10.11999/JEIT190598

基于流形學(xué)習(xí)能量數(shù)據(jù)預(yù)處理的模板攻擊優(yōu)化方法

doi: 10.11999/JEIT190598 cstr: 32379.14.JEIT190598

袁慶軍^{1, 2},
王安³,
王永娟^{1, 2, ,},
王濤^{1, 2}

1.
戰(zhàn)略支援部隊信息工程大學(xué) 鄭州 450001
2.
河南省網(wǎng)絡(luò)密碼技術(shù)重點實驗室鄭州 450001
3.
北京理工大學(xué)計算機學(xué)院北京 100081

基金項目: 國家自然科學(xué)基金(61872040)，河南省網(wǎng)絡(luò)密碼技術(shù)重點實驗室開放基金(LNCT2019-S02)，“十三五”國家密碼發(fā)展基金(MMJJ20170201)

詳細信息

作者簡介:
袁慶軍：男，1993年生，講師，研究方向為機器學(xué)習(xí)側(cè)信道分析

王安：男，1983年生，副教授，研究方向為側(cè)信道分析與防護技術(shù)

王永娟：女，1982年生，研究員，研究方向為側(cè)信道分析與密碼系統(tǒng)安全

王濤：男，1995年生，碩士生，研究方向為機器學(xué)習(xí)側(cè)信道分析

通訊作者:
王永娟　pinkywyj@163.com

中圖分類號: TP309.7
計量
- 文章訪問數(shù): 2999
- HTML全文瀏覽量: 1333
- PDF下載量: 103
- 被引次數(shù): 0
出版歷程
- 收稿日期: 2019-08-07
- 修回日期: 2019-10-31
- 網(wǎng)絡(luò)出版日期: 2019-11-27
- 刊出日期: 2020-08-18

An Improved Template Analysis Method Based on Power Traces Preprocessing with Manifold Learning

Qingjun YUAN^{1, 2},
An WANG³,
Yongjuan WANG^{1, 2
, ,},
Tao WANG^{1, 2}

1.
PLA Strategic Support Force Information Engineering University , Zhengzhou 450001, China
2.
Henan Key Laboratory of Network Cryptography Technology, Zhengzhou 450001, China
3.
School of Computer Science, Beijing Institute of Technology, Beijing 100081, China

Funds: The National Natural Science Foundation of China (61872040), The Fund of Henan Key Laboratory of Network Cryptography Technology (LNCT2019-S02), The National Cryptographic Development Fund of the 13th Five-Year Plan (MMJJ20170201)

摘要

摘要: 能量數(shù)據(jù)作為模板攻擊過程中的關(guān)鍵對象，具有維度高、有效維度少、不對齊的特點，在進行有效的預(yù)處理之前，模板攻擊難以奏效。針對能量數(shù)據(jù)的特性，該文提出一種基于流形學(xué)習(xí)思想進行整體對齊的方法，以保留能量數(shù)據(jù)的變化特征，隨后通過線性投影的方法降低數(shù)據(jù)的維度。使用該方法在Panda 2018 challenge1標準數(shù)據(jù)集進行了驗證，實驗結(jié)果表明，該方法的特征提取效果優(yōu)于傳統(tǒng)的PCA和LDA方法，能大幅度提高模板攻擊的成功率。最后采用模板攻擊恢復(fù)密鑰，僅使用兩條能量跡密鑰恢復(fù)成功率即可達到80%以上。
- 信息安全 /
- 模板攻擊 /
- 流形學(xué)習(xí) /
- 能量數(shù)據(jù) /
- 對齊算法 /
- 降維算法
Abstract: As the key object in the process of template analysis, power traces have the characteristics of high dimension, less effective dimension and unaligned. Before effective preprocessing, template attack is difficult to work. Based on the characteristics of energy data, a global alignment method based on manifold learning is proposed to preserve the changing characteristics of power traces, and then the dimensionality of data is reduced by linear projection. The method is validated in Panda 2018 challenge1 standard datasets respectively. The experimental results show that the feature extraction effect of this method is superior over that of traditional PCA and LDA methods. Finally, the method of template analysis is used to recover the key, and the recovery success rates can reach 80% with only two traces.
- Information security /
- Template analysis /
- Manifold learning /
- Power traces /
- Alignment algorithm /
- Dimension reduction algorithm

HTML全文

圖 1 PANDA 2018 Challenge1 前3條能量跡

下載: 全尺寸圖片幻燈片

圖 2 PANDA 2018 Challenge1 能量跡與密鑰相關(guān)系數(shù)

下載: 全尺寸圖片幻燈片

圖 3 PANDA 2018 Challenge1能量數(shù)據(jù)對齊后

下載: 全尺寸圖片幻燈片

圖 4 PANDA 2018 Challenge1 能量跡降維后

下載: 全尺寸圖片幻燈片

圖 5 PANDA 2018 Challenge1 能量跡PCA-20和LDA-20降維后

下載: 全尺寸圖片幻燈片

表 1 向量矩陣計算算法

　輸入：能量數(shù)據(jù)${T_\alpha } = {\rm{\{ } }{T_i},0 \le i \le \alpha ,i \in N\}$，對齊參數(shù)$k$。

　輸出：對齊后的能量數(shù)據(jù)${T'_\alpha }$

　(1)　for j in range(α), do

　(2)　　計算與${T_j}$ 歐式距離最近的$k$條能量跡${\rm{\{ }}{T_{j1}},{T_{j2}}, ··· ,{T_{jk}}\} $;

　(3)　end

　(4)　for j in range (α), do

　(5)　　計算關(guān)系向量矩陣${ {{W} }_{{j} } } = \dfrac{ {\left( { {{C} }_i^{ - 1} \cdot { {{1} }_k} } \right)} }{ { { {{\textit{1} } } }_k^{\rm T} \cdot {{C} }_i^{ - 1} \cdot { {{{\textit{1}}} }_k} } }$，其中${ { C}_i} $為
　　　 ${\rm{\{ }}{T_{j1}},{T_{j2}}, ··· ,{T_{jk}}\} $的協(xié)方差矩陣，${ {{{\textit{1}}} }_k}$為$k$維全1向量；

　(6)　end

　(7)　計算矩陣${{M} } = ({{ {I} } } - {{W} }){({{I} } - {{W} })^{\rm{T} } }$;

　(8)　設(shè)$\beta = \alpha /2$從矩陣M中選擇較小的$\beta $個特征值，記為${{{M}}_\beta }$，
　　　計算${T'_\alpha } = T \cdot {{{M}}_\beta }$;

　(9)　return ${T_\alpha }^\prime $。

下載: 導(dǎo)出CSV

表 2 PANDA 2018 Challenge1數(shù)據(jù)集預(yù)處理后方差(×10⁴)表(漢明重量不同)

方差	0	1	3	7	15	31	63	127	255
0	4.08	10.99	14.31	16.61	9.80	15.80	18.32	13.02	10.19
1	10.99	2.67	12.49	8.83	7.34	9.50	11.48	5.00	6.33
3	14.31	12.49	8.53	13.62	15.21	12.67	11.73	13.00	15.81
7	16.61	8.83	13.62	3.62	16.24	8.13	11.60	4.99	10.73
15	9.80	7.34	15.21	16.24	4.23	12.21	12.85	9.23	9.84
31	15.80	9.50	12.67	8.13	12.21	4.17	11.62	8.86	9.61
63	18.32	11.48	11.73	11.60	12.85	11.62	4.54	9.26	9.73
127	13.02	5.00	13.00	4.99	9.23	8.86	9.26	1.97	5.23
255	10.19	6.33	15.81	10.73	9.84	9.61	9.73	5.23	4.26

下載: 導(dǎo)出CSV

表 3 PANDA 2018 Challenge1數(shù)據(jù)集預(yù)處理后方差(×10⁴)表(漢明重量相同)

方差	7	11	13	14	19	35	67	131	224
7	3.62	11.23	23.70	12.19	13.35	13.52	11.55	14.04	9.86
11	11.23	2.60	18.80	11.73	12.07	11.85	12.43	10.97	10.21
13	23.70	18.80	31.91	23.04	27.09	22.52	23.58	56.33	19.22
14	12.19	11.73	23.04	3.89	12.54	9.52	14.47	14.96	12.70
19	13.35	12.07	27.09	12.54	4.78	13.86	15.33	17.68	11.98
35	13.52	11.85	22.52	9.52	13.86	3.15	15.07	15.10	10.67
67	11.55	12.43	23.58	14.47	15.33	15.07	4.98	17.73	9.50
131	14.04	10.97	56.33	14.96	17.68	15.10	17.73	37.04	20.31
224	9.86	10.21	19.22	12.70	11.98	10.67	9.50	20.31	3.91

下載: 導(dǎo)出CSV

表 4 PANDA 2018 Challenge1數(shù)據(jù)集PCA-20處理后方差(×10⁴)表(漢明重量不同)

方差	0	1	3	7	15	31	63	127	255
0	33.00	27.97	30.58	29.58	28.96	30.91	29.07	31.04	31.06
1	27.97	13.72	15.97	16.05	15.23	16.10	15.99	20.49	14.26
3	30.58	15.97	13.79	16.97	15.97	17.57	15.58	23.60	16.56
7	29.58	16.05	16.97	17.04	16.70	17.60	17.34	22.65	17.31
15	28.96	15.23	15.97	16.70	14.53	16.83	16.07	21.60	16.43
31	30.91	16.10	17.57	17.60	16.83	16.64	16.65	22.57	17.06
63	29.07	15.99	15.58	17.34	16.07	16.65	15.41	22.27	16.76
127	31.04	20.49	23.60	22.65	21.60	22.57	22.27	24.36	22.35
255	31.06	14.26	16.56	17.31	16.43	17.06	16.76	22.35	13.91

下載: 導(dǎo)出CSV

表 5 PANDA 2018 Challenge1數(shù)據(jù)集LDA-20處理后方差(×10⁴)表(漢明重量不同)

方差	0	1	3	7	15	31	63	127	255
0	0.95	1.21	0.93	0.99	1.07	1.09	1.08	1.12	1.13
1	1.21	1.13	1.07	1.17	1.20	1.11	1.24	1.15	1.20
3	0.93	1.07	0.65	0.90	0.99	0.93	1.00	1.05	1.01
7	0.99	1.17	0.90	0.84	0.97	1.02	1.10	1.09	1.06
15	1.07	1.20	0.99	0.97	0.92	1.08	1.17	1.16	1.11
31	1.09	1.11	0.93	1.02	1.08	0.89	1.10	1.10	1.02
63	1.08	1.24	1.00	1.10	1.17	1.10	1.07	1.18	1.15
127	1.12	1.15	1.05	1.09	1.16	1.10	1.18	0.98	1.15
255	1.13	1.20	1.01	1.06	1.11	1.02	1.15	1.15	0.97

下載: 導(dǎo)出CSV

參考文獻(23)

KOCHER P, JAFFE J, and JUN B. Differential power analysis[C]. The 13th Annual International Cryptology Conference, Santa Barbara, USA, 1999: 388–397. doi: 10.1007/3-540-48405-1_25.

ERNST D and MARTIN S. The common criteria for information technology security evaluation: Implications for China’s policy on information security standards[R]. East-West Center Working Papers, No. 108, 2010. doi: 10.2139/ssrn.2770146.

VAN TILBORG H C A and JAJODIA S. Encyclopedia of Cryptography and Security[M]. Boston: Springer, 2011: 468–471. doi: 10.1007/978-1-4419-5906-5.

CHARI S, RAO J R, and ROHATGI P. Template attacks[C]. The 4th International Workshop on Cryptographic Hardware and Embedded Systems, Redwood Shores, USA, 2002: 13–28. doi: 10.1007/3-540-36400-5_3.

BRIER E, CLAVIER C, and OLIVIER F. Correlation power analysis with a leakage model[C]. The 6th International Workshop on Cryptographic Hardware and Embedded Systems, Cambridge, USA, 2004: 16–29. doi: 10.1007/978-3-540-28632-5_2.

BOGDANOV A. Improved side-channel collision attacks on AES[C]. The 14th International Workshop on Selected Areas in Cryptography, Ottawa, Canada, 2007: 84–95. doi: 10.1007/978-3-540-77360-3_6.

RIVAIN M, PROUFF E, and DOGET J. Higher-order masking and shuffling for software implementations of block ciphers[C]. The 11th International Workshop on Cryptographic Hardware and Embedded Systems, Lausanne, Switzerland, 2009: 171–188. doi: 10.1007/978-3-642-04138-9_13.

CORON J S and KIZHVATOV I. Analysis and improvement of the random delay countermeasure of CHES 2009[C]. The 12th International Workshop on Cryptographic Hardware and Embedded Systems, Santa Barbara, USA, 2010: 95–109. doi: 10.1007/978-3-642-15031-9_7.

黃海, 馮新新, 劉紅雨, 等. 基于隨機加法鏈的高級加密標準抗側(cè)信道攻擊對策[J]. 電子與信息學(xué)報, 2019, 41(2): 348–354. doi: 10.11999/JEIT171211

HUANG Hai, FENG Xinxin, LIU Hongyu, et al. Random addition-chain based countermeasure against side-channel attack for advanced encryption standard[J]. Journal of Electronics &Information Technology, 2019, 41(2): 348–354. doi: 10.11999/JEIT171211

LERMAN L, BONTEMPI G, and MARKOWITCH O. Power analysis attack: An approach based on machine learning[J]. International Journal of Applied Cryptography, 2014, 3(2): 97–115. doi: 10.1504/IJACT.2014.062722

ARCHAMBEAU C, PEETERS E, STANDAERT F X, et al. Template attacks in principal subspaces[C]. The 8th International Workshop on Cryptographic Hardware and Embedded Systems, Yokohama, Japan, 2006: 1–14. doi: 10.1007/11894063_1.

STANDAERT F X and ARCHAMBEAU C. Using subspace-based template attacks to compare and combine power and electromagnetic information leakages[C]. The 10th International Workshop on Cryptographic Hardware and Embedded Systems, Washington, USA, 2008: 411–425. doi: 10.1007/978-3-540-85053-3_26.

HETTWER B, GEHRER S, and GüNEYSU T. Applications of machine learning techniques in side-channel attacks: A survey[J]. Journal of Cryptographic Engineering, 2020(10): 85–95. doi: 10.1007/s13389-019-00212-8

王燚, 吳震, 藺冰. 對加掩加密算法的盲掩碼模板攻擊[J]. 通信學(xué)報, 2019, 40(1): 1–14. doi: 10.11959/j.issn.1000-436x.2019007

WANG Yi, WU Zhen, and LIN Bing. Blind mask template attacks on masked cryptographic algorithm[J]. Journal on Communications, 2019, 40(1): 1–14. doi: 10.11959/j.issn.1000-436x.2019007

CAGLI E, DUMAS C, and PROUFF E. Convolutional neural networks with data augmentation against jitter-based countermeasures: Profiling attacks without pre-processing[C]. The 19th International Conference on Cryptographic Hardware and Embedded Systems, Taipei, China, 2017: 45–68. doi: 10.1007/978-3-319-66787-4_3.

ZHOU Yuanyuan and STANDAERT F X. Deep learning mitigates but does not annihilate the need of aligned traces and a generalized ResNet model for side-channel attacks[J]. Journal of Cryptographic Engineering, 2020(10): 135–162. doi: 10.1007/s13389-019-00209-3

WANG Z. The data of PANDA challeng1[EB/OL]. https://github.com/kistoday/Panda2018/tree/master/challeng1, 2019.

CRIMINISI A, SHOTTON J, and KONUKOGLU E. Decision forests: A unified framework for classification, regression, density estimation, manifold learning and semi-supervised learning[J]. Foundations and Trends? in Computer Graphics and Vision, 2012, 7(2/3): 81–227. doi: 10.1561/0600000035

HOMMA N, NAGASHIMA S, IMAI Y, et al. High-resolution side-channel attack using phase-based waveform matching[C]. The 8th International Workshop on Cryptographic Hardware and Embedded Systems - CHES 2006, Yokohama, Japan, 2006: 187–200. doi: 10.1007/11894063_15.

GUILLEY S, KHALFALLAH K, LOMNE V, et al. Formal framework for the evaluation of waveform resynchronization algorithms[C]. The 5th IFIP WG 11.2 International Workshop on Information Security Theory and Practice. Security and Privacy of Mobile Devices in Wireless Communication, Heraklion, Greece, 2011: 100–115. doi: 10.1007/978-3-642-21040-2_7.

MATEOS E and GEBOTYS C H. A new correlation frequency analysis of the side channel[C]. The 5th Workshop on Embedded Systems Security, Scottsdale, USA, 2010: 4. doi: 10.1145/1873548.1873552.

GIERLICHS B, LEMKE-RUST K, and PAAR C. Templates vs. stochastic methods: A performance analysis for side channel cryptanalysis[C]. The 8th International Workshop on Cryptographic Hardware and Embedded Systems, Yokohama, Japan, 2006: 15–29. doi: 10.1007/11894063_2.

ZHANG Hailong and ZHOU Yongbin. Template attack vs. stochastic model: An empirical study on the performances of profiling attacks in real scenarios[J]. Microprocessors and Microsystems, 2019, 66: 43–54. doi: 10.1016/j.micpro.2019.02.010

相關(guān)文章

施引文獻

資源附件(0)

訪問統(tǒng)計