A security protection method for industrial Internet big data

Abstract

The present invention relates to the field of data encryption technology, and in particular to a security protection method for industrial Internet big data, the method comprising: obtaining each data sequence in an industrial park and encrypting it to obtain a key byte matrix; constructing a matching sequence based on the difference between each key byte sequence in the key byte matrix and a randomly generated sequence; recording any two key bytes as the first and second key bytes, respectively, and obtaining a byte similarity index based on the alignment sequence and distance of the first and second key byte matching sequences and the difference between each matching sequence; using a Markov clustering algorithm to construct a first key byte matrix for all byte similarity indices; constructing an initial key byte matrix based on the byte similarity index between each key byte in any two key words in the first key byte matrix, and optimizing the AES encryption algorithm to achieve encryption of industrial production process data. The present invention improves the security of the AES encryption algorithm for industrial data.

Description

Technical Field

[0001] The present application relates to the field of data encryption technology, and specifically to a security protection method for industrial Internet big data. Background Art

[0002] Industrial parks typically deploy a large number of sensors to collect production data for monitoring and controlling the production process. This sensor data contains crucial information about industrial production, such as parameter adjustments. A data leak could result in significant losses. Furthermore, sensor data is often used to monitor production operations. Tampering with this data could cause equipment failures and even pose security risks. Therefore, it's essential to implement security protection for Industrial Internet big data.

[0003] The key byte matrix of the traditional AES encryption algorithm directly arranges the key bytes in order, with four adjacent bytes used as a key word. This key byte matrix has a high degree of correlation between adjacent bytes, resulting in high correlation between subsequent rounds of keys and poor data encryption effect. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides a security protection method for industrial Internet big data to solve the existing problems.

[0005] The security protection method of industrial Internet big data of the present invention adopts the following technical solutions:

[0006] An embodiment of the present invention provides a method for protecting industrial Internet big data security, the method comprising the following steps:

[0007] Obtaining each data sequence of each industrial area in the industrial park; encrypting all data sequences in the industrial park to obtain a key byte matrix;

[0008] Representing each key byte in the key byte matrix in binary form to obtain each key byte sequence; obtaining a byte XOR sequence based on the difference between the key byte sequence and the randomly generated sequence; constructing a matching sequence based on the byte XOR sequence; recording any two key bytes as the first and second key bytes, respectively, and recording the matching sequences of the first and second key bytes as the first and second matching sequences; obtaining an alignment sequence and a distance between the first and second matching sequences; and obtaining a byte similarity index between the first and second key bytes based on the distance and the difference between the alignment sequence and each matching sequence.

[0009] A Markov clustering algorithm is used to obtain Markov clusters based on the byte similarity index between the key bytes; a first key byte matrix is ​​constructed based on the elements in each Markov cluster; and for any two key words in the first key byte matrix, a word similarity correlation index of the two key words is obtained based on the byte similarity index between the key bytes in the two key words;

[0010] An initial key byte matrix is ​​constructed according to the word similarity association index between the first key byte matrix and each key word, and the industrial production process data is encrypted using the AES encryption algorithm combined with the initial key byte matrix.

[0011] Preferably, obtaining a byte XOR sequence based on the difference between the key byte sequence and the randomly generated sequence includes:

[0012] A random number generator is used to generate a random key byte sequence with the same length as the key byte sequence; an XOR operation is performed on the elements at the same position of the random key byte sequence and the key byte sequence to obtain a byte XOR sequence.

[0013] Preferably, constructing a matching sequence based on a byte XOR sequence includes:

[0014] The subscript of each element in the byte XOR sequence is used as the horizontal coordinate of the key byte matching graph, and the vertical coordinate of the key byte matching graph is:

[0015] When the element value is 0, the ordinate of the current element is the ordinate of the previous element plus 1; when the element value is 1, the ordinate of the current element is the ordinate of the previous element minus 1;

[0016] The key bytes are matched with each element value in the graph to form a matching sequence.

[0017] Preferably, obtaining the alignment sequence and distance of the first and second matching sequences includes:

[0018] The SBD distance is obtained by using the SBD measurement method for the two matching sequences, which is recorded as the distance between the first and second matching sequences;

[0019] An alignment sequence of the second matching sequence relative to the first matching sequence is obtained according to the SBD metric method.

[0020] Preferably, obtaining the byte similarity index of the first and second key bytes according to the distance, the difference between the aligned sequence and each matching sequence includes:

[0021] Obtaining byte similarity difference coefficients between the first and second matching sequences based on the distance, the alignment sequence, and the first and second matching sequences;

[0022] The byte similarity difference coefficient between the first and second matching sequences is used as a benchmark coefficient;

[0023] For each matching sequence of key bytes, calculate the difference between the byte similarity difference coefficient between the matching sequence and the first matching sequence and the reference coefficient;

[0024] The normalized value of the ratio of the sum of the difference values ​​of all matching sequences to the reference coefficient is used as the byte similarity index of the first and second key bytes.

[0025] Preferably, obtaining the byte similarity difference coefficient between the first and second matching sequences according to the distance, the alignment sequence, and the first matching sequence includes:

[0026] Calculating the absolute value of the difference between each element of the aligned sequence and the first matching sequence, and calculating the sum of the absolute values ​​of the differences of all elements;

[0027] The sum of the sum value and the distance is used as the byte similarity difference coefficient between the first and second matching sequences.

[0028] Preferably, the Markov clustering algorithm is used to obtain Markov clusters based on the byte similarity index between the key bytes, including:

[0029] Each key byte is regarded as a node of an undirected graph, and the byte similarity index between any two nodes is regarded as the weight of the connection between the two nodes;

[0030] The undirected graph is input into the Markov clustering algorithm, and Markov clusters are output. The number of the Markov clusters is set to 4.

[0031] Preferably, the step of constructing a first key byte matrix according to the elements in each Markov cluster comprises:

[0032] Sort the Markov clusters from large to small according to the number of elements in the cluster; Sort the elements in each Markov cluster from small to large according to the key byte subscript;

[0033] Sequentially fill each row of the first key byte matrix with each element in each sorted Markov cluster, and sequentially push the redundant key bytes in each row into the remaining key byte stack;

[0034] After all Markov clusters are filled into the first key byte matrix in order, the elements in the remaining key byte stack are popped out and filled into the free area of ​​the first key byte matrix in order from left to right and from top to bottom.

[0035] Preferably, obtaining the word similarity correlation index of the two key words according to the byte similarity index between the key bytes in the two key words comprises:

[0036] For any two key bytes in the two key words, obtain the absolute value of the difference between the row numbers corresponding to the two key bytes, and use the opposite of the absolute value of the difference as the exponent of an exponential function with a natural constant as the base;

[0037] Calculating the product of the byte similarity index of the two key bytes and the calculation result of the exponential function;

[0038] The sum of the products of all arbitrary two key bytes in the two key words is taken as the word similarity association index of the two key words.

[0039] Preferably, the step of constructing the initial key byte matrix according to the word similarity correlation index between the first key byte matrix and each key word comprises:

[0040] Using the first key word in the first key byte matrix as the first key word in the initial key byte matrix;

[0041] Selecting a key word having the smallest word similarity correlation index with the first key word from the remaining key words in the first key byte matrix as the second key word in the initial key byte matrix;

[0042] The filling sequence of the remaining key words in the initial key byte matrix is ​​filled using the same filling method as the second key word.

[0043] The present invention has at least the following beneficial effects:

[0044] The present invention uses the SBD measurement method and the Markov clustering algorithm based on the similarity between each key byte in the key byte matrix to comprehensively construct a word similarity correlation index to reflect the similarity between key words. The beneficial effect is to avoid the situation where the correlation between key bytes in the initial key byte matrix is ​​low but the correlation between key words is high, prevent the subsequent round key expansion from resulting in high similarity between each round key, and reduce the risk of encrypted data being cracked.

[0045] At the same time, based on the above construction of the initial key byte matrix, the beneficial effect is to avoid the problem of large correlation between key bytes and adjacent key bytes, key words and adjacent key words in the traditional AES encryption algorithm, improve the excellence of the AES encryption algorithm in obtaining keys, and thus improve the security of the AES encryption algorithm for industrial data. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0047] Figure 1 A flowchart of a security protection method for industrial Internet big data provided by the present invention;

[0048] Figure 2 is the key word in the key byte matrix;

[0049] Figure 3 is the key byte matching graph. DETAILED DESCRIPTION

[0050] To further illustrate the technical means and effectiveness of the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a method for protecting industrial Internet big data security, including its specific implementation, structure, features, and effectiveness. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0051] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0052] The following describes in detail a specific solution of a security protection method for industrial Internet big data provided by the present invention with reference to the accompanying drawings.

[0053] An embodiment of the present invention provides a method for protecting the security of industrial Internet big data.

[0054] Specifically, the following security protection method for industrial Internet big data is provided. Please refer to Figure 1 , the method comprises the following steps:

[0055] Step S001: collect industrial production data in the industrial park through relevant sensors and pre-process it to obtain a key byte matrix.

[0056] In an industrial park, different industrial areas may have different industrial production. Therefore, the entire industrial park is evenly divided into N industrial areas according to their locations. The same sensors are placed in each industrial area to collect production process data from each industrial area. For example, temperature and humidity sensors are used to obtain humidity and temperature data during the production process of the industrial area, sound sensors are used to obtain sound data during the production process of the industrial area, and air detectors are used to obtain dust concentration during the production process of the industrial area. The specific types of sensors used are adjusted by the implementers according to the industrial production situation.

[0057] Assume that there are m data categories, the interval between two consecutive data collections is t, and the collection time is T, where t takes an empirical value of 1 second and T takes an empirical value of 600. Due to the complexity of the industrial production environment, collected data may be missing and the dimensions of different data may be inconsistent. Therefore, this embodiment uses linear interpolation to first fill in the data and then performs Z-Score normalization on the filled data to remove the dimensionality effect.

[0058] Based on the above collection process, this embodiment collects the data sequence of each production data of each industrial area in the industrial park. The data sequence of the production data of all industrial areas in the industrial park is encrypted using the AES encryption algorithm to obtain a key byte matrix.

[0059] Step S002: determining an initial key byte matrix of the AES encryption algorithm based on the similarity between the key bytes in the key byte matrix.

[0060] For the traditional AES encryption algorithm, when the key length is 256 bits, the dimension of the key byte matrix is ​​4*8. The 4 bytes in the adjacent columns form a key word, and a total of 8 key words are obtained, which are recorded as R0, R1, R2, R3, R4, R5, R6, and R7, respectively. Figure 2 shown.

[0061] Due to the large correlation between industrial data, the key byte matrix constructed by the traditional AES encryption algorithm has a strong correlation between adjacent bytes. In addition, the AES encryption algorithm uses a key expansion algorithm to expand the initial key into multiple round keys for subsequent round encryption. This will cause the round keys obtained by subsequent AES encryption algorithm to have a high degree of similarity. Once an attacker obtains one of the round keys, he may obtain all the keys based on the strong similarity between the round keys, which greatly affects the security of industrial data.

[0062] Therefore, the purpose of this embodiment is to set the bytes with higher similarity to a farther distribution distance in the key byte matrix based on the similarity between the key bytes generated by the industrial data, thereby reducing the similarity between the key word and the round key and improving the security of the industrial data.

[0063] The key bytes in the AES encryption algorithm are all composed of binary digits. For each key byte in the key byte matrix, a corresponding binary key byte sequence is obtained. A random number generator is used to generate a random key byte sequence r′ with the same length as the above key byte sequence. The binary AES key byte sequences obtained based on industrial production sensor data often have certain similarities. An exclusive OR logic operation is performed on the elements in the same position in the key byte sequence and the random key byte sequence r′, and the exclusive OR logic operation result is recorded as the byte exclusive OR sequence of the two key bytes.

[0064] If the key byte sequence μ0 is [0,1,1,0,0,1,1] and the random key byte sequence μ′ is [0,1,1,1,0,1,1], the elements in the same position are subjected to the XOR logic operation, and the result is recorded as the byte XOR sequence μp0 = [0,0,0,1,0,0,0].

[0065] A key byte matching graph is constructed for the byte XOR sequence μp0. Specifically, the subscript of the data element is used as the horizontal coordinate, starting from the origin of the rectangular coordinate system, that is, the initial point coordinate is (0,0). When the data element in the byte XOR sequence μp0 is 0, the vertical coordinate of the key byte matching graph is the vertical coordinate of the previous data element +1. When the data element in the byte XOR sequence μp0 is 1, the vertical coordinate of the key byte matching graph is the vertical coordinate of the previous data element -1. The key byte matching graph is obtained by connecting the data elements in the byte XOR sequence μp0 in order, and the matching sequence can be obtained from the key byte matching graph.

[0066] Taking the above byte XOR sequence μp0=[0,0,0,1,0,0,0] as an example, the corresponding key byte matching diagram is as follows Figure 3 As shown, the matching sequence is μs0 = [1, 2, 3, 2, 3, 4, 5].

[0067] At this point, the matching sequence between each key byte sequence and the random key byte sequence is obtained. Taking the αth and βth key bytes in the AES encryption algorithm as an example, the matching sequence rs corresponding to the αth and βth key bytes can be obtained. α rs β Using the SBD (Shape-based Distance) metric, input two matching sequences rs α rs β, output the SBD distance between the two matching sequences and the matching sequence rs β Compared to rs α The alignment sequence rs′ β , where the SBD measurement method is a well-known technique and will not be described in detail in this embodiment. Based on the above analysis, the byte similarity index H between the αth and βth key bytes can be obtained: α,β :

[0068]

[0069] Among them, HD α,β is the byte similarity difference coefficient between the αth and βth matching sequences, SBD(rs α ,rs β ) is the matching sequence rs α rs β The SBD distance output by the SBD measurement method, rs′ β is the output of the SBD measurement method rs β Relative to rs α The aligned sequence, rs α,δ is the matching sequence rs α The value of the δth element in rs′ β,δ is the alignment sequence rs′ β The value of the δth element in , η is the length of the matching sequence.

[0070] When the shape similarity between the two matching sequences is lower, that is, SBD(rs α ,rs β ) is larger, indicating that the two matching sequences are more dissimilar in sequence shape, and the greater the difference in sequence shape, the greater the difference between the data elements at the same position in the matching sequence and the aligned sequence, that is, |rs α,δ -rs′ β,δ The larger the | is, the greater the difference in the numerical value of the matching sequence corresponding to the two key bytes is, and the byte similarity difference coefficient HD α,β The bigger.

[0071] H α,β is the byte similarity index between the αth and βth key bytes, HD α,l is the byte similarity metric distance between the αth and lth key bytes, k is the total number of all key bytes, and Norm[*] is the linear normalization function, where HD α,β is the base coefficient.

[0072] When the byte similarity difference coefficient between the α key byte and the β key byte is smaller than the byte similarity difference coefficient of other key bytes, that is, (HD α,l -HD α,β ) is larger, and HDα,β The smaller the value, the more similar the β key byte is to the α key byte than to the other bytes. α,β The bigger.

[0073] At this point, the byte similarity index between any two key bytes is obtained. Each key byte is treated as a node in an undirected graph, and the byte similarity index between each key byte is used as the weight of the connection between the nodes in the undirected graph. A Markov clustering algorithm is used, taking the obtained undirected graph of key bytes as input, and partitioning each key byte node into Markov clusters. The output is K Markov clusters. The Markov clustering algorithm is well known and will not be described in detail in this embodiment.

[0074] When the number of Markov clusters K=4, the number of key bytes in the Markov clusters is sorted from large to small, and the first-ranked Markov cluster is placed in the first row of the first key byte matrix in the order of the internal key byte subscripts from small to large. The remaining key bytes in the first-ranked cluster are stored in the remaining key byte stack in the order of the subscripts from small to large. Then, the second-ranked Markov cluster is placed in the second row of the first key byte matrix in the same way, and the remaining key bytes are also stored in the remaining key byte stack in sequence. The remaining clusters perform the same operation according to the above process, and the specific process is not repeated here.

[0075] Until all clusters have been placed in the first key byte matrix, finally all the remaining key bytes in the remaining key byte stack are placed in the free element area of ​​the first key byte matrix in the order from left to right and from top to bottom according to the stack structure (first in, last out).

[0076] At this point, the construction of the first key byte matrix is ​​completed, and the eight key words obtained are recorded as R'0, R'1, R'2, ..., R'7 respectively. The similarity between the key bytes in the obtained key words is low. In industrial data, the correlation similarity between the key words obtained by the AES encryption algorithm may be high. In order to prevent the high similarity between adjacent key words, which leads to the high similarity of each round key during the subsequent round key expansion, affecting the security of AES encryption of industrial data. Based on the above analysis, the word similarity correlation index B between each key word is obtained according to the byte similarity index. i,j :

[0077]

[0078] Among them, B i,j is the word similarity correlation index between the i-th keyword and the j-th keyword, is the byte similarity index between the xth key byte in the i-th key word and the yth key byte in the j-th key word, P x 、P y are the row numbers corresponding to the x and y key bytes in the first key byte matrix, respectively. Z is the number of key bytes in the first key byte matrix. In this embodiment, Z is set to 4.

[0079] In the first key byte matrix, when the distance between the key bytes of two key words is closer, that is, The larger the value is, the greater the adjacent weight of the two key bytes is, and the greater the correlation should be. At the same time, when the byte similarity index between the two key bytes is larger, that is, The larger the value, the more relevant the two key bytes are in terms of shape and value, and the higher the similarity between the two key words, that is, the word similarity correlation index B i,j The bigger.

[0080] The larger the word similarity correlation index between two key words, the higher the degree of similarity correlation between the two key words. The positions of the two key words in the key matrix should be farther apart, thereby reducing the correlation between the round keys when the industrial data is encrypted by AES and ensuring the security of industrial data encryption.

[0081] Based on the word similarity correlation index between the first key byte matrix and each key word obtained above, this embodiment also uses the first key word as the first key word in the initial key byte matrix, and selects the key word with the smallest word similarity correlation index with the first key word from the remaining key words as the second key word in the initial key byte matrix; the key word with the smallest word similarity correlation index with the second key word will be selected from the remaining key words as the third key word, and so on. The remaining key words are processed using the same operations as above to construct the initial key byte matrix.

[0082] Step S003: improving the AES encryption algorithm based on the obtained initial key byte matrix to achieve encryption of industrial production process data.

[0083] According to the above steps, the initial key byte matrix for encrypting the industrial production process data is obtained, and the initial key byte matrix is ​​used as the key byte matrix in the traditional AES encryption algorithm. The AES encryption algorithm is improved, and all industrial production process data are encrypted based on the improved AES encryption algorithm to obtain the corresponding ciphertext. The subsequent AES encryption process is a well-known technology, and the specific process will not be repeated here.

[0084] The embodiment of the present invention uses the SBD measurement method and the Markov clustering algorithm based on the similarity between each key byte in the key byte matrix to comprehensively construct a word similarity correlation index to reflect the similarity between key words. The beneficial effect is to avoid the situation where the correlation between key bytes in the initial key byte matrix is ​​low but the correlation between key words is high. This prevents the subsequent round key expansion from resulting in high similarity between each round key, thereby reducing the risk of encrypted data being compromised.

[0085] At the same time, based on the above construction of the initial key byte matrix, the beneficial effect is to avoid the problem of large correlation between key bytes and adjacent key bytes, key words and adjacent key words in the traditional AES encryption algorithm, improve the excellence of the AES encryption algorithm in obtaining keys, and thus improve the security of the AES encryption algorithm for industrial data.

[0086] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0087] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0088] The above-described embodiments are only used to illustrate the technical solutions of the present application, and not to limit them. Modifications to the technical solutions described in the aforementioned embodiments, or equivalent replacements of some of the technical features therein, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A security protection method for industrial Internet big data, characterized in that: The method comprises the following steps: Obtaining each data sequence of each industrial area in the industrial park; encrypting all data sequences in the industrial park to obtain a key byte matrix; Representing each key byte in the key byte matrix in binary format to obtain each key byte sequence; obtaining a byte XOR sequence based on the difference between the key byte sequence and the randomly generated sequence; constructing a matching sequence based on the byte XOR sequence; recording any two key bytes as the first and second key bytes, respectively, and recording the matching sequences of the first and second key bytes as the first and second matching sequences; obtaining an alignment sequence and a distance between the first and second matching sequences; and obtaining a byte similarity index between the first and second key bytes based on the distance and the difference between the alignment sequence and each matching sequence. A Markov clustering algorithm is used to obtain Markov clusters based on the byte similarity index between the key bytes; a first key byte matrix is ​​constructed based on the elements in each Markov cluster; and for any two key words in the first key byte matrix, a word similarity correlation index of the two key words is obtained based on the byte similarity index between the key bytes in the two key words; An initial key byte matrix is ​​constructed based on the word similarity correlation index between the first key byte matrix and each key word, and the industrial production process data is encrypted using the AES encryption algorithm combined with the initial key byte matrix; The step of constructing a matching sequence based on a byte XOR sequence includes: The subscript of each element in the byte XOR sequence is used as the horizontal coordinate of the key byte matching graph, and the vertical coordinate of the key byte matching graph is: When the element value is 0, the ordinate of the current element is the ordinate of the previous element plus 1; when the element value is 1, the ordinate of the current element is the ordinate of the previous element minus 1; The key byte matching graph is used to form a matching sequence of each element value; The obtaining of the alignment sequence and the distance between the first and second matching sequences includes: The SBD distance is obtained by using the SBD measurement method for the two matching sequences, which is recorded as the distance between the first and second matching sequences; An alignment sequence of the second matching sequence relative to the first matching sequence is obtained according to the SBD metric method.

2. The method for protecting industrial Internet big data security according to claim 1, characterized in that: The step of obtaining a byte XOR sequence based on the difference between the key byte sequence and the randomly generated sequence includes: A random number generator is used to generate a random key byte sequence with the same length as the key byte sequence; an XOR operation is performed on the elements at the same position of the random key byte sequence and the key byte sequence to obtain a byte XOR sequence.

3. The method for protecting the security of industrial Internet big data according to claim 1, wherein: The obtaining of the byte similarity index of the first and second key bytes according to the distance, the difference between the aligned sequence and each matching sequence includes: Obtaining byte similarity difference coefficients between the first and second matching sequences based on the distance, the alignment sequence, and the first and second matching sequences; The byte similarity difference coefficient between the first and second matching sequences is used as a benchmark coefficient; For each matching sequence of key bytes, calculate the difference between the byte similarity difference coefficient between the matching sequence and the first matching sequence and the reference coefficient; The normalized value of the ratio of the sum of the difference values ​​of all matching sequences to the reference coefficient is used as the byte similarity index of the first and second key bytes.

4. The method for protecting the security of industrial Internet big data according to claim 3, characterized in that: The obtaining of the byte similarity difference coefficients between the first and second matching sequences according to the distance, the alignment sequence, and the first matching sequence includes: Calculating the absolute value of the difference between each element of the aligned sequence and the first matching sequence, and calculating the sum of the absolute values ​​of the differences of all elements; The sum of the sum value and the distance is used as the byte similarity difference coefficient between the first and second matching sequences.

5. The method for protecting the security of industrial Internet big data according to claim 3, characterized in that: The Markov clustering algorithm is used to obtain Markov clusters based on the byte similarity index between the key bytes, including: Each key byte is regarded as a node of an undirected graph, and the byte similarity index between any two nodes is regarded as the weight of the connection between the two nodes; The undirected graph is input into the Markov clustering algorithm, and Markov clusters are output. The number of the Markov clusters is set to 4.

6. The method for protecting industrial Internet big data security according to claim 5, characterized in that: The step of constructing a first key byte matrix according to the elements in each Markov cluster comprises: Sort the Markov clusters from large to small according to the number of elements in the cluster; Sort the elements in each Markov cluster from small to large according to the key byte subscript; Fill each element in each sorted Markov cluster into each row of the first key byte matrix in order, and push the redundant key bytes in each row into the remaining key byte stack in order; After all Markov clusters are filled into the first key byte matrix in order, the elements in the remaining key byte stack are popped out and filled into the free area of ​​the first key byte matrix in order from left to right and from top to bottom.

7. The method for protecting industrial Internet big data security according to claim 1, characterized in that: The method of obtaining the word similarity correlation index of the two key words according to the byte similarity index between the key bytes in the two key words comprises: For any two key bytes in the two key words, obtain the absolute value of the difference between the row numbers corresponding to the two key bytes, and use the opposite of the absolute value of the difference as the exponent of an exponential function with a natural constant as the base; Calculating the product of the byte similarity index of the two key bytes and the calculation result of the exponential function; The sum of the products of all arbitrary two key bytes in the two key words is taken as the word similarity association index of the two key words.

8. The method for protecting the security of industrial Internet big data according to claim 7, characterized in that: The method of constructing an initial key byte matrix according to the word similarity association index between the first key byte matrix and each key word includes: Using the first key word in the first key byte matrix as the first key word in the initial key byte matrix; Selecting a key word having the smallest word similarity correlation index with the first key word from the remaining key words in the first key byte matrix as the second key word in the initial key byte matrix; The remaining key words in the initial key byte matrix are filled in the same filling method as the second key word.

Images