Data security storage method under 5G network security management

Abstract

The invention relates to the technical field of data security, in particular to a data security storage method under 5G network security management, comprising: collecting communication data in a 5G network, and determining to-be-encrypted data; clustering the communication data to obtain a data cluster, performing interpolation processing on the numerical sequence based on the length of standard data of the data cluster to obtain a communication feature matrix, and determining a template feature vector by using a dimension reduction technology; determining the regional difference degree of the maximum value in the neighborhood range based on the distribution condition of the elements on the two sides of the maximum value in the distance sequence of each row of elements; determining a key sequence based on the regional difference degree; determining an offset coefficient according to the secret key sequence; determining an encryption key of data in each row of elements based on the offset coefficient by utilizing a logistic transformation encryption technology; encryption of communication data is completed based on the encryption key, and ciphertexts of all the data are stored. According to the method, the security encryption effect of the communication data is improved through the encryption mode of the self-adaptive key.

Description

Technical Field

[0001] The present application relates to the field of data security technology, and specifically to a data security storage method under 5G network security management. Background Art

[0002] With the rapid development of network communication technology, the amount of communication data in different fields continues to grow, and the number of 5G network base stations is also gradually increasing, making the traffic of communication data in 5G networks larger and larger, and the information communication load of network bandwidth, receiving end and communication sending end gradually increasing. The high-speed connection and huge data transmission volume brought by 5G networks also cause the possibility of information leakage to gradually increase, and it is necessary to ensure data security through technologies such as identity authentication, data encryption and intrusion detection.

[0003] In the process of big data transmission in 5G networks, data encryption and privacy protection become crucial. By encrypting the data transmitted in 5G networks, sensitive information and private information can be ensured not to be stolen or tampered with during the transmission process. At present, the security protection of communication data is achieved by using encryption algorithms. However, when using encryption algorithms to protect data, the data to be encrypted is usually encrypted to the same degree or with the same key, ignoring the differences between the data, resulting in the security of the encrypted ciphertext being insufficient to resist malicious attacks, and still facing network security issues. Summary of the invention

[0004] The present application provides a data security storage method under 5G network security management to solve the existing problems.

[0005] An embodiment of the present application provides a data security storage method under 5G network security management, and the technical solution adopted is as follows:

[0006] Collect communication data in the 5G network and determine the data to be encrypted;

[0007] The communication data is clustered to obtain multiple data clusters, and the communication feature matrix is ​​obtained by interpolating the numerical sequence based on the length of the communication data corresponding to the data point in each data cluster, and the template feature vector of the communication feature matrix is ​​determined by using the dimensionality reduction technology;

[0008] Determine the regional difference degree of the maximum value in the neighborhood range based on the distribution of elements on both sides of the maximum value in the distance sequence of each row element in the communication feature matrix; determine the key sequence of each row element based on the regional difference degree; determine the offset coefficient of each row element according to the secret key sequence of each row element in the communication feature matrix;

[0009] The encryption key of each data in each row element is determined based on the offset coefficient of each row element using the logistics transformation encryption technology;

[0010] The communication data is encrypted based on the encryption key, and the ciphertext of all data is saved.

[0011] In one embodiment, clustering the communication data to obtain multiple data clusters includes the following specific steps:

[0012] Each communication data is converted into a decimal form through UTF-8 encoding conversion technology to obtain a numerical sequence corresponding to each communication data;

[0013] The mean of all values ​​in the numerical sequence corresponding to each communication data is taken as the first eigenvalue of each communication data;

[0014] The number of values ​​in the numerical sequence corresponding to each communication data is used as the data length of each communication data, and a clustering space is constructed according to the length and the first eigenvalue as the horizontal and vertical coordinates;

[0015] A data clustering algorithm is used to divide all communication data into multiple data clusters based on the first eigenvalue and the data points corresponding to the data length in the clustering space.

[0016] In one embodiment, the interpolation process of the numerical sequence based on the length of the communication data corresponding to the data point in each data cluster to obtain the communication feature matrix includes the following specific steps:

[0017] For any data cluster, the communication data with the largest data length among the communication data corresponding to the data points in each data cluster is used as the standard data of each data cluster;

[0018] For communication data corresponding to other data points other than the data points corresponding to the standard data, respectively, the numerical sequences corresponding to the other data points are interpolated and expanded by using a mean filling method to reach the data length of the standard data;

[0019] The standard data of each data cluster and the interpolation results of the numerical sequence corresponding to the data points in each data cluster are used as a row element in the matrix to obtain the communication feature matrix of each data cluster.

[0020] In one embodiment, the use of dimensionality reduction technology to determine the template feature vector of the communication feature matrix includes the following specific steps:

[0021] The data dimension reduction technology is used to obtain multiple eigenvectors of the communication feature matrix and the eigenvalues ​​of each eigenvector, and the eigenvector corresponding to the maximum value of all eigenvalues ​​is used as the template eigenvector of each data cluster.

[0022] In one embodiment, the degree of regional difference of the maximum value in the neighborhood range is determined as follows:

[0023] Calculate the absolute value of the difference between each row element in the communication feature matrix of each data cluster and the template feature vector of each data cluster in the same order, and use the sequence composed of the absolute values ​​of all the differences corresponding to each row element as the distance sequence of each row element;

[0024] For any maximum value in the distance sequence, take the first number of adjacent elements on the left and right of each maximum value as the center point to form the first neighborhood range of each maximum value;

[0025] In the value range determined according to the length of the distance sequence, the first quantity is iterated once with a step size of 1 to obtain the second quantity, and the second neighborhood range of each maximum value is formed by taking the second number of adjacent elements on the left and right of each maximum value as the center point;

[0026] The mapping result of the cumulative sum of all elements in the second neighborhood of each maximum value on the sigmoid function and the mapping result of the cumulative sum of all elements in the first neighborhood of each maximum value on the sigmoid function are taken as the regional difference degree of each maximum value in the second neighborhood.

[0027] In one embodiment, the determining of the key sequence of each row element based on the regional difference degree comprises the following specific steps:

[0028] t0: Get the regional difference degree of each maximum value in the same neighborhood range respectively, and calculate the first quartile of the regional difference degree of all maximum values ​​in the same neighborhood range as the decision threshold of the neighborhood range;

[0029] t1: Calculate the regional difference degree of each maximum value in each neighborhood range. If the regional difference degree of the maximum value in a neighborhood range is less than the decision threshold of this neighborhood range, iterate the neighborhood range until the regional difference degree of the maximum value in the neighborhood range is greater than the decision threshold of the neighborhood range;

[0030] t2: When the regional difference degree of each maximum value in the neighborhood range is greater than the decision threshold of the neighborhood range, the set of all elements of each maximum value in the neighborhood range is used as a screening area set of the distance sequence where the maximum value is located;

[0031] The sequence of all the screening area sets in each distance sequence in ascending order of the first element order value is used as the key sequence of the elements corresponding to a row in the communication feature matrix of each distance sequence;

[0032] t3: Determine the maximum value in the distance sequence of each row of elements in the communication feature matrix, repeat the above steps t0-t2, and obtain the key sequence of each row of elements in the communication feature matrix respectively.

[0033] In one embodiment, the offset coefficient of each row of elements is determined as follows:

[0034] The product of the number of maximum values ​​in the distance sequence of each row of elements and the kurtosis of the key sequence of each row of elements is used as the offset coefficient of each row of elements.

[0035] In one embodiment, the encryption key of each data in each row element is determined as follows:

[0036] Determine an iterative key sequence of each row of elements based on a first characteristic value of communication data corresponding to each row of elements; determine an encryption key for each data in each row of elements based on the iterative key sequence:

[0037]

[0038] In the formula, key i，m is the encryption key for the mth data in the i-th row, bin is the binary conversion symbol, 256 is the base of the binary number, V m，i is the iterative key for the mth data in the i-th row, >> is a binary right shift, length(i) is the number of data in the i-th row, mod is the remainder operation, q i is the offset coefficient of the i-th row element, The round-up symbol.

[0039] In one embodiment, the iterative key sequence of each row of elements is determined as follows:

[0040] Arrange the iteration keys of all elements in each row of elements in order of element positions to obtain an iteration key sequence for each row of elements.

[0041] In one embodiment, the encryption of the communication data based on the encryption key includes the following specific formula:

[0042] IM i，m =bin(Z i，m )xor key i,m

[0043] Among them, IM i，m is the ciphertext of the mth data in the i-th row element, Z i，m is the mth data in the i-th row, xor is the exclusive OR operator, key i，m is the encryption key for the mth data in the i-th row, and bin is the binary conversion operation symbol.

[0044] The beneficial effects of the technical solution of the present application are: first, a clustering space is constructed by the first eigenvalue and data length of the communication data to realize the clustering division of the communication data; secondly, the numerical sequence corresponding to all data points is interpolated and filled by the length of the standard data in each data cluster, which can ensure that the distribution of the original data in the numerical sequence is not affected as much as possible while filling the length, and the data security can be further enhanced after filling the additional data; then, considering that there may be repetitive content and high privacy content between different communication data, the distribution of elements on both sides of the maximum value in the distance sequence of each row of elements is determined based on the distribution of the elements on both sides of the maximum value in the neighborhood range, and the data points corresponding to the important content in the communication data can be screened by gradually expanding the neighborhood range; then, the key sequence and iteration coefficient of each row of elements are determined by a cyclic iteration method, which is convenient for the subsequent adaptive adjustment of the shift length when the key is generated; finally, the encryption of the communication data is completed based on the generated encryption key, and the ciphertext of all data is safely stored. The present application enhances the security encryption effect of communication data by performing adaptive key encryption. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 A flowchart of a method for secure data storage under 5G network security management provided in an embodiment of the present application;

[0047] Figure 2 A flow chart for obtaining a template feature vector of a data cluster provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Given the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0049] See also Figure 1 , which shows a step flow chart of a data security storage method under 5G network security management provided by an embodiment of the present application, the method comprising the following steps:

[0050] Step S001: Collect communication data in the 5G network and determine the data to be encrypted.

[0051] The purpose of this application is to protect the security of communication data by encrypting the communication data in the 5G network, so it is necessary to collect the communication data first.

[0052] Specifically, in the present application, the modes of data communication in the 5G network include simplex communication, half-duplex communication and full-duplex communication, and the modes of obtaining communication data include but are not limited to a communication transmitter, a communication receiver and a network node. The present application does not impose any special restrictions on the modes of data communication in the 5G network and the locations for obtaining communication data.

[0053] Preferably, as an embodiment of the present application, taking the communication in the call-answer mode in full-duplex communication mode as an example, the communication data in units of pieces generated by the caller and the answerer in the communication are collected, and each piece of communication data is converted into a decimal form through the UTF-8 (Unicode Transformation Format-8) encoding conversion technology, and a numerical sequence corresponding to each piece of communication data is obtained as the encrypted data to be encrypted later. Among them, the UTF-8 encoding conversion technology is a commonly used technology in the field of data communication, and the specific process is not repeated here.

[0054] Step S002: clustering the communication data to obtain multiple data clusters, constructing a communication feature matrix of the data clusters, and determining a template feature vector of the data clusters.

[0055] It should be noted that since the communication exchange in the call-answer mode usually relies on standard forms for reply, the communication content of the caller in the same field is highly similar. For example, in the process of civil aviation tower communication in the aviation field, the same department only has differences in flight number and content when communicating with the tower. Such information with differences is usually considered to be highly private data information; in the process of ship communication in the navigation field, it is necessary to first confirm the ship's call sign, and then conduct information transmission such as inquiry and notification. When different ships communicate with the coast station, the ship's call sign and the information transmitted are obviously different, while the position, direction, and navigation speed are usually similar due to the constraints of relevant systems. Therefore, such information with obvious differences is usually used as important information for encryption and protection.

[0056] Therefore, the embodiment of the present application constructs a clustering space according to the overall distribution of each communication data, and clusters all communication data to obtain several data clusters; the communication data in each data cluster has a high degree of similarity, and only the elements belonging to the important information area are different. Therefore, the embodiment of the present application obtains the difference in the important information area of ​​each communication data in a data cluster, and obtains the offset coefficient according to the difference in the important area, which is used to generate the key.

[0057] Specifically, this embodiment obtains the mean of all values ​​in the numerical sequence corresponding to each communication data as the first eigenvalue of each communication data, and obtains the data length of each communication data, where the data length is the number of values ​​in the numerical sequence of each communication data, and constructs a clustering space based on the length and the first eigenvalue, wherein the horizontal axis of the clustering space is the data length, and the vertical axis is the first eigenvalue.

[0058] Furthermore, the first eigenvalues ​​and data lengths of all communication data are obtained respectively, and based on the first eigenvalues ​​and data lengths, the data points corresponding to each communication data in the clustering space are determined, and all the data points are divided into k data clusters using a data clustering algorithm. The data lengths and first eigenvalues ​​of the communication data corresponding to the data points in each data cluster are relatively close.

[0059] It should be noted that data clustering is a commonly used technology in the field of data processing. The data clustering algorithm includes but is not limited to AP (Affinity Propagation) clustering algorithm, K-means clustering algorithm, and hierarchical clustering algorithm. This application does not impose any special restrictions on this.

[0060] Preferably, as an embodiment, all data points are divided into k data clusters using the AP clustering algorithm. In particular, for clustering algorithms that require a preset number of clusters, the elbow graph method is used to determine the preset number of clusters. The elbow graph method is a commonly used technology in the field of data processing, and the specific process is not repeated here.

[0061] Furthermore, for any data cluster, the embodiment of the present application further obtains a communication feature matrix for constructing the data cluster and determines a template feature vector of the data cluster. The process of obtaining the template feature vector is as follows: Figure 2 As shown, it includes steps 201 and 202:

[0062] 201 , obtaining standard data of each data cluster, and performing interpolation processing on the numerical sequences of other communication data.

[0063] First, the communication data with the largest data length among the communication data corresponding to the data points in each data cluster is used as the standard data of each data cluster. In particular, if there are multiple communication data corresponding to the maximum data length, the communication data corresponding to the data point with the closest metric distance to the cluster center of the data cluster is used as the standard data. The metric distance is the metric distance between the data points in the clustering process.

[0064] Secondly, in each data cluster, for the communication data corresponding to other data points other than the data points corresponding to the standard data, the other communication data are interpolated and expanded to reach the data length of the standard data. The interpolation method is the mean interpolation method, which performs mean interpolation based on the size of the data in the numerical sequence corresponding to each communication data. The purpose of this processing is, on the one hand, to ensure that the distribution of the original data in the numerical sequence is not affected as much as possible while filling the length; on the other hand, after filling in the additional data, the data security can be further enhanced, which is convenient for subsequent encryption processing. The mean interpolation method is a commonly used technology in the field of data processing, and the specific process will not be repeated.

[0065] 202, constructing a communication feature matrix of the data cluster and determining a template feature vector of the data cluster.

[0066] Subsequently, the communication feature matrix of each data cluster is constructed using the interpolated numerical sequences of all communication data corresponding to the data points in each data cluster. Specifically, the numerical sequence corresponding to the standard data of each data cluster is used as a row of elements in the matrix, and each interpolated numerical sequence is used as a row of elements in the matrix, and the resulting matrix is ​​used as the communication feature matrix.

[0067] Afterwards, data dimensionality reduction technology is used to obtain multiple eigenvectors of the communication feature matrix and the eigenvalues ​​of each eigenvector, and the eigenvector corresponding to the maximum value of all eigenvalues ​​is used as the template eigenvector of each data cluster.

[0068] The data dimensionality reduction technology is a commonly used technology in the field of data processing, including but not limited to principal component analysis PCA and linear discriminant analysis LDA. This application does not impose any special restrictions on data dimensionality reduction technology. Preferably, as an embodiment, PCA is used to obtain the template feature vector of each data cluster.

[0069] Step S003: Determine the regional difference degree of the maximum value in the neighborhood range based on the distribution of elements on both sides of the maximum value in the distance sequence of each row of elements in the communication feature matrix; and determine the key sequence of each row of elements based on the regional difference degree.

[0070] Specifically, the absolute values ​​of the differences between the elements in each row of the communication feature matrix of each data cluster and the template feature vector of each data cluster in the same order are calculated, and the sequence composed of the absolute values ​​of all the differences corresponding to the elements in each row is taken as the distance sequence A of each row of elements, and the regional difference degree of each maximum value in the distance sequence A is determined according to the degree of difference between each maximum value in the distance sequence A and adjacent elements in different neighborhood ranges.

[0071] Furthermore, for any maximum value in the distance sequence A, the neighborhood elements within the neighborhood range n are obtained with each maximum value as the center point, that is, n adjacent elements are taken on the left and right of the maximum value, and if the number of elements on the left and right sides is less than n, 0 is added to analyze the distribution difference of the adjacent elements of the maximum value in different neighborhood ranges, thereby evaluating the diffusion degree of the unstable characteristics of each maximum value in each neighborhood range. If the difference between adjacent elements increases with the increase of the neighborhood range, it means that there may be a large difference between the adjacent elements and the template feature vector of the data cluster, and the neighborhood range determined by the adjacent elements belongs to the unimportant information interval in the corresponding row of elements of the distance sequence A in the communication feature matrix. Among them, the value range of the neighborhood range n is M is the length of the distance sequence A, and round() is the rounding function.

[0072] Among them, the calculation formula for the regional difference degree of the pth maximum value in the distance sequence A under the neighborhood range n is:

[0073] U p，n (A) = sigmoid(Δp A,n )-sigmoid(Δp A,n-1 )

[0074] Where U p，n (A) is the regional difference degree of the pth maximum value in the distance sequence A under the neighborhood range n, sigmoid() is the S-shaped growth curve function, Δp A,n is the cumulative sum of all elements in the neighborhood range n of the pth maximum value in the distance sequence A, Δp A,n-1 It is the cumulative sum of all elements in the neighborhood range n-1 of the pth maximum value in the distance sequence A.

[0075] Among them, the greater the difference between the adjacent elements of the pth maximum in the neighborhood range n and the neighborhood range n-1 and the elements in the template feature vector, the greater the difference between the elements of a row in the communication feature matrix corresponding to the adjacent elements on both sides of the pth maximum and the template feature vector of the data cluster, and the greater the deviation from the overall data distribution of the communication feature matrix, that is, U p，n The larger the value of (A), the greater the difference between the pth maximum value in the distance sequence A and the elements in the template feature vector within the neighborhood range n.

[0076] Furthermore, the regional difference degree of each maximum value in the same neighborhood range is obtained respectively, and the first quartile of the regional difference degree of all maximum values ​​in the same neighborhood range is calculated as the decision threshold of the neighborhood range. The key sequence of each row element in the communication feature matrix is ​​determined based on the decision threshold, and the process of obtaining the key sequence is as follows:

[0077] T1: Calculate the degree of regional difference of each maximum value in each neighborhood range. If the degree of regional difference of a certain maximum value in the neighborhood range n is less than the decision threshold of the neighborhood range, it means that there is a small difference between the data in the communication feature matrix corresponding to the elements on both sides of the maximum value and the elements in the template feature vector of the data cluster, and the data in the communication feature matrix may be data existing in multiple interpolated numerical sequences; then let n=n+1, calculate the degree of regional difference of this maximum value in the neighborhood range n+1, and continue to iterate within the value range of n until the degree of regional difference in the neighborhood range is greater than the decision threshold of the neighborhood range. At this time, it means that there is a large difference between the data in the communication feature matrix corresponding to the elements on both sides of the maximum value and the elements in the template feature vector of the data cluster, and the data in the communication feature matrix may be data with higher privacy and lower repetitiveness in each communication data, which is significantly different from the common data corresponding to the repeated content in multiple communication data.

[0078] T2: When the regional difference degree of each maximum value in the neighborhood range is greater than the decision threshold of the neighborhood range, the set composed of all elements of each maximum value in the neighborhood range is used as the screening area set of the distance sequence where the maximum value is located, and the sequence composed of all screening area sets in each distance sequence A in the order of the first element order value in the set from small to large is used as the key sequence C of the corresponding row of elements in the communication feature matrix.

[0079] T3: Repeat the above steps T1-T2 to obtain the key sequence of each row element in the communication feature matrix respectively.

[0080] Step S004: Obtain the offset coefficient of each row of elements according to the secret key sequence of each row of elements in the communication feature matrix.

[0081] The key sequence of each row element in the communication feature matrix is ​​determined based on the common data features of each row element and all row elements, that is, the difference between the template feature vectors. Before encrypting the communication data, the application considers the distribution characteristics of the key sequence of each row element and the distance sequence A of each row element to determine the degree of aggregation of data with high privacy and low repetitiveness in the communication data corresponding to each row element, and obtains the offset coefficient of each row element, which is convenient for the subsequent determination of the shift of the numerical sequence corresponding to each row element when generating the key.

[0082] Specifically, the calculation formula for the offset coefficient of the i-th row element is:

[0083] q i = count(A i )*l(C i )

[0084] In the formula, q i is the offset coefficient of the i-th row element, A i is the distance sequence of the elements in the i-th row, count(A i ) is the number of maximum values ​​in the distance sequence of the elements in the i-th row, C i is the key sequence of the elements in row i, l(C i ) is the kurtosis of the key sequence of the i-th row elements.

[0085] Among them, the more maximum values ​​there are, the more personalized important data there are in the elements of the i-th row. The larger the kurtosis of the key sequence, the more concentrated the important data is, and the more it is necessary to adjust the distribution of the key to achieve enhanced encryption effect.

[0086] Step S005: using the logistics transformation encryption technology to determine the encryption key of each data in each row of elements based on the offset coefficient of each row of elements, and completing the encryption of the communication data based on the encryption key.

[0087] According to the above steps, the offset coefficients of each row element in the communication feature matrix are calculated respectively. Furthermore, the communication data is encrypted using the logistics transformation encryption technology in combination with the offset coefficients of each row element in the communication feature matrix and the first eigenvalue of the communication data corresponding to each row element. The specific process is as follows:

[0088] First, the normalized result of the first eigenvalue of the communication data corresponding to each row element is used as the initial key for encrypting each row element, and iteration is performed based on the initial key. Wherein, normalization is a commonly used technology in the field of data processing, and the normalization includes but is not limited to Min-Max normalization and Z-score standardization. Preferably, as an embodiment, Min-Max normalization is used to determine the initial key for encrypting each row element.

[0089] Secondly, the iterative formula of the initial key is determined by using the logistics transformation encryption technology, and the initial key for encrypting the i-th row element is recorded as V 0，i , the iteration formula for m iterations is:

[0090] V m，i =μ*V m-1，i ×(1-V m-1，i )

[0091] Where V m，i is the initial key V 0,i The result of m iterations, μ is the preset parameter in the logistics transformation encryption technology, and the value of μ is 4, V m-1，i is the initial key V 0,i The result of iterating m-1 times.

[0092] Afterwards, according to the above process, the initial key V is calculated in sequence 0，i The result of each iteration is used as the iteration key for each data in the i-th row element. The corresponding relationship is: 0,i The result of one iteration is used as the iteration key of the first data in the i-th row element, and the initial key V 0,i The result of m iterations is used as the iteration key for the mth data in the i-th row, and so on. The iteration keys of all elements in the i-th row are arranged in order of element position to obtain the iteration key sequence of the i-th row elements.

[0093] Then, the encryption key of each data in the i-th row of elements is obtained by combining the offset coefficient of the i-th row of elements and the iterative key sequence of the i-th row of elements. The encryption key of the m-th data in the i-th row of elements is obtained as follows:

[0094]

[0095] In the formula, key i，m is the encryption key for the mth data in the i-th row, bin is the binary conversion operator, 256 represents the base of the binary number, V m，i is the iterative key for the mth data in the i-th row, >> is a binary right shift, length(i) is the number of data in the i-th row, mod is the remainder operation, q i is the offset coefficient of the i-th row element, To round up.

[0096] Finally, the numerical sequence corresponding to the elements in the i-th row is encrypted based on the encryption key of each data in the i-th row element to obtain the ciphertext.

[0097] Encrypt the mth data using the encryption key of the mth data in the i-th row element to obtain the ciphertext of the mth data:

[0098] IM i，m =bin(Z i，m )xor key i,m

[0099] Among them, IM i，m is the ciphertext of the mth data in the i-th row element, Z i，mis the mth data in the i-th row, xor is the exclusive OR operator, key i，m is the encryption key for the mth data in the i-th row, and bin is the binary conversion operator.

[0100] Furthermore, according to the above steps, the ciphertext of each data in each row element in the communication feature matrix is ​​obtained respectively, and the ciphertext of all data is saved in a database or hardware device to complete the encryption of the communication data.

[0101] It should be noted that the above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

[0102] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A data security storage method under 5G network security management, characterized in that: The method comprises the following steps: Collect communication data in the 5G network and determine the data to be encrypted; The communication data is clustered to obtain multiple data clusters, and the communication feature matrix is ​​obtained by interpolating the numerical sequence based on the length of the communication data corresponding to the data point in each data cluster, and the template feature vector of the communication feature matrix is ​​determined by using the dimensionality reduction technology; Determine the regional difference degree of the maximum value in the neighborhood range based on the distribution of elements on both sides of the maximum value in the distance sequence of each row element in the communication feature matrix; determine the key sequence of each row element based on the regional difference degree; determine the offset coefficient of each row element according to the secret key sequence of each row element in the communication feature matrix; The encryption key of each data in each row element is determined based on the offset coefficient of each row element using the logistics transformation encryption technology; The communication data is encrypted based on the encryption key, and the ciphertext of all data is saved.

2. According to a method for secure data storage under 5G network security management according to claim 1, it is characterized in that: The clustering process for the communication data to obtain multiple data clusters includes the following specific steps: Each communication data is converted into a decimal form through UTF-8 encoding conversion technology to obtain a numerical sequence corresponding to each communication data; The mean of all values ​​in the numerical sequence corresponding to each communication data is taken as the first eigenvalue of each communication data; The number of values ​​in the numerical sequence corresponding to each communication data is used as the data length of each communication data, and a clustering space is constructed according to the length and the first eigenvalue as the horizontal and vertical coordinates; A data clustering algorithm is used to divide all communication data into multiple data clusters based on the first eigenvalue and the data points corresponding to the data length in the clustering space.

3. According to the method for secure data storage under 5G network security management according to claim 1, it is characterized in that: The interpolation process of the numerical sequence based on the length of the communication data corresponding to the data point in each data cluster to obtain the communication feature matrix includes the following specific steps: For any data cluster, the communication data with the largest data length among the communication data corresponding to the data points in each data cluster is used as the standard data of each data cluster; For communication data corresponding to other data points other than the data points corresponding to the standard data, respectively, the numerical sequences corresponding to the other data points are interpolated and expanded by using a mean filling method to reach the data length of the standard data; The standard data of each data cluster and the interpolation results of the numerical sequence corresponding to the data points in each data cluster are used as a row element in the matrix to obtain the communication feature matrix of each data cluster.

4. According to a method for secure data storage under 5G network security management according to claim 3, it is characterized in that: The specific steps of determining the template feature vector of the communication feature matrix by using the dimensionality reduction technology are as follows: Data dimensionality reduction technology is used to obtain multiple eigenvectors of the communication feature matrix and the eigenvalues ​​of each eigenvector, and the eigenvector corresponding to the maximum value of all eigenvalues ​​is used as the template eigenvector of each data cluster.

5. According to a method for secure data storage under 5G network security management as described in claim 1, it is characterized in that: The determination method of the regional difference degree of the maximum value in the neighborhood range is: Calculate the absolute value of the difference between each row element in the communication feature matrix of each data cluster and the template feature vector of each data cluster in the same order, and use the sequence composed of the absolute values ​​of all the differences corresponding to each row element as the distance sequence of each row element; For any maximum value in the distance sequence, take the first number of adjacent elements on the left and right of each maximum value as the center point to form the first neighborhood range of each maximum value; In the value range determined according to the length of the distance sequence, the first quantity is iterated once with a step size of 1 to obtain the second quantity, and the second neighborhood range of each maximum value is formed by taking the second number of adjacent elements on the left and right of each maximum value as the center point; The mapping result of the cumulative sum of all elements in the second neighborhood of each maximum value on the sigmoid function and the mapping result of the cumulative sum of all elements in the first neighborhood of each maximum value on the sigmoid function are taken as the regional difference degree of each maximum value in the second neighborhood.

6. According to a method for secure data storage under 5G network security management as described in claim 1, it is characterized in that: The step of determining the key sequence of each row element based on the regional difference degree includes the following specific steps: t0: Get the regional difference degree of each maximum value in the same neighborhood range respectively, and calculate the first quartile of the regional difference degree of all maximum values ​​in the same neighborhood range as the decision threshold of the neighborhood range; t1: Calculate the regional difference degree of each maximum value in each neighborhood range. If the regional difference degree of the maximum value in a neighborhood range is less than the decision threshold of this neighborhood range, iterate the neighborhood range until the regional difference degree of the maximum value in the neighborhood range is greater than the decision threshold of the neighborhood range; t2: When the regional difference degree of each maximum value in the neighborhood range is greater than the decision threshold of the neighborhood range, the set of all elements of each maximum value in the neighborhood range is used as a screening area set of the distance sequence where the maximum value is located; The sequence of all the screening area sets in each distance sequence in ascending order of the first element order value is used as the key sequence of the elements corresponding to a row in the communication feature matrix of each distance sequence; t3: Determine the maximum value in the distance sequence of each row element in the communication feature matrix, repeat the above steps t0-t2, and obtain the key sequence of each row element in the communication feature matrix respectively.

7. According to a method for secure data storage under 5G network security management as described in claim 1, it is characterized in that: The offset coefficient of each row of elements is determined as follows: The product of the number of maximum values ​​in the distance sequence of each row of elements and the kurtosis of the key sequence of each row of elements is used as the offset coefficient of each row of elements.

8. According to a method for secure data storage under 5G network security management as described in claim 1, it is characterized in that: The encryption key of each data in each row element is determined as follows: Determine an iterative key sequence of each row of elements based on a first characteristic value of communication data corresponding to each row of elements; determine an encryption key for each data in each row of elements based on the iterative key sequence: In the formula, key i，m is the encryption key for the mth data in the i-th row, bin is the binary conversion operator, 256 is the base of binary numbers, V m，i is the iterative key for the mth data in the i-th row, >> is a binary right shift, length(i) is the number of data in the i-th row, mod is the remainder operation, q i is the offset coefficient of the i-th row element, The round-up symbol.

9. According to the method for secure data storage under 5G network security management of claim 8, it is characterized in that: The iterative key sequence of each row of elements is determined as follows: Arrange the iteration keys of all elements in each row of elements in order of element positions to obtain an iteration key sequence for each row of elements.

10. According to a method for secure data storage under 5G network security management as described in claim 1, it is characterized in that: The encryption of the communication data based on the encryption key includes the following specific formula: IM i，m =bin(Z i，m )xor key i,m Among them, IM i，m is the ciphertext of the mth data in the i-th row element, Z i，m is the mth data in the i-th row, xor is the exclusive OR operator, key i，m is the encryption key for the mth data in the i-th row, and bin is the binary conversion operator.

Images