CN114531232B - Multichannel bypass signal safety analysis and detection system - Google Patents
Multichannel bypass signal safety analysis and detection system Download PDFInfo
- Publication number
- CN114531232B CN114531232B CN202111673193.3A CN202111673193A CN114531232B CN 114531232 B CN114531232 B CN 114531232B CN 202111673193 A CN202111673193 A CN 202111673193A CN 114531232 B CN114531232 B CN 114531232B
- Authority
- CN
- China
- Prior art keywords
- channel
- attack
- bypass
- signal
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 62
- 238000004458 analytical method Methods 0.000 title claims abstract description 38
- 230000004927 fusion Effects 0.000 claims abstract description 64
- 238000000034 method Methods 0.000 claims abstract description 55
- 238000011156 evaluation Methods 0.000 claims abstract description 31
- 238000007781 pre-processing Methods 0.000 claims abstract description 31
- 238000012360 testing method Methods 0.000 claims abstract description 5
- 239000013598 vector Substances 0.000 claims description 19
- 238000000546 chi-square test Methods 0.000 claims description 13
- 230000008859 change Effects 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 8
- 238000000354 decomposition reaction Methods 0.000 claims description 7
- 238000005315 distribution function Methods 0.000 claims description 5
- 238000005259 measurement Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000013139 quantization Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 2
- 230000000007 visual effect Effects 0.000 claims description 2
- 150000004667 medium chain fatty acids Chemical class 0.000 claims 6
- 230000004931 aggregating effect Effects 0.000 claims 1
- 230000000873 masking effect Effects 0.000 claims 1
- 238000006243 chemical reaction Methods 0.000 abstract description 2
- 230000007246 mechanism Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000013461 design Methods 0.000 description 5
- 238000011084 recovery Methods 0.000 description 4
- 238000012163 sequencing technique Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 238000007476 Maximum Likelihood Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005670 electromagnetic radiation Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/002—Countermeasures against attacks on cryptographic mechanisms
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/20—Network architectures or network communication protocols for network security for managing network security; network security policies in general
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/08—Key distribution or management, e.g. generation, sharing or updating, of cryptographic keys or passwords
- H04L9/088—Usage controlling of secret information, e.g. techniques for restricting cryptographic keys to pre-authorized uses, different access levels, validity of crypto-period, different key- or password length, or different strong and weak cryptographic algorithms
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/36—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols with means for detecting characters not meant for transmission
Landscapes
- Engineering & Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Storage Device Security (AREA)
Abstract
The invention discloses a multi-channel bypass signal security analysis and detection system, which can integrate and utilize physical information of multiple or multiple channels leaked during the operation of a cryptographic module, and complete analysis and evaluation of the capability of resisting bypass attack, leakage characteristics and security level of the cryptographic module by means of multi-channel bypass signal preprocessing, multi-channel fusion attack, multi-channel leakage detection, key enumeration attack, key ordering evaluation and the like. The invention can pre-examine and improve the actual password module product before the product is sent to a professional security assessment mechanism or a laboratory, reduce the assessment period and the test cost of the product, and accelerate the conversion of the product into a practical progress. Compared with the threat that the bypass analysis detection system in the current market only can utilize single-channel information leakage and does not evaluate key enumeration attack, the method can provide more comprehensive depth safety evaluation conclusion and suggestion.
Description
Technical Field
The invention belongs to the technical field of information security, and particularly relates to security analysis and detection of a cryptographic module.
Background
Bypass attacks pose a serious threat to the physical security of a cryptographic module by exploiting the physical information leakage (e.g., energy consumption, electromagnetic radiation, etc.) of the cryptographic module to recover secret information. Therefore, it is important and necessary to evaluate the physical security level of an actual cryptographic module product against bypass analysis before it is marketed. The current bypass attack research is basically a single-channel attack, and reflects the security level of the cryptographic module under the condition that a certain channel of certain bypass information leaks. The existing bypass security analysis and detection products in the market use single-channel bypass information leakage to perform security assessment on the cryptographic module without assessing the threat of key enumeration attack, such as SmartSIC analysis platform of Inspector, secureIC of Riscure company.
However, in real-world applications, the bypass leakage of the cryptographic module is multi-source, simultaneous and multi-directional, the security threat faced by the cryptographic module is multi-dimensional, and the bypass security analysis and evaluation requirements are more comprehensive and deeper.
Disclosure of Invention
The present invention aims to solve the above-mentioned problems of the prior art and provide a multi-channel bypass signal security analysis and detection system. Based on the multi-channel bypass leakage fusion technology, the invention comprehensively utilizes the bypass information leakage of a plurality of channels, develops a plurality of multi-channel bypass security analysis and detection methods, including multi-channel bypass signal preprocessing, multi-channel fusion attack, multi-channel bypass leakage detection, key enumeration attack based on multi-channel leakage fusion, key sequencing evaluation and the like, and can provide more comprehensive depth security assessment conclusion and suggestion. The method can meet the requirements of real safety analysis and evaluation, and has great innovation and practicability.
The technical solution for realizing the purpose of the invention is as follows: a multi-channel bypass signal security analysis and detection system, the system comprising:
the file processing module is used for reading the multi-channel bypass signal and the plaintext and ciphertext information leaked during the operation of the password module and processing the multi-channel bypass signal and the plaintext and ciphertext information to generate a data matrix;
the signal preprocessing module is used for preprocessing bypass signals, including multi-channel bypass signal alignment, multi-channel bypass signal fusion degree judgment and second-order attack preprocessing;
the multi-channel fusion attack module is an MCFA module and is used for evaluating the security level of the cryptographic module by recovering the key realized by the cryptographic algorithm;
the multi-channel leakage detection module is an LD module and is used for screening leakage characteristic points containing secret information in the bypass signal and eliminating non-characteristic point interference;
the key enumeration attack and key ordering evaluation module is used for estimating the security level of the cryptographic module and giving feasibility suggestions for when the cryptographic module changes keys or algorithms.
Further, the system also comprises an evaluation report generation module which is used for summarizing the bypass analysis and detection results and generating a safety evaluation report so as to facilitate visual analysis of a user.
Further, the object aimed by the multi-channel bypass signal alignment is a cryptographic module embedded with a hidden protection countermeasure, and the object aimed by the second-order attack preprocessing is a cryptographic module embedded with a first-order mask protection countermeasure.
Further, the alignment method for alignment of the multi-channel bypass signal includes a global alignment method, specifically an alignment method based on Needleman-Wunsch, abbreviated as nw_la, and the specific process of the method includes:
1) Carrying out data standardization processing on the bypass signal and carrying out non-uniform quantization to obtain a corresponding symbol sequence;
2) The most similar part of the symbol sequence set, namely the longest common subsequence, is obtained by global alignment with different scoring matrices and rules using the Needleman-Wunsch algorithm, i.e. the optimization matching algorithm, and the global sequence comparison method, thereby aligning the bypass signal.
Further, the alignment method for alignment of the multi-channel bypass signal further includes a local alignment method, specifically, a Smith-Waterman-based alignment method, abbreviated as sw_la, which is similar to the nw_la, except that: after converting the bypass signal data into symbol sequences, the sw_la uses the Smith-Waterman algorithm to obtain the longest common subsequence of the symbol sequence set through local alignment, so as to align the bypass signal.
Further, the multi-channel bypass signal fusion degree judgment is used for judging whether the bypass signals of some channels contain useful secret information or not before safety analysis and detection are carried out; specifically, a fusion standard based on chi-square test is simply called CQM for judgment, when the CQM value calculated by using the multi-channel bypass signal is higher than a preset threshold value, the multi-channel bypass signal fusion can be considered, the next fusion, attack or leakage detection link is entered, and otherwise, the bypass signals of other channels are selected again for recalculation;
the specific judging steps of the CQM are as follows:
let two single channel bypass signal sets be L respectively 1 And L 2 Obviously L 1 And L 2 There should be an intersection part F between which depends on the common secret information, the more suitable the two single channel bypass signals are for combining, the larger the intersection F should be, L 1 And L 2 The greater the correlation between them; at this time L 1 And L 2 E obtained by removing the common portion F 1 And E is 2 Should be uncorrelated signals;
using chi-square independenceChi-square test L for checking R.times.C column-linked list 1 And L 2 Then there is a statistic:
wherein R and C independently represent L 1 And L 2 The number of the samples of the medium leakage signal is N, N is the total number of the samples, V ij And P ij Respectively the actual observation frequency and the theoretical frequency of the sample population; if statistics χ 2 If the value of (2) exceeds the threshold value at the corresponding confidence level, then L 1 And L 2 The correlation is large, and the fusion is suitable; otherwise, fusion is not appropriate.
Further, the second order attack preprocessing includes two preprocessing operations, namely regularization product and absolute value difference.
Further, the MCFA module is divided into 3 MCFA methods in terms of implementation, namely a data level, a feature level and a decision level;
data-level MCFA method: the simple leakage data fusion attack is called S_DAFA for short, which assumes that a certain cipher algorithm running on a cipher module performs one-time encryption or decryption operation, and obtains a single channel bypass signal of an operation in the realization of the cipher algorithm in between; the method comprises the steps of connecting a plurality of single-channel bypass signals in series, simultaneously subtracting the average value of each signal, and then enabling the dimension of the bypass signals to be consistent through vector normalization operation;
feature level MCFA method: the feature fusion attack based on singular value decomposition is called SVD_FEFA for short, which assumes that a certain bypass signal set of a cryptographic module is L, wherein the L contains m bypass signals, and each signal has n sample points; if each row vector of L represents a signal, and each column vector represents a leakage sample point of the same intermediate value at the same moment in the operation of the cryptographic algorithm, then from the perspective of signal time-frequency analysis, singular value decomposition SVD is carried out on L, and the left singular value vector of L contains frequency domain information of L, namely the change condition of all bypass signals in L at the same moment; the singular values contain energy information of L, namely the magnitude of the signal; the right singular value vector contains time domain information of L, namely the change condition of a single signal in L at different moments; the SVD_FEFA is that firstly, left singular value vectors of a plurality of single-channel bypass signal sets are fused into a new leakage set, and then attack is carried out;
decision-level MCFA method: the decision fusion attack based on maximum likelihood is called ML_DEFA for short, which assumes that in the operation of a cryptographic algorithm, M groups of different bypass leakage signal sets are respectively acquired from M channels, and l is used 1 ,l 2 …, lM represent these leakage signal sets; let L M ＝{l 1 ,l 2 ,…,l M -and represents a set of possible candidate values for a single subkey as k= { K 1 ,k 2 ,…,k N Symbol P (·) represents a probability distribution function of a random variable; leakage L at known M channels based on conditional independence, bayesian theorem, and sub-key ith candidate M The posterior probability under the condition can be obtained as follows:
wherein P (L) and P (L) j ) Probability distribution functions, k, representing the bypass signal set and the j-th set of leakage signals, respectively i And l j Representing sets K and L, respectively M Since P (L), P (lj) is independent of the distribution of subkey guesses, and subkey guesses are generally considered to be uniformly distributed, the final subkey guesses are:
and after the subkey fraction obtained by each single-channel bypass attack is converted into posterior probability by using the Bayes theorem, multiplying the posterior probability by the ML_DEFA to obtain the maximum value as the final score of the subkey candidate value, and outputting the subkey corresponding to the maximum score as the guessing key.
Further, the LD module performs leak detection by a leak detection method cs_ld based on chi-square test;
the cs_ld method is similar to the CQM principle, based on chi-square test of the rxc list, with the statistics unchanged, except that: the rows and columns of the column-linked list become target intermediate values realized by the cryptographic algorithm, and the multi-channel bypass fusion leaks; or after each single-channel bypass signal set is detected, a weighted Bayesian inference is utilized to infer and set a threshold value to integrate single-channel leakage detection results, and then a final detection result is obtained.
Further, the specific implementation process of the key enumeration attack and key sequence evaluation module comprises the following steps:
dividing the master key into a plurality of parts for side channel attack, obtaining the sequence and the score of all candidate values of each sub key, and enumerating the candidate values of the master key according to the sequence and the score so as to recover the complete master key;
the key enumeration algorithm of the module combines the Riemann integral of the ranking position of the candidate values of the sub-keys along with the change of the number of leakage signals, firstly weights the two to form a new measurement index, secondly randomly selects a certain number of leakage signals, calculates the ranking of all candidate values of each sub-key for a plurality of times, observes and uses an index to measure the change of the ranking position in each candidate value of the sub-key, then reorders the ranking positions of all correct sub-keys according to the index value, and further advances the ranking position of the correct main key.
Compared with the prior art, the invention has the remarkable advantages that: in view of the actual demand, a plurality of evaluation links such as signal preprocessing, attack, leakage detection, report generation and the like are covered, a set of more complete multi-channel bypass analysis and detection tool is innovatively developed, single-channel bypass analysis and detection is compatible (only one bypass leakage data is needed to be imported), and the situation which is more likely to be encountered in practice is considered particularly, namely, when the leakage quantity is too small to meet the key recovery requirement, how to evaluate the security level of the cryptographic module by utilizing key enumeration attack and key sequencing evaluation is suggested, and the key or algorithm implementation should be replaced when the leakage quantity is too large so as to ensure security.
The invention is described in further detail below with reference to the accompanying drawings.
Drawings
FIG. 1 is a diagram of a multi-channel bypass signal security analysis and detection system.
Fig. 2 is a schematic diagram of a signal preprocessing module.
Fig. 3 is a schematic diagram of a multi-channel fusion attack module.
Fig. 4 is a schematic diagram of a multi-channel leak detection module.
Fig. 5 is a schematic diagram of a key enumeration attack and key ordering evaluation module.
Fig. 6 is a schematic diagram of two single channel bypass signal sets with the common portion removed to obtain uncorrelated signals.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
The invention adopts a modularized design idea, and has 6 modules as shown in figure 1:
1. and a file processing module. The main function is to process bypass signal data, read the collected electromagnetic file and plaintext file into matrix, and facilitate the use of the following functions.
2. And the signal preprocessing module. The main functions are to preprocess the bypass signal, including multi-channel bypass signal alignment, multi-channel bypass signal fusion degree judgment, and second order attack preprocessing. Wherein:
1) The object of the multi-channel bypass signal alignment method is to take a cryptographic module embedded with hidden protection measures, and the cryptographic module comprises a global alignment method (alignment method based on Needleman-Wunsch, abbreviated as nw_la) and a local alignment method (alignment method based on Smith-Waterman, abbreviated as sw_la).
The nw_la includes the following steps: firstly, the bypass signal is subjected to data standardization processing and non-uniform quantization to obtain a corresponding symbol sequence. The bypass signal is then aligned by global alignment using the Needleman-Wunsch algorithm (optimized matching algorithm and overall sequence comparison) to obtain the most similar part of the set of symbol sequences, i.e., the longest common subsequence, with different scoring matrices and scoring rules. The step of sw_la is similar to nw_la except that sw_la, after converting the bypass signal data into symbol sequences, uses Smith-Waterman algorithm to obtain the longest common subsequence of the set of symbol sequences by local alignment, thereby aligning the bypass signal.
2) The multi-channel bypass signal fusion degree judgment is mainly used for judging whether bypass signals of certain channels contain useful secret information before safety analysis and detection, and is a precondition of signal fusion, attack and leakage detection. The system uses a fusion standard (CQM for short) based on chi-square test, when the CQM value calculated by using the multi-channel bypass signal is higher than a certain threshold value, the multi-channel bypass signal fusion can be considered to be carried out, the next fusion, attack or leakage detection link is carried out, and otherwise, the bypass signals of other channels are re-selected for re-calculation.
The steps of the CQM are as follows: first, assume that two single-channel bypass signal sets are L respectively 1 And L 2 . Obviously L 1 And L 2 There should be an intersection part F between which depends on the common secret information, the more suitable the two single channel bypass signals are for combining, the larger the intersection F should be, L 1 And L 2 The greater the correlation between them. At this time, L 1 And L 2 E obtained by removing the common portion F 1 And E is 2 Should be uncorrelated signals (such as electronic noise of different channels, etc., as in fig. 6). Then, using chi-square independence test (i.e., chi-square test of RxC column-Union) L 1 And L 2 Then there is a statistic:
wherein R and C independently represent L 1 And L 2 The number of samples of the medium leakage signal, N is the total number of samples,V ij and P ij Respectively the actual observation frequency and the theoretical frequency of the sample population; if statistics χ 2 If the value of (2) exceeds the threshold value at the corresponding confidence level, then L 1 And L 2 The correlation is large, and the fusion is suitable; otherwise, fusion is not appropriate.
3) The object aimed at by the second-order attack preprocessing is a cryptographic module embedded with a first-order mask protection countermeasure, and is a link before the attack, and the second-order attack preprocessing comprises two classical preprocessing operations, namely regularization product and absolute value difference.
3. A Multi-channel fusion attack (Multi-Channel Fusion Attacks, MCFA) module, which is compatible with single-channel attacks and second-order attacks. The multi-channel fusion attack evaluates the security level of the cryptographic module by recovering the key realized by the cryptographic algorithm, and is divided into 3 MCFA methods of data level, feature level and decision level from the realization mode.
Wherein, data level MCFA: a simple leakage data fusion attack (S_DAFA for short) is disclosed, which assumes that a certain cipher algorithm running on a cipher module performs one encryption or decryption operation, and obtains a single channel bypass signal of an operation in the implementation of the cipher algorithm. The bypass signals of a plurality of single channels are connected in series, and the number of the obtained fusion leakage signals is multiplied by a plurality of times compared with the number of the leakage signals of the single channels. On the basis, the attack can improve the side information leakage utilization rate and increase the correlation coefficient corresponding to the correct key, thereby increasing the bypass analysis efficiency. However, direct serial connection can reduce the signal-to-noise ratio of the fusion signal. The S_DAFA algorithm eliminates the difference in amplitude of the bypass signals of different channels by subtracting the mean value of each signal, and then keeps the bypass signal dimension consistent through vector normalization operation, thereby improving the signal-to-noise ratio and being superior to single-channel attack.
Feature level MCFA: based on singular value decomposition feature fusion attack (SVD_FEFA for short), a certain bypass signal set of a cryptographic module is assumed to be L, and m bypass signals are contained in L, and each signal has n sample points. If each row vector of L represents a signal, and each column vector represents a leakage sample point of the same intermediate value at the same moment in the operation of the cryptographic algorithm, singular Value Decomposition (SVD) is performed on L from the perspective of signal time-frequency analysis, and the left singular value vector of L contains frequency domain information of L, namely the change condition of all bypass signals in L at the same moment; the singular values contain energy information of L, namely the magnitude of the signal; the right singular value vector contains the time domain information of L, that is, the variation condition of a single signal in L at different moments. Since most bypass attacks use statistical means to reveal secret information, it is the change information of the target intermediate value in all signals at the same time that is utilized, so the left singular vector of L can be regarded as the main leakage feature of L and can be used to characterize L. The svd_fefa is to first fuse the left singular value vectors of the multiple single-channel bypass signal sets into a new leakage set, and then attack the leakage set.
Decision level MCFA: a decision fusion attack (ML_DEFA) based on maximum likelihood, which assumes that in the operation of a cryptographic algorithm, M groups of different bypass leakage signal sets are respectively acquired from M channels, and l is used 1 ,l 2 …, lM represent these leakage signal sets; let L M ＝{l 1 ,l 2 ,…,l M -and represents a set of possible candidate values for a single subkey as k= { K 1 ,k 2 ,…,k N Symbol P (·) represents a probability distribution function of a random variable; leakage L at known M channels based on conditional independence, bayesian theorem, and sub-key ith candidate M The posterior probability under the condition can be obtained as follows:
because P (L), P (lj) is independent of the distribution of subkey guesses, and subkey guesses are generally considered to be uniformly distributed, the final subkey guesses are:
and after the subkey fraction obtained by each single-channel bypass attack is converted into posterior probability by using the Bayes theorem, multiplying the posterior probability by the ML_DEFA to obtain the maximum value as the final score of the subkey candidate value, and outputting the subkey corresponding to the maximum score as the guessing key. The method combines the posterior probability of the subkey obtained by each single-channel attack, and further obtains a result superior to the single-channel attack.
4. A multi-channel leak detection (Leakage Detection, LD for short) module. The module is used for screening out leakage characteristic points containing secret information in the bypass signal, eliminating interference of non-characteristic points, reducing data space to be processed and improving evaluation efficiency. The autonomous design of the module provides 1 leak detection scheme: a leakage detection method (cs_ld) based on chi-square test.
The scheme is similar to the CQM principle, based on chi-square test of an RxC column list, the statistics are unchanged, but the rows and columns of the column list become target intermediate values realized by a cryptographic algorithm, and multi-channel bypass fusion leakage is realized; or after each single-channel bypass signal set is detected, a weighted Bayesian inference is utilized to infer and set a threshold value to integrate single-channel leakage detection results, and then a final detection result is obtained.
5. And a key enumeration attack module. Considering that in practice an adversary is more likely to encounter this situation: that is, there is insufficient environmental conditions to obtain a sufficient amount of leakage of a certain cryptographic module to complete a successful key recovery attack, at this time, when the bypass signal data amount is very small, the system provides a key enumeration attack method to estimate the security level of the cryptographic module, and a feasibility suggestion can be given when the cryptographic module changes keys or algorithms.
The flow of the module can be briefly described as follows: after the main key is divided into a plurality of parts to carry out side channel attack, the ordering and the score of all candidate values of each sub key are obtained, and then the candidate values of the main key are enumerated according to the ordering and the score, so that the complete main key is expected to be restored. The key enumeration algorithm of the module combines the Riemann integral of the ranking position of the candidate values of the sub-keys along with the change of the number of leakage signals, firstly weights the two to form a new measurement index, secondly randomly selects a certain number of leakage signals, calculates the ranking of all candidate values of each sub-key for a plurality of times, observes and uses an index to measure the change of the ranking position in each candidate value of the sub-key, then reorders the ranking positions of all correct sub-keys according to the index value, further advances the ranking position of the correct main key, and fundamentally reduces the enumeration times after key recovery attack.
6. And an evaluation report generation module. The module gathers the bypass analysis and detection results, generates a security evaluation report, and gives the security level of the cryptographic module against bypass attack and suggestions about key and cryptographic algorithm replacement or protective measures aiming at leakage characteristics.
The finished product of the multi-channel bypass signal security analysis and detection system exists in a software form, and the generated 'exe' file can be installed on a common personal computer (requiring an operating system to be win7 or win 10) to run, so as to support bypass security level evaluation for mainstream cryptographic algorithms such as DES, 3DES, AES, RC, SM4 and RSA, TEA, RSA. Before the security evaluation, the system needs to import the multichannel bypass leakage signal file and the explicit ciphertext information of the pre-acquired password module, then respectively realize signal preprocessing, bypass attack, leakage detection, key enumeration attack, key sequencing evaluation and the like according to the user requirements, and generate an evaluation report.
The overall operation flow of the system is described herein with reference to the accompanying drawings:
firstly, importing a file in a format of csv or bin in a system, wherein the files in the two formats are multi-channel bypass signals leaked when a cryptographic module operates; the system is then selected to preprocess the bypass signal by a preprocessing module (fig. 2), wherein the preprocessing module has 3 menus for selection, namely multi-channel bypass signal alignment, multi-channel bypass signal fusion degree judgment and second-order attack preprocessing.
The user selects according to the knowledge of the cryptographic module, if the cryptographic module is embedded with privacy protection countermeasures, the multi-channel signal alignment is selected: under the function, 1 global alignment method (alignment based on Needleman-Wunsch algorithm, abbreviated as nw_la) and local alignment method (alignment based on Smith-Waterman algorithm, abbreviated as sw_la) are designed, and the above alignment methods are compatible with single/multi-channel bypass signal alignment, and a user selects a corresponding algorithm according to the signal type to perform alignment. After the alignment operation is completed, the system interface popup window prompts the alignment operation to complete-! ". If the first-order mask protection countermeasure embedded cryptographic module is adopted, selecting second-order attack preprocessing: this is the link before the attack, and includes two classical preprocessing operations, namely regularization product and absolute value difference.
And then the user judges the fusion degree of the multi-channel bypass signals, the function depends on 1 measurement standard (CQM for short based on the fusion standard of chi-square test), if the CQM value calculated by using the multi-channel bypass signals is higher than a certain preset threshold value, the multi-channel bypass signals can be considered to be fused, the next fusion, attack or leakage detection link is entered, and otherwise, the bypass signals of other channels should be reselected for recalculation.
After the signal preprocessing is completed, a multi-channel fusion attack (MCFA) module (as shown in fig. 3) is selected, and the security level of the cryptographic module is evaluated by recovering the key realized by the cryptographic algorithm, so that the multi-channel fusion attack (MCFA) module is compatible with single-channel attack and second-order attack. The attack method can be divided into three major classes of data level, feature level and decision level from the security level, and each class comprises 1 attack algorithm, and 3 algorithms in total are all designed independently. The data level fusion attack method comprises the following steps: a simple leaky data fusion attack (s_dafa for short); the feature level fusion attack method comprises the following steps: feature fusion attack (SVD_FEFA for short) based on singular value decomposition; the decision-level fusion attack method comprises the following steps: decision fusion attack (ml_defa for short) based on weighted bayesian inference. The algorithm is provided for a user to select in a form of a drop-down menu on a system interface, if the attack is successful, a popup window prompts ' successful attack ', otherwise prompts ' failed attack, and the password security-! By considering that the fusion attack method can not reach the optimal effect, the method utilizes the negative correlation analysis technology to respectively obtain the linear optimal fusion result for the multi-channel bypass signal data, the characteristics and the attack result, thereby providing the optimal MCFA method of the data level, the characteristic level and the decision level under the constraint of the linear condition, and being capable of evaluating the security level of the bypass attack resistance of the password module under the worst leakage scene.
A multi-channel leak detection (Leakage Detection, LD for short) module (fig. 4). The module is used for screening out leakage characteristic points containing secret information in the bypass signal, eliminating interference of non-characteristic points, reducing data space to be processed and improving evaluation efficiency. The co-autonomous design of the module provides 1 leak detection scheme: a leakage detection method (cs_ld) based on chi-square test.
In particular, if the user does not have sufficient environmental conditions to obtain a sufficient leakage amount of a certain cryptographic module to complete a successful key recovery attack, and when the bypass signal data amount is very small, a multi-channel fusion attack module cannot be selected, and for this purpose, the system designs a key enumeration attack and key ordering evaluation module (such as fig. 5) to estimate the security level of the cryptographic module by using the key enumeration attack and key ordering evaluation method, and may give a feasibility suggestion when the cryptographic module changes keys or algorithms.
Finally, after all the modules are completed, an evaluation report generation module may be performed. The module gathers the results of the multi-channel fusion attack module, the multi-channel leakage detection module and the key enumeration attack and key ordering evaluation module, generates a security evaluation report, and reports security suggestions containing attack results, leakage feature points of secret information and a password module.
The multichannel bypass signal security analysis and detection system is developed by using Java programming language, eclipse2019, jdk1.7.0 environment, and 7 Java classes are designed, which are respectively:
(1) an attach_window class inherits from the JFrame class, in which a JPanel is created, and a plurality of button components, implementing the design of the system interface.
(2) File_choose is a File selection class which defines a function for selecting files, and the files are read in a computer.
(3) Prehandle is a signal preprocessing class which realizes the algorithm designed by the preprocessing module.
(4) The attach_method is an Attack class which defines 6 Attack methods of the multi-channel fusion Attack module.
(5) Leak detection is a Leak detection class that implements 3 Leak detection schemes for a Leak detection module.
(6) Print is a display class that defines a function of a display result for displaying execution results of other modules.
(7) Data is a Data class, which contains Data used in the whole project and is a class specially used for storing general Data.
In summary, the invention can integrate and utilize the physical information of a plurality of or a plurality of channels leaked when the cryptographic module runs, and complete the analysis and evaluation of the capability of resisting the bypass attack, the leakage characteristic and the security level of the cryptographic module by means of multichannel bypass signal preprocessing, multichannel fusion attack, multichannel leakage detection, key enumeration attack, key sequencing evaluation and the like. The invention can pre-examine and improve the actual password module product before the product is sent to a professional security assessment mechanism or a laboratory, reduce the assessment period and the test cost of the product, and accelerate the conversion of the product into a practical progress. Compared with the threat that the bypass analysis detection system in the current market only can utilize single-channel information leakage and does not evaluate key enumeration attack, the method can provide more comprehensive depth safety evaluation conclusion and suggestion.
The foregoing has outlined and described the basic principles, features, and advantages of the present invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made without departing from the spirit and scope of the invention, which is defined in the appended claims.
Claims (10)
1. A multi-channel bypass signal security analysis and detection system, the system comprising:
the file processing module is used for reading the multi-channel bypass signal and the plaintext and ciphertext information leaked during the operation of the password module and processing the multi-channel bypass signal and the plaintext and ciphertext information to generate a data matrix;
the signal preprocessing module is used for preprocessing bypass signals, including multi-channel bypass signal alignment, multi-channel bypass signal fusion degree judgment and second-order attack preprocessing;
the multi-channel fusion attack module is an MCFA module and is used for evaluating the security level of the cryptographic module by recovering the key realized by the cryptographic algorithm;
the multi-channel leakage detection module is an LD module and is used for screening leakage characteristic points containing secret information in the bypass signal and eliminating non-characteristic point interference;
the key enumeration attack and key ordering evaluation module is used for estimating the security level of the cryptographic module and giving feasibility suggestions for when the cryptographic module changes keys or algorithms.
2. The system for multi-channel bypass signal security analysis and detection of claim 1, further comprising an assessment report generation module for aggregating the results of the bypass analysis and detection to generate a security assessment report for visual analysis by a user.
3. The system of claim 1, wherein the object to which the multi-channel bypass signal is aligned is a cryptographic module embedded with a hidden security countermeasure, and the object to which the second-order attack pre-processing is directed is a cryptographic module embedded with a first-order masking security countermeasure.
4. The system according to claim 1, wherein the alignment method of the multi-channel bypass signal alignment comprises a global alignment method, in particular an alignment method based on Needleman-Wunsch, abbreviated nw_la, comprising the steps of:
1) Carrying out data standardization processing on the bypass signal and carrying out non-uniform quantization to obtain a corresponding symbol sequence;
2) The most similar part of the symbol sequence set, namely the longest common subsequence, is obtained by global alignment with different scoring matrices and rules using the Needleman-Wunsch algorithm, i.e. the optimization matching algorithm, and the global sequence comparison method, thereby aligning the bypass signal.
5. The system according to claim 4, wherein the alignment method of the multi-channel bypass signal alignment further comprises a local alignment method, specifically a Smith-Waterman-based alignment method, abbreviated as sw_la, which is: after converting the bypass signal data into symbol sequences, the sw_la uses the Smith-Waterman algorithm to obtain the longest common subsequence of the symbol sequence set through local alignment, so as to align the bypass signal.
6. The system of claim 1, wherein the multi-channel bypass signal fusion level determination is used to determine whether the bypass signals of certain channels contain useful secret information before performing the security analysis and detection; specifically, a fusion standard based on chi-square test is simply called CQM for judging, when the CQM value calculated by using the multi-channel bypass signal is higher than a preset threshold value, multi-channel bypass signal fusion is carried out, the next fusion, attack or leakage detection link is entered, and otherwise, the bypass signals of other channels are selected again for recalculation;
the specific judging steps of the CQM are as follows:
let two single channel bypass signal sets be L respectively 1 And L 2 ；
Chi-square test L using chi-square independence test, i.e. RXC column-Union 1 And L 2 Then there is a statistic:
wherein R and C independently represent L 1 And L 2 The number of the samples of the medium leakage signal is N, N is the total number of the samples, V ij And P ij Respectively the actual observation frequency and the theoretical frequency of the sample population; if statistics χ 2 The value of (2) exceeds the corresponding confidence levelThreshold value, L 1 And L 2 The correlation is large, and the fusion is suitable; otherwise, fusion is not appropriate.
7. The multi-channel bypass signal security analysis and detection system of claim 1 wherein the second order attack pre-processing comprises two pre-processing operations, regularized product and absolute value difference, respectively.
8. The multi-channel bypass signal security analysis and detection system of claim 1 wherein the MCFA module is divided in implementation into 3 MCFA methods, data level, feature level and decision level;
data-level MCFA method: the simple leakage data fusion attack is called S_DAFA for short, which assumes that a certain cipher algorithm running on a cipher module performs one-time encryption or decryption operation, and obtains a single channel bypass signal of an operation in the realization of the cipher algorithm in between; the method comprises the steps of connecting a plurality of single-channel bypass signals in series, simultaneously subtracting the average value of each signal, and then enabling the dimension of the bypass signals to be consistent through vector normalization operation;
feature level MCFA method: the feature fusion attack based on singular value decomposition is called SVD_FEFA for short, which assumes that a certain bypass signal set of a cryptographic module is L, wherein the L contains m bypass signals, and each signal has n sample points; if each row vector of L represents a signal, and each column vector represents a leakage sample point of the same intermediate value at the same moment in the operation of the cryptographic algorithm, then from the perspective of signal time-frequency analysis, singular value decomposition SVD is carried out on L, and the left singular value vector of L contains frequency domain information of L, namely the change condition of all bypass signals in L at the same moment; the singular values contain energy information of L, namely the magnitude of the signal; the right singular value vector contains time domain information of L, namely the change condition of a single signal in L at different moments; the SVD_FEFA is that firstly, left singular value vectors of a plurality of single-channel bypass signal sets are fused into a new leakage set, and then attack is carried out;
decision-level MCFA method: decision based on maximum likelihoodThe fusion attack is called ML_DEFA, which assumes that in the operation of a cryptographic algorithm, M groups of different bypass leakage signal sets are respectively acquired from M channels, and l is used 1 ,l 2 ,…,l M Representing these leakage signal sets; let L M ＝{l 1 ,l 2 ,…,l M -and represents a set of possible candidate values for a single subkey as k= { K 1 ,k 2 ,…,k N Symbol P (·) represents a probability distribution function of a random variable; leakage L at known M channels based on conditional independence, bayesian theorem, and sub-key ith candidate M The posterior probability under the condition can be obtained as follows:
wherein P (L) and P (L) j ) Probability distribution functions, k, representing the bypass signal set and the j-th set of leakage signals, respectively i And l j Representing sets K and L, respectively M Due to P (L), P (L) j ) Independent of the distribution of sub-key guesses, and the sub-key guesses are generally considered to be uniformly distributed, the last sub-key guesses k guess The method comprises the following steps:
and after the subkey fraction obtained by each single-channel bypass attack is converted into posterior probability by using the Bayes theorem, multiplying the posterior probability by the ML_DEFA to obtain the maximum value as the final score of the subkey candidate value, and outputting the subkey corresponding to the maximum score as the guessing key.
9. The multi-channel bypass signal security analysis and detection system according to claim 1 or 6, wherein the LD module performs leak detection by a leak detection method cs_ld based on chi-square test;
the CS_LD method is based on chi-square test of an RxC list, and the statistic is unchanged: the rows and columns of the column-linked list become target intermediate values realized by the cryptographic algorithm, and the multi-channel bypass fusion leaks; or after each single-channel bypass signal set is detected, a weighted Bayesian inference is utilized to infer and set a threshold value to integrate single-channel leakage detection results, and then a final detection result is obtained.
10. The system for securely analyzing and detecting a multi-channel bypass signal according to claim 1, wherein the specific implementation procedure of the key enumeration attack and key ordering evaluation module comprises:
dividing the master key into a plurality of parts for side channel attack, obtaining the sequence and the score of all candidate values of each sub key, and enumerating the candidate values of the master key according to the sequence and the score so as to recover the complete master key;
the key enumeration algorithm of the module combines the Riemann integral of the ranking position of the candidate values of the sub-keys along with the change of the number of leakage signals, firstly weights the two to form a new measurement index, secondly randomly selects a certain number of leakage signals, calculates the ranking of all candidate values of each sub-key for a plurality of times, observes and uses an index to measure the change of the ranking position in each candidate value of the sub-key, then reorders the ranking positions of all correct sub-keys according to the index value, and further advances the ranking position of the correct main key.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111673193.3A CN114531232B (en) | 2021-12-31 | 2021-12-31 | Multichannel bypass signal safety analysis and detection system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111673193.3A CN114531232B (en) | 2021-12-31 | 2021-12-31 | Multichannel bypass signal safety analysis and detection system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114531232A CN114531232A (en) | 2022-05-24 |
| CN114531232B true CN114531232B (en) | 2024-03-19 |
Family
ID=81621450
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111673193.3A Active CN114531232B (en) | 2021-12-31 | 2021-12-31 | Multichannel bypass signal safety analysis and detection system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114531232B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104811295A (en) * | 2015-05-05 | 2015-07-29 | 国家密码管理局商用密码检测中心 | Side channel energy analysis method for ZUC cryptographic algorithm with mask protection |
| CN112787971A (en) * | 2019-11-01 | 2021-05-11 | 国民技术股份有限公司 | Construction method of side channel attack model, password attack equipment and computer storage medium |
-
2021
- 2021-12-31 CN CN202111673193.3A patent/CN114531232B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104811295A (en) * | 2015-05-05 | 2015-07-29 | 国家密码管理局商用密码检测中心 | Side channel energy analysis method for ZUC cryptographic algorithm with mask protection |
| CN112787971A (en) * | 2019-11-01 | 2021-05-11 | 国民技术股份有限公司 | Construction method of side channel attack model, password attack equipment and computer storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 基于侧信道分析的密码算法安全评估技术研究;傅山;《中国博士学位论文全文数据库 信息科技辑》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114531232A (en) | 2022-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lerman et al. | Power analysis attack: an approach based on machine learning | |
| Fronza et al. | Failure prediction based on log files using random indexing and support vector machines | |
| EP2814218B1 (en) | Detecting anomalies in work practice data by combining multiple domains of information | |
| Zou et al. | Regularized simultaneous model selection in multiple quantiles regression | |
| Lee et al. | Bayesian multi-SNP genetic association analysis: control of FDR and use of summary statistics | |
| CN106060008A (en) | Network invasion abnormity detection method | |
| KR101623493B1 (en) | Appropriate Countermeasure against Side Channel Analysis on Cryptogram Generating Process of Financial IC Cards | |
| Zhang et al. | Machine-learning-based side-channel leakage detection in electronic system-level synthesis | |
| Taouil et al. | Power side channel attacks: Where are we standing? | |
| CN114531232B (en) | Multichannel bypass signal safety analysis and detection system | |
| CN112615714B (en) | Side channel analysis method, device, equipment and storage medium | |
| Picek et al. | When theory meets practice: A framework for robust profiled side-channel analysis | |
| Cagli | Feature extraction for side-channel attacks | |
| He et al. | Image-Based Zero-Day Malware Detection in IoMT Devices: A Hybrid AI-Enabled Method | |
| Patel et al. | Random forest profiling attack on advanced encryption standard | |
| KR102308517B1 (en) | Correlation power analysis method and apparatus for block cipher | |
| Akter et al. | Exploring the vulnerabilities of machine learning and quantum machine learning to adversarial attacks using a malware dataset: a comparative analysis | |
| He et al. | ResNeXt+: Attention Mechanisms Based on ResNeXt for Malware Detection and Classification | |
| Vishwakarma et al. | Risk-Aware and Explainable Framework for Ensuring Guaranteed Coverage in Evolving Hardware Trojan Detection | |
| KR20200109677A (en) | An apparatus and method for detecting malicious codes using ai based machine running cross validation techniques | |
| Alghazali et al. | Using a hybrid algorithm with intrusion detection system based on hierarchical deep learning for smart meter communication network | |
| Forbes et al. | Metric Ensembles For Hallucination Detection | |
| Wu et al. | An end-to-end plaintext-based side-channel collision attack without trace segmentation | |
| Attarian et al. | Mixflow: Assessing mixnets anonymity with contrastive architectures and semantic network information | |
| Chunikhina et al. | The C-SHIFT algorithm for normalizing covariances |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |