A data dynamic fusion iterative training modeling and application method

Abstract

The present invention discloses a data dynamic fusion iterative training modeling and application method, which finds the maximum value point in the curve and dynamically divides the ratio threshold interval based on it, and encrypts each adjacent node in the proposed path step by step according to different networking urgency values ​​through steps D1-D2, and finds a balance between ensuring that the exchanged data has a sufficient encryption level and ensuring the data exchange transmission efficiency for each adjacent node. The networking urgency value corresponding to the adjacent node whose distance characteristic value jumps is corrected by the maximum networking urgency value, and the networking urgency values ​​between adjacent nodes in the same proposed path that are in a disconnected state at the current time point are balanced, so as to avoid the situation where the networking urgency value of some adjacent nodes at the current time point has too large deviation and affects the overall data exchange transmission efficiency of the proposed path.

Description

Technical Field

[0001] The present invention relates to the field of data security technology, and in particular to a data dynamic fusion iterative training modeling and application method. Background Art

[0002] In order to increase data security, different types of business systems are usually isolated from each other by the network. However, in some specific scenarios, when it is necessary to use the business nodes in different types of business systems to cooperate with each other to complete the execution of a certain type of business service, it is necessary to connect the business nodes in the business service chain as data exchange nodes from the disconnected state to the connected state. In order to ensure the security of data exchange and transmission between different business systems, the exchanged data is usually encrypted before transmission. However, existing encryption algorithms usually have rules to follow, and in theory, as long as the computing power is sufficient, they can be easily cracked. Therefore, for scenarios with high data exchange security requirements, designing a set of algorithms with difficult encryption rules to ensure the security of data exchange and transmission is the first technical problem to be solved by this application.

[0003] In addition, when the amount of data that needs to be transmitted from the pointing node to the pointed node in the adjacent nodes in the above-mentioned business service chain at the current point in time is larger, the urgency value of the pointing node and the pointed node to be connected to the network is higher. At this time, if the intended exchange data of the pointing node is complexly encrypted, it is necessary to wait for a longer encryption time, which will extend the waiting time for the adjacent nodes to be connected to the network, and thus affect the efficiency of data exchange and transmission. Therefore, it is necessary to simplify the encryption process, but how much encryption complexity should be sacrificed in order to find a balance between ensuring that the exchanged data has a sufficient encryption level and ensuring the efficiency of data exchange and transmission for each adjacent node is the second technical problem that this application needs to solve.

[0004] Finally, the network urgency values ​​of each adjacent node in the business service chain are usually different, which means that each adjacent node has a different waiting time when exchanging data. However, the execution of a specified type of business service in the above specific scenario usually has time requirements. If the data exchange and transmission action is performed after the allowed waiting time is reached according to the network urgency value of each adjacent node, it will affect the overall execution efficiency of the specified type of business service. However, how to balance the network urgency values ​​between adjacent nodes in the business service chain has become the third technical problem that needs to be solved in this field. Summary of the invention

[0005] The present invention aims to improve the security of cross-system data exchange and transmission, and ensure that each adjacent node in the business service chain of cross-system data exchange and transmission finds a balance between ensuring that the exchanged data has a sufficient encryption level and ensuring the efficiency of data exchange and transmission. It provides a data dynamic fusion iterative training modeling and application method.

[0006] To achieve this object, the present invention adopts the following technical solutions:

[0007] Provided is a data dynamic fusion iterative training modeling and application method, comprising the steps of:

[0008] L1, for each group of adjacent nodes in the proposed path for data exchange transmission identified at the current time point and in a disconnected state, a network urgency value is calculated, and then, in the form of a sliding "delimited window", an encryption complexity requirement value interval that slides synchronously with the dynamically determined ratio threshold interval into which the network urgency value falls is obtained;

[0009] L2, solving the first mean value of the encryption complexity requirement value corresponding to each encryption link in the national secret generation algorithm whose encryption complexity requirement value falls within the encryption complexity requirement value interval;

[0010] L3, solving the distance characteristic value corresponding to the adjacent nodes in the proposed path , and obtain the latest updated online urgency value correction dataset;

[0011] L4, judge Is there a jump?

[0012] If yes, the networking urgency value of the adjacent node is corrected by using the second mean of the networking urgency fitting values ​​corresponding to each maximum point in the networking urgency value correction data set; if no, no correction is performed;

[0013] L5, using the revised data as the newly added samples to update the training revised data set to build a model, so as to obtain the updated revised data set of the network urgency value.

[0014] Preferably, in step L1, the method for identifying the proposed path comprises the steps of:

[0015] S1, by constructing hierarchical networks of different types of business systems and reducing the dimensionality of node space, in order to find data exchange nodes with cross-layer relationships that have the ability to execute specified business services at the current time point in one-dimensional space, and form several available data exchange paths from the data exchange start node to the data exchange target node;

[0016] S2, calculating the comprehensive urgency value of changing the interconnection state between adjacent nodes at different levels of each of the available data exchange paths at the current time point, and obtaining the historically used path with the maximum morphological similarity to each of the available data exchange paths,

[0017] If the acquisition is successful, then the historical comprehensive urgency value of each of the historically used paths when forming is further acquired, and the data exchange available path with the maximum morphological similarity and the historically used path are formed into a corresponding path combination, and then the process proceeds to step S31;

[0018] If the acquisition fails, go to step S32;

[0019] S31, solving the urgency value deviation of each of the path combinations, and taking the available data exchange path in the path combination with the minimum urgency value deviation as the proposed path;

[0020] S32, taking the available data exchange path with the largest comprehensive urgency value as the proposed path.

[0021] Preferably, the method for encrypting the service data exchanged from the pointing node among the adjacent nodes to the pointed node comprises the steps of:

[0022] C1, determining whether the proposed path is the first available path for data exchange in the path combination having the maximum morphological similarity and the minimum urgency value deviation obtained by step S31,

[0023] If so, in the historically used paths in the path combination, identify a second adjacent node having a corresponding relationship with a first adjacent node in the first data exchange available path that needs to perform data exchange transmission at the current time point, and then obtain historical dimensionality reduction information of each data exchange node in the second adjacent node;

[0024] If not, then calculating the dimension reduction information of the first adjacent node that needs to perform data exchange transmission at the current time point in the second data exchange available path determined as the proposed path in step S32;

[0025] C2, obtaining random number types bound to encryption links respectively corresponding to the encryption complexity requirement values ​​of the encryption complexity requirement value interval obtained by synchronously sliding following the ratio threshold interval into which the networking urgency value of the first neighboring node falls in step L1;

[0026] C3, extracting random numbers corresponding to each of the random number types obtained in step C2 from the historical dimensionality reduction information or the dimensionality reduction information and adding them to the corresponding encryption links.

[0027] Preferably, the method for encrypting the data to be exchanged and transmitted of the next adjacent node which is continuously arranged after the previous adjacent node in the proposed path with the encryption level gradually increased comprises the steps of:

[0028] D1, determining whether the encryption complexity requirement value interval of the next adjacent node is increased compared with the previous adjacent node,

[0029] If so, then obtaining each encryption link in which the data exchanged and transmitted between the previous adjacent nodes is encrypted with the dimensionality reduction information, and then screening out the remaining encryption links that have not been encrypted in the national encryption generation algorithm;

[0030] If not, refusing to re-encrypt the encrypted data that has been exchanged and transmitted at the previous adjacent node and is to be exchanged and transmitted at the next adjacent node;

[0031] D2, for the data to be exchanged and transmitted at the next adjacent node, through steps C1-C3, the random number of the corresponding type is added to each of the corresponding remaining encryption links.

[0032] Preferably, the networking urgency value is: at the current time point, the ratio of the difference between the amount of data to be sent from the pointing node to the pointed node among the adjacent nodes and the amount of data to be released from the pointed node to the amount of data to be sent.

[0033] Preferably, the method for dynamically determining the ratio threshold interval comprises the steps of:

[0034] L11, the networking urgency value associated with each adjacent node in each historically used path having the same data exchange start node and the same data exchange target node as the proposed path and the mean of the encryption complexity requirement , drawn in the same plane coordinate system Curve and curve; Indicates group the adjacent nodes;

[0035] L12, calculate the Curve and The distance characteristic value between points with positional symmetry in the curve , and then and The corresponding networking urgency value For fitting points, a high-order function curve is fitted to the fitting points associated with each of the adjacent nodes to obtain a curve representing the complex trend of encryption under different network urgency values. curve;

[0036] L13, find and obtain the The networking urgency fitting value corresponding to the extreme point in the curve, and the networking urgency fitting value corresponding to each of the extreme points are used as dividing points, and the historical networking urgency value interval composed of the minimum historical networking urgency value and the maximum historical networking urgency value of each data exchange node in each of the historically used paths is divided into a number of ratio threshold intervals.

[0037] Preferably, in step L1, the method for obtaining the encryption complexity requirement value interval that slides synchronously with the dynamically determined ratio threshold interval into which the networking urgency value falls in the form of a sliding "delimited window" is:

[0038] Calculating the ratio of the first width of each ratio threshold interval divided by the latest updated historical networking urgency value interval to the second width of the historical networking urgency value interval;

[0039] The overall interval of encryption complexity requirement values ​​composed of the minimum complexity requirement value and the maximum complexity requirement value in each encryption link in the national secret generation algorithm is taken as the demarcation window, and each of the proportions is taken as a sliding stop position. The encryption complexity requirement value interval having the proportions corresponding to the width of the demarcation window is obtained by sliding in the demarcation window.

[0040] Preferably, in step L3, the method for solving the distance characteristic value corresponding to the adjacent nodes in the proposed path is:

[0041] The proposed path is drawn in the same plane coordinate system by the method described in step L11. Curve and curve, and then calculate the Curve and The distance between points with symmetrical position in the curve is taken as the distance characteristic value ;

[0042] The acquired latest updated network urgency value correction data set includes the latest updated The coordinate information of each of the maximum points in the curve;

[0043] In step L4, when and or and When There is a jump; Indicates that the proposed path is sorted in the order of data exchange transmission. The node after the adjacent node of the group and with the closest sorting distance The distance characteristic value of the adjacent nodes of the group; Indicates that the proposed path is sorted in the order of data exchange transmission. The node before the adjacent node of the group and with the closest sorting distance The distance characteristic value of the adjacent nodes of the group.

[0044] Preferably, in step S1, the method for constructing the hierarchical network of different types of business systems comprises the steps of:

[0045] A1, for the data exchange nodes distributed at various locations under each of the business systems, a rectangular area is selected with a minimum rectangular selection area, and then the plane coordinates of any vertex of the rectangular area corresponding to each of the business systems are solved;

[0046] A2, taking the first vertex of the solved plane coordinates in the first rectangular area containing the data exchange start node as the origin of the three-dimensional space coordinate system, and for each of the remaining second rectangular areas, calculating the distance between the second vertex of the solved plane coordinates and the origin as the z-axis coordinate value of the corresponding second rectangular area in the three-dimensional space coordinate system;

[0047] A3, taking the first vertex and each of the second vertices as the z-axis positions of the corresponding rectangular areas in the three-dimensional space coordinate system, loading each rectangular area into the three-dimensional space coordinate system to form the hierarchical network that is parallel to the upper and lower parts in the three-dimensional space.

[0048] Preferably, the dimension reduction information includes: at least one of the node coordinate value of the data exchange node after three-dimensional dimension reduction, the z-axis coordinate value feature, the network hierarchical relationship with the adjacent nodes, the coordinate mapping relationship of dimension increase and dimension reduction, and the system type of the business system to which the data exchange node belongs;

[0049] The method for reducing the dimensionality of the data exchange nodes in three-dimensional space is: using the dimensionality reduction condition that the number of coordinate values ​​with corresponding axes and the same three-dimensional coordinate values ​​is at least "2", the data exchange nodes under each network level are reduced to a straight line on the xz axis or yz axis plane, so as to retain the z-axis coordinate value characteristics of the data exchange nodes after dimensionality reduction.

[0050] The present invention has the following beneficial effects:

[0051] 1. By building a hierarchical network of different types of business systems, As one of the dimensionality reduction information after the spatial dimensionality reduction of the data exchange nodes in the hierarchical network, when encrypting the data to be exchanged and transmitted between each adjacent node in the proposed path, due to the dimensionality reduction information of each adjacent node, such as The randomness of the data to be exchanged between each adjacent node is increased, which improves the security of data exchange and transmission. In addition, by using the dimensionality reduction information as the random number of the national secret encryption process, once the starting node of the data exchange changes, the hierarchical network structure will change, the random number will change dynamically, and it will be difficult to find the change pattern, which greatly improves the security of the exchange and transmission of encrypted data.

[0052] 2. Through steps D1-D2, each adjacent node in the proposed path is encrypted step by step according to different networking urgency values, finding a balance between data exchange transmission efficiency and data exchange security.

[0053] 3. In step S31, the available path for data exchange in the path combination with the maximum morphological similarity and the minimum urgency value deviation is used as the proposed path. When the business data to be exchanged and transmitted is subsequently encrypted, after the relationship between the proposed path and the data exchange nodes of the historically used paths in the path combination is matched, the dimensionality reduction information of the corresponding data exchange nodes in the historically used paths can be directly obtained to encrypt the business data. Under the condition of urgent time for data exchange and transmission, there is no need to calculate the dimensionality reduction information of each data exchange node in the proposed path at that time, but the dimensionality reduction information can be calculated after the data exchange is completed, which greatly improves the efficiency of data exchange and transmission.

[0054] 4. The higher the networking urgency value, the higher the efficiency requirement for data exchange and transmission. However, the more complex the encryption, the longer it usually takes. Therefore, in order to improve the efficiency of data exchange and transmission, it is necessary to sacrifice the complex encryption process and lower the encryption level. Similarly, the higher the network disconnection urgency value, the more urgent it is for adjacent nodes to change from a connected state to a disconnected state. In order to exchange data within the limited remaining networking time, it is also necessary to sacrifice the complex encryption process. However, how much encryption complexity is sacrificed can ensure a sufficient encryption level while ensuring the efficiency of data exchange and transmission. This application finds The maximum point in the curve is used as the basis for dynamically dividing the ratio threshold interval, and a balance is found for each adjacent node between ensuring that the exchanged data has a sufficient encryption level and ensuring the efficiency of data exchange transmission.

[0055] 5. The maximum value of network urgency is used to correct the network urgency value corresponding to the adjacent node where the distance characteristic value jumps. The network urgency values ​​of the adjacent nodes in the same proposed path (or historically used path) that are disconnected at the current time point are balanced to avoid the network urgency value deviation of some adjacent nodes at the current time point being too large, which affects the overall data exchange and transmission efficiency of the proposed path.

[0056] 6. Using any vertex in the rectangular area corresponding to the business system as the origin of the constructed hierarchical network enhances the randomness of the dimensionality reduction information, thereby making data encryption exchange and transmission more secure. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0058] Figure 1It is a diagram of implementation steps of a data dynamic fusion iterative training modeling and application method provided by one embodiment of the present invention;

[0059] Figure 2 is a schematic diagram of the structure of a hierarchical network constructed in an embodiment of the present invention;

[0060] Figure 3 It is an example diagram of the calculation of the morphological similarity between the available paths for data exchange and the paths used historically;

[0061] Figure 4 It is an example diagram for calculating the network disconnection urgency value or the network connection urgency value;

[0062] Figure 5 yes Curve and Schematic diagram of the curves drawn in the same plane coordinate system;

[0063] Figure 6 yes Example graph of the curve. DETAILED DESCRIPTION

[0064] The technical solution of the present invention is further described below with reference to the accompanying drawings and through specific implementation methods.

[0065] Among them, the drawings are only used for illustrative explanations, and they only represent schematic diagrams rather than actual pictures, and should not be understood as limitations on the present invention. In order to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0066] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right", "inside", "outside", etc. appear to indicate the orientation or position relationship, it is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the terms describing the position relationship in the drawings are only used for illustrative purposes and cannot be understood as a limitation on the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0067] In the description of the present invention, unless otherwise clearly specified and limited, if the term "connection" or the like appears to indicate the connection relationship between components, the term should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two components or the interaction relationship between two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0068] The embodiment of the present invention provides a data dynamic fusion iterative training modeling and application method, such as Figure 1 As shown, the steps include:

[0069] L1, for each group of adjacent nodes in the proposed path for data exchange transmission identified at the current time point that are in a disconnected state, a network urgency value is calculated, and then, in the form of a sliding "delineated window", an encryption complexity requirement value interval that slides synchronously with the dynamically determined ratio threshold interval into which the network urgency value falls is obtained;

[0070] The following is an explanation of the proposed path identification method:

[0071] The method for identifying the proposed path comprises the steps of:

[0072] S1, through the hierarchical network construction and node space dimensionality reduction of different types of business systems, in order to find the data exchange nodes with cross-layer relationships that have the ability to execute the specified business service at the current time point in one-dimensional space, and form several data exchange available paths from the data exchange starting node to the data exchange target node

[0073] The method of constructing a hierarchical network of different types of business systems includes the steps of:

[0074] A1, for the data exchange nodes distributed in various places under each business system, a rectangular area is selected with the minimum rectangular selection area, and then the plane coordinates of any vertex of the rectangular area corresponding to each business system are solved;

[0075] Assume that the servers in the first business system are distributed in various places in Hangzhou. These servers are used to provide a certain type of business service. Each server is used as a data exchange node. First, the coordinate position of each data exchange node in the first business system in the real physical space is obtained. Then, the rectangular area formed by these data exchange nodes is selected in the minimum area rectangular selection method, such as Figure 2 The rectangular area 100 shown in FIG. 1 is then solved for the plane coordinates of any vertex of the rectangular area 100. The plane coordinates of any vertex of the rectangular area 100 can be solved by an existing calculation method, for example, Figure 2The long side of the rectangle where the data exchange node 2 is shown in the figure is the x-axis, and the short side of the rectangle where the data exchange node 1 is shown in the figure is the y-axis. When the coordinate positions of the data exchange nodes 1 and 2 in the physical space are known, the plane coordinates of the vertex 3 can be solved according to the cosine theorem.

[0076] A2, taking the first vertex of the solved plane coordinates in the first rectangular area containing the data exchange start node as the origin of the three-dimensional space coordinate system, for each of the remaining second rectangular areas, calculating the distance between the second vertex of the solved plane coordinates and the origin as the z-axis coordinate value of the corresponding second rectangular area in the three-dimensional space coordinate system;

[0077] Assume that Figure 2 In the example, the first vertex 3 of the first rectangular area 100 containing the data exchange start node 4 is used as the origin of the three-dimensional space coordinate system, and the remaining second rectangular areas (such as Figure 2 The second rectangular area 200 in the plane coordinates is calculated and solved, and the distance between the second vertex 5 and the origin (i.e., the first vertex 3) is used as the z-axis coordinate value of the corresponding second rectangular area 200 in the three-dimensional space coordinate system. This z-axis coordinate value is preferably:

[0078] The average distance between each data exchange node in the second rectangular area and the data exchange start node in the first rectangular area in real space is converted to the conversion value of the three-dimensional space coordinate system. For example, Figure 2 The data exchange nodes (such as data exchange nodes 6 and 7) in the second rectangular area 200 are distributed in Ningbo, and the distances from the data exchange start node 4 in the first rectangular area 100 distributed in Hangzhou are, for example, as follows: the distances in real space between the data exchange nodes 6 and 7 and the data exchange start node 4 are respectively Assuming that there are only two data exchange nodes 6 and 7 in the second rectangular area 200, the z-axis coordinate value of the second rectangular area 200 in the three-dimensional space coordinate system is .

[0079] It should be noted here that the conversion value of the z-axis coordinate value of the rectangular area in the three-dimensional space coordinate system is the scaling ratio of the distance mean value in the above example. The distance mean value of 100km is converted to 1cm in the three-dimensional space coordinate system, and the scaling ratio is .

[0080] A3, taking the first vertex and each second vertex as the z-axis position of the corresponding rectangular area in the three-dimensional space coordinate system, loading each rectangular area into the three-dimensional space coordinate system, forming a three-dimensional space coordinate system as shown in FIG. Figure 2 The upper and lower parallel hierarchical networks shown in .

[0081] The following is an explanation of the method for performing spatial dimensionality reduction on each data exchange node at different levels in the constructed hierarchical network in step S1, which specifically includes the following steps:

[0082] B1, construct the coordinate mapping relationship between the real space and the three-dimensional space as expressed by the following formula (1):

[0083]

[0084] In formula (1), , Respectively represent the x-axis value, y-axis value or z-axis value of the data exchange node under the same business system in three-dimensional space and real space; Representation and The x-axis value, y-axis value or z-axis value of the vertex of the rectangular area with a coordinate correspondence as the origin in the three-dimensional space in the real space; Indicates the three-dimensional space conversion ratio; Represents the first The conversion value corresponding to the second rectangular area; when When is the x-axis value or y-axis value to be mapped, .

[0085] For example, suppose Figure 2 The x-axis coordinate value of the data exchange node 7 in real space is , the x-axis coordinate value of the first vertex 3 as the origin in real space is ,because is the x-axis value to be mapped, then , When solving the z-axis coordinate value of the data exchange node 7 in the three-dimensional space, since the z-axis coordinate value of the data exchange node 7 and the first vertex 3 as the origin in the real space is "0", the z-axis value of the data exchange node 7 in the three-dimensional space is .

[0086] B2, with the dimension reduction condition that the number of coordinate values ​​with corresponding axes and the same three-dimensional coordinate values ​​is at least "2", the data exchange nodes under each network level are reduced to the xz axis plane or the yz axis plane.

[0087] For example, building Figure 2After the hierarchical network of the example, for the data exchange starting node 4 and the data exchange node 2 in the first business system in the first network level, it is assumed that both are reduced to a straight line in the xz-axis plane. After the data exchange starting node 4 and the data exchange node 2 are reduced to a one-dimensional space, the coordinate values ​​of the z-axis corresponding to the axis are the same, the coordinate values ​​of the y-axis corresponding to the axis are the same, and the coordinate values ​​of the x-axis corresponding to the axis are different, satisfying the dimensionality reduction condition that the number of coordinate values ​​corresponding to the axis and having the same three-dimensional coordinate values ​​is at least "2", thereby Figure 2 The first rectangular area 100 in the first network level and the second rectangular area 200 in the second network level are reduced to the xz axis plane. The method of reducing the data exchange nodes in the rectangular area to the straight line on the yz axis plane is the same as the method of reducing the dimension to the xz axis plane, and will not be repeated.

[0088] The following is an explanation of the method for finding an available path for data exchange in step S1:

[0089] In this embodiment, in the one-dimensional space after dimensionality reduction, the method for finding available paths for data exchange is as follows: after identifying each data exchange node that has business service capabilities for the data to be exchanged from each network layer, a number of available paths for data exchange are formed from the data exchange start node to the data exchange target node, based on the condition that adjacent nodes in the path have a cross-layer relationship.

[0090] It should be emphasized here that the application scenario of the technical solution provided by this application is: how to ensure the security of data exchange and transmission in the network connection state between cross-systems where network isolation is the norm, therefore, the adjacent nodes in the available path must have a cross-layer relationship. The data exchange nodes in the same network layer are usually distributed in the same city or the same building or even the same computer room, so the distance between these data exchange nodes has a negligible effect on the rate of data exchange and transmission. Therefore, when looking for an available path, in this application, it is determined that there is a cross-layer relationship between adjacent nodes. The purpose is to select only one data exchange node in the same network layer when looking for an available path, so as to avoid increasing the data exchange transmission time when adjacent nodes are in the same network layer. In addition, the number of nodes constituting the available path is reduced, thereby reducing the number of data exchanges and maximizing the security of data transmission.

[0091] For example, suppose that the data to be exchanged is from Figure 2 Starting from the data exchange starting node 4 in the third network level, and taking the data exchange node 8 in the third network level as the target node for data exchange, in the process of finding the available path, it is assumed that Figure 2The data exchange start node 4 in the first network layer can transmit the data to be exchanged under the specified service type to the data exchange target node 8 through the data exchange nodes 6, 7, and 9 in the second network layer, or can transmit the data to be exchanged to the data exchange node 11 in the third network layer first, then to the data exchange node 10 in the second network layer, and finally to the data exchange target node 8. The available data exchange paths found are: the four available paths expressed as 4-6-8, 4-7-8, 4-9-8, and 4-11-10-8 formed by each data exchange node in the order of receiving the data to be exchanged.

[0092] The dimension reduction information of each data exchange node in the available data exchange path includes: the node coordinate value after three-dimensional dimension reduction, the z-axis coordinate value feature (i.e., the conversion value), the network hierarchy relationship with the adjacent nodes, the coordinate mapping relationship, and at least one of the system type of the business system to which it belongs. The system type of the business system is expressed by a specific code, such as the code "001" to represent the first business system; the coordinate value after three-dimensional dimension reduction is the coordinate value in the above formula (1) The coordinate value after dimension reduction to the xz-axis plane or the yz-axis plane; the network hierarchical relationship with the adjacent nodes, for example, in the data exchange available path of "4-6-8", the network hierarchical relationship between the data exchange node 6 and the adjacent nodes is: as the next network layer of the data exchange node 4, and as the previous network layer of the data exchange node 8, for example, it can be expressed as ; The mapping relationship of the coordinates is the relationship expressed by the above formula (1), which includes the hierarchical distance from the origin.

[0093] After finding the available paths for data exchange and obtaining the dimension reduction information of each data exchange node in each path, the method for identifying the proposed path proceeds to step:

[0094] S2, through path similarity matching and data traffic prediction, calculates the comprehensive urgency value of each available data exchange path to change the interconnection status between adjacent nodes at different levels, and obtains the historical used path with the maximum morphological similarity to each available data exchange path,

[0095] If the acquisition is successful, the historical comprehensive urgency value of each historically used path when it is formed is further acquired, and the data exchange available path and the historically used path with the maximum morphological similarity (the primary condition for having morphological similarity is that the data exchange available path and the historically used path have the same data exchange starting node and data exchange target node) are formed into a corresponding path combination, and then the process goes to step S31;

[0096] If the acquisition fails, go to step S32;

[0097] The calculation method of the comprehensive urgency value specifically includes the following steps:

[0098] S21, determining whether the number of historically used paths matched from the historical path database and having the same service type as the available path for data exchange and whose morphological similarity is greater than a similarity threshold is "0",

[0099] If so, sort the available paths for data exchange from small to large according to the sum of the hierarchical distances between adjacent nodes in the path to form a list of available paths;

[0100] If not, extract the historically used path with the greatest morphological similarity to each available data exchange path, and then arrange the extracted historically used paths from high to low according to the morphological similarity to form a historically used path list;

[0101] In step S21, the matching parameters of the morphological similarity include at least one of the angle information between the direct line and the horizontal line between adjacent nodes in the path, the number of nodes, and the number of network level jumps between adjacent nodes; the angle information includes the degree of the angle and the quadrant feature of the angle;

[0102] The following is an explanation of the calculation method of morphological similarity, which specifically includes the following steps:

[0103] S211, calculating the angle information between the straight line and the horizontal line between the adjacent nodes in the available path for data exchange, and then determining whether a historically used path having a first similarity with the angle information corresponding to each group of adjacent nodes in the available path is matched from the historical path library, and if so, proceeding to step S212;

[0104] For example, Figure 3 The left side (a) is an example of a path available for data exchange; Figure 3 The right side (b) of the figure is an example of a historically used path. In step S211, first, the angle information between the direct line between adjacent nodes in the available path for data exchange and the horizontal line is calculated, such as Figure 3 The angle information between the direct line L1 between the data exchange nodes 10 and 20 and the horizontal line includes the angle ,as well as The quadrant characteristics are: In the fourth quadrant; The quadrant characteristics are: The quadrant in which the angle is located is determined by taking the data exchange node as the origin, dividing the plane space into four quadrants by a horizontal line passing through the data exchange node and a vertical line passing through the data exchange node perpendicular to the horizontal line, and then determining the angle less than 90°. The quadrant in which the angle is located is used as the quadrant characteristic of the angle.

[0105] Having the first similarity means that the corresponding angles have the same quadrant characteristics, that is, they are in the same quadrant space, such as Figure 3 The angle between the available paths for data exchange in The angle with the historical path have the same quadrant characteristics, and the angle deviation is less than the preset deviation threshold, such as and The absolute value of the difference is less than the preset absolute value threshold of the difference. It should be noted that the first similarity is the comprehensive similarity between the adjacent nodes in the available path and the adjacent nodes in the historically used path, that is, Figure 3 The angle in and ,as well as and When both reach the first similarity, it is determined Figure 3 There is a first similarity between the first adjacent nodes formed by the data exchange nodes 10 and 20 and the second adjacent nodes formed by the historical data exchange nodes 30 and 40. When all the first adjacent nodes in the data exchange available path are matched to the corresponding second adjacent nodes with the first similarity in the same historically used path, the historically used path is filtered out and the process proceeds to step S212; otherwise, it is determined that the first similarity matching between the historically used path and the data exchange available path fails, and the historically available path is filtered out from the historical path library.

[0106] S212, for each second adjacent node in each historically used path remaining after filtering in step S211, determine whether the first network level hop number of the second adjacent node has a second similarity with the second network level hop number of the first adjacent node whose angle information in the available path has a first similarity, and if so, proceed to step S213;

[0107] For example, assuming that the remaining historically used paths filtered in step S211 are Figure 3 In the figure (b) on the right side of the figure, the historical data exchange nodes 30 and 40 in the figure (b) constitute one of the second adjacent nodes, and the first network level jump number of the second adjacent node is assumed to be "3" (for example, the historical data exchange node 30 is at the first network level, and the historical data exchange node 40 is at the fourth network level, then the first network level jump number between nodes 40 and 30 is "3"). In the available path, the first adjacent node (such as Figure 3The second network layer hop number of the data exchange node 10 and the data exchange node 20 shown in the figure is assumed to be "2" (for example, the data exchange node 10 is at the first network layer, and the data exchange node 20 is at the third network layer, then the second network layer hop number between nodes 10 and 20 is "2"). Since the ratio of the first network layer hop number to the second network layer hop number is two-thirds, which is not an integer ratio relationship, it is determined that the second adjacent node does not have a second similarity with the first adjacent node. When all second adjacent nodes in the same historically used path have first adjacent nodes that satisfy the same second similarity (the same integer ratio relationship) in the available data exchange path, it is determined that the historically used path has a second similarity with the available data exchange path, otherwise, the historically used path is filtered out from the historical path library. It should be noted here that in order to ensure an integer proportional relationship, when calculating the second similarity, the larger value of the level jump number is used as the denominator and the smaller value is used as the numerator. Moreover, once the level jump number in the historically used path or the available path for data exchange is determined as the denominator, when performing the second similarity matching, the level jump numbers of all adjacent nodes in the previously determined historically used path or the available path for data exchange are used as the denominator in the second similarity calculation.

[0108] S213, for each historically used path with a second similarity in the number of network level jumps, determine whether the second number of nodes in the used path and the second number of nodes in the available path have a third similarity, and if so, combine the first similarity, the second similarity and the third similarity to form a morphological similarity.

[0109] The third similarity is set, for example, as follows: the difference between the number of nodes in the historically used path (defined as the first number of nodes) and the number of nodes in the data exchange available path (defined as the second number of nodes) is less than a preset difference threshold.

[0110] In addition, in order to facilitate the calculation of the morphological similarity between the historically used paths and the available paths for data exchange, the above-mentioned first similarity, second similarity, and third similarity can be quantified into a first similarity value, a second similarity value, and a third similarity value, respectively. There are many existing methods for quantifying the similarity values, so they are not elaborated in detail.

[0111] It should be noted that when any step in steps S211-S213 is judged as "if no", it indicates that no historical available path with morphological similarity to the data exchange available path is matched, and the judgment process of the subsequent steps is terminated, and an available path list is formed in the current step.

[0112] After the above step S21, after forming the available path list or the historically used path list, in step S2, the calculation and screening method of the comprehensive urgency value is transferred to the step:

[0113] S22, extract the top ranking in the available path list of elements, and for each group of adjacent nodes in each element that are in a connected state or in an isolated state, respectively calculate the disconnection urgency value and the connection urgency value, and then merge the disconnection urgency values ​​and the connection urgency values ​​calculated for the same element into a comprehensive urgency value (for example, by weighted summation); or of elements, and obtain the historical comprehensive urgency value corresponding to the element;

[0114] In this embodiment, for the elements in the available path list, the ratio of the difference between the amount of data to be sent from the pointing node to the pointed node and the amount of data to be released at the pointed node to the amount of data to be sent at the current time point is used as the network disconnection urgency value or network connection urgency value between the adjacent nodes.

[0115] by Figure 4 For example, suppose Figure 4 The adjacent nodes composed of data exchange nodes 10 and 20 in the available path for data exchange represented are in a networked state at the current time point. The data exchange node 10 has a data intended to be sent with the data exchange node 20 as the target, and the data exchange nodes 10 and 20 are respectively defined as the pointing node and the pointed node at the current time point. And at the current time point, the data exchange node 20 as the pointed node has a corresponding amount of data to be released. In order to improve the security of cross-system data transmission and reception, the shorter the networking time, the higher the security. In this embodiment, in order to achieve this goal, the above ratio is used as the basis for judging the urgency of disconnection of the adjacent nodes. When the difference between the amount of data intended to be sent by the pointing node and the amount of data to be released by the pointed node is greater than the ratio of the amount of data intended to be sent, it means that the data transmission and reception of the pointing node and the pointed node are more unbalanced. At this time, the pointing node lacking the amount of data intended to be sent or the pointed node lacking the amount of data to be released is eager to change from the networked state to the disconnected state with another of the adjacent nodes. Therefore, in this embodiment, the above ratio is used as the disconnection urgency value between adjacent nodes.

[0116] The larger the amount of data that the pointing node intends to send to the pointed node, the more urgent it is to establish a networking relationship with the pointed node; and when the pointed node has data to be released, the smaller the amount of data to be released within a certain time period, the greater the probability that the data will be used across layers to execute designated business services within the time period, and the lower the urgency of the designated node changing from a connected state to a disconnected state. This is because, once the network is disconnected, the data currently owned by the pointed node cannot be successfully consumed by the designated business service. At this time, it is necessary to find a parasitic node, transfer the data currently owned by the pointed node to the parasitic node, and then disconnect from the network. This will increase the complexity of business service execution. Therefore, the lower the urgency of the pointed node changing from a connected state to a disconnected state. In order to achieve the above-mentioned technical effects, in this embodiment, when calculating the networking urgency value, the amount of data to be released is cleverly set to a negative value, so that the more balanced the amount of data to be sent to the pointed node and the amount of data to be released to the pointed node, the higher the urgency value for the network between adjacent nodes at different network levels to change from an isolated state to a networking state. This prevents the adjacent node from being changed from an isolated state to a networking state when the amount of data to be sent is insufficient or the pointed node needs to remain in an isolated state, thereby increasing the network security risk caused by the transmission and reception of large amounts of data.

[0117] The comprehensive urgency value is the weighted sum of the disconnection urgency value and the connection urgency value.

[0118] In this embodiment, the method for obtaining the historical comprehensive urgency value of each group of adjacent nodes in the elements in the historical used path list includes the steps of:

[0119] S221, for each group of adjacent nodes in the data exchange available path that have a corresponding relationship with the historically used path with morphological similarity when used as the historically available path, calculate the fourth similarity of the connected or isolated state, if both are in the connected or disconnected state, the fourth similarity is "1", otherwise it is "0", and then the sum of the fourth similarities is used as the comprehensive similarity;

[0120] For example, suppose Figure 3 When the historically used path represented by (b) is used as a historically available path, its data exchange node 30 and the data exchange node 40 are in a networked state. Figure 3 If the first adjacent node composed of data exchange nodes 10 and 20 in the available data exchange path expressed in Figure (a) and having morphological similarity with the second adjacent node composed of data exchange nodes 30 and 40 is also in a networked state, then the fourth similarity between the second adjacent node and the second adjacent node is "1", otherwise it is "0", and then the fourth similarities between each group of adjacent nodes in the historically used path and the corresponding adjacent nodes in the available data exchange path with morphological similarity are summed, and the sum result is used as the comprehensive similarity of the historically used path.

[0121] S222, ranked first from the historical used path list of Among the elements, extract and obtain the top comprehensive similarity ranking The historical comprehensive urgency values ​​corresponding to each historically used path.

[0122] After step S2, the calculation of the comprehensive urgency value of each available data exchange path is completed, or the historical comprehensive urgency value of the historically used path with the maximum morphological similarity to each available data exchange path is obtained, and under the "if" judgment condition of step S2, the method of identifying the intended path proceeds to step S31:

[0123] Solve the urgency value deviation in each path combination (preferably the absolute value of the difference between the comprehensive urgency value of the available data exchange path in the combination at the current time point and the historical comprehensive urgency value of the historically used path in the combination when it was formed), and use the available data exchange path in the path combination with the minimum urgency value deviation as the proposed path;

[0124] Under the "if no" determination condition of step S2, the method for identifying the proposed path proceeds to step:

[0125] S32, taking the available data exchange path with the largest comprehensive urgency value as the proposed path.

[0126] The following describes the method for dynamically determining the ratio threshold interval in step L1:

[0127] The method for dynamically determining the ratio threshold interval includes the steps of:

[0128] L11, the networking urgency value associated with each adjacent node in each historically used path with the same data exchange start node and the same data exchange target node and the mean of the encryption complexity requirement , drawn in the same plane coordinate system Curve and curve; Indicates Group adjacent nodes;

[0129] Figure 5 b1 and b2 represent Curve and Curve. Complexity requirement mean The method is as follows: First, obtain the network urgency value by sliding the "delimited window" The dynamically determined ratio threshold interval falls into the encryption complexity requirement interval of the synchronous sliding. Then, extract The encryption complexity requirement values ​​of each encryption link of the data exchanged and transmitted by the adjacent nodes are calculated. Finally, the average value of the extracted encryption complexity requirement values ​​is calculated as Corresponding .

[0130] For example, the national encryption algorithm used to encrypt data to be exchanged and transmitted has 8 encryption links as shown in Table a below, and the complexity requirement values ​​corresponding to each encryption link are as follows:

[0131] Table a

[0132]

[0133] Assume that the overall interval of the encryption complexity requirement values ​​of each encryption link of the national encryption algorithm is [0,1], and the obtained encryption complexity requirement value interval is assumed to be [0,0.5]; Assume that V in Table a above E2 、V E4 、V E5 If the encryption complexity requirement value of falls into the encryption complexity requirement value interval [0,0.5], then calculate V E2 、V E4 、V E5 The average value of Corresponding .

[0134] L12, calculation Curve and The distance characteristic value between points with positional symmetry in the curve , and then and Corresponding network urgency value To fit the points, a high-order function curve is fitted to the fitting points of each adjacent node to obtain the complex trend of encryption under different networking urgency values. curve;

[0135] Figure 5 The horizontal and vertical coordinates of the fitting point P1 in are The horizontal and vertical coordinates of the fitting point P1' which is symmetrical to the fitting point P1 are , , the distance characteristic value between the fitting point P1 and the fitting point P1' , and then use each and Corresponding network urgency value To fit the points, a high-order function curve is fitted to the fitting points associated with each adjacent node. The high-order function can be expressed by the following formula (2):

[0136]

[0137] In formula (2), is the term coefficient, , are the independent variable and dependent variable of the high-order function respectively. Curve like Figure 6 Example.

[0138] L13, Find and Get The network urgency fitting value corresponding to the extreme point in the curve is used as the dividing point, and the historical network urgency value interval composed of the minimum historical network urgency value and the maximum historical network urgency value of each data exchange node in each historically used path is divided into several ratio threshold intervals.

[0139] For example, suppose that Figure 6 shown The maximum points found in the curve include , , ,and The corresponding network urgency fitting values ​​are from large to small, that is, Corresponding network emergency fitting value Greater than Corresponding network emergency fitting value Greater than Corresponding network emergency fitting value Assume that at the current time point, the minimum historical networking urgency value among the historical networking urgency values ​​corresponding to the adjacent nodes in the historically used paths having the same data exchange starting node and the same data exchange target node as the proposed path identified in step L1 is , the maximum historical network urgency value is , then the latest updated historical network urgency value interval is [ .

[0140] Then with The corresponding network urgent fitting values ​​are the dividing points, and [ Divided into [ , ),[ , ),[ , ),[ , These four ratio threshold intervals. It should be noted here that when When filtering out [ , ,when , then filter out [ , ).

[0141] The following specifically describes how to obtain the encryption complexity requirement value interval that slides synchronously with the dynamically determined ratio threshold interval into which the networking urgency value falls in the form of a "defined window" in this embodiment:

[0142] First, the ratio of the first width of each ratio threshold interval divided by the latest updated historical networking urgency value interval to the second width of the historical networking urgency value interval is calculated;

[0143] For example, the first width of the ratio threshold interval is: , ), whose first width is and The difference (assuming it is defined as the first difference). Historical network urgency value interval [ The second width is and The ratio of the first width to the second width is: the ratio of the first difference to the second difference.

[0144] Then, the overall interval of encryption complexity requirement values ​​composed of the minimum complexity requirement value and the maximum complexity requirement value in each encryption link in the national secret generation algorithm is used as the demarcated window, and each proportion is used as the sliding stop position. The encryption complexity requirement value interval with a proportion corresponding to the width of the demarcated window is obtained by sliding in the demarcated window.

[0145] For example, in Table a above, assuming V E3 The encryption complexity requirement of V is the smallest. E8 The encryption complexity requirement value of is the largest, then the overall interval of the encryption complexity requirement value is expressed as [V E3 ,V E8 ]. Since the networking urgency value and the encryption complexity requirement value have an inverse relationship (the larger the networking urgency value, the lower the encryption complexity requirement value), [V E3 ,V E8 The “delimited window” of ] is expressed as [V E8 ,V E3 ].

[0146] Assume that , )exist[ The proportion of [V E8 ,V E3 The stop position in ] is ,Right now[ , )The encryption complexity requirement range is [ , ].

[0147] Through step L1, after obtaining the encryption complexity requirement value interval that slides synchronously with the dynamically determined ratio threshold interval into which the networking urgency value of the adjacent node at the current time point falls, the data dynamic fusion iterative training modeling and application method provided in this embodiment proceeds to step:

[0148] L2, solving the first mean value of the encryption complexity requirement value corresponding to each encryption link in the national secret generation algorithm for the encryption complexity requirement value interval into which the encryption complexity requirement value falls;

[0149] Continuing with the above example, suppose we fall into [ , ] The encryption links in the national encryption generation algorithm in this encryption complexity requirement value range include the , , , then the first mean is .

[0150] L3, solve the distance eigenvalues ​​corresponding to the adjacent nodes in the proposed path , and obtain the latest updated online urgency value correction dataset;

[0151] Draw the proposed path in the same plane coordinate system using the method described in step L11 Curve and curve, and then calculate Curve and The distance between points with symmetrical position in the curve is used as the distance characteristic value of the adjacent nodes in the proposed path. It should be noted that in the same plane coordinate system, the proposed path is drawn Curve and curve, with all adjacent nodes in the proposed path corresponding to is the curve fitting point. Curve and The function of the curve is preferably a quadratic function, Any one of them is the independent variable of the quadratic function, and the other is the dependent variable.

[0152] The latest updated network urgency value correction data set includes the corresponding values ​​of each adjacent node in each of the latest generated historical used paths. , and the latest update The coordinate information of each maximum point in the curve.

[0153] L4, judge Is there a jump?

[0154] If so, the networking urgency value of the adjacent node is corrected by using the second mean of the networking urgency fitting values ​​corresponding to each maximum point in the networking urgency value correction data set; if not, no correction is performed.

[0155] In this embodiment, when and or and When There is a jump; Indicates that the proposed path is sorted in the order of data exchange transmission. The node after the adjacent node of the group and with the closest sorting distance The distance characteristic value of the adjacent nodes of the group; Indicates that the proposed path is sorted in the order of data exchange transmission. The node before the adjacent node of the group and with the closest sorting distance The distance characteristic value of the adjacent nodes of the group.

[0156] For example, suppose Figure 4 The distance characteristic value corresponding to the first adjacent node formed by the data exchange nodes 10 and 20 in the proposed path is The distance characteristic value corresponding to the second adjacent node formed by the data exchange nodes 20 and 50 is ; The distance characteristic value corresponding to the third adjacent node formed by the data exchange nodes 50 and 60 is , then when and or and ,determination A jump occurs. At this time, the network urgency of the second adjacent node at the current time point needs to be corrected. The correction method is: use the latest updated network urgency value to correct the mean of the network urgency fitting values ​​corresponding to each extreme point in the data set (defined as the second mean) to correct the network urgency value of the second adjacent node at the current time point, that is, replace the network urgency value of the second adjacent node at the current time point with the second mean.

[0157] Finally, the data dynamic fusion iterative training modeling and application method provided in this embodiment enters the steps:

[0158] L5, use the correction data as the new sample to update the training correction data set to build the model, so as to obtain the updated network urgency value correction data set. The correction data set construction model includes fitting , curve, Models such as curves, ratio threshold interval division, encryption complexity requirement value interval division, etc.

[0159] The following describes the application scenarios and application methods of the data dynamic fusion iterative training modeling method provided in this embodiment:

[0160] The application scenario is to encrypt the data to be exchanged when data is exchanged and transmitted between the pointing node and the pointed node in the adjacent nodes having a cross-layer network relationship. The encryption method specifically includes the following steps:

[0161] C1, determine whether the proposed path is an available data exchange path (defined as the first available data exchange path) in the path combination with the maximum morphological similarity and the minimum urgency value deviation obtained by step S31,

[0162] If so, then in the historically used paths in the path combination (the path combination where the first data exchange available path is located), identify the second adjacent node that has a corresponding relationship with the first adjacent node in the first data exchange available path that needs to perform data exchange transmission at the current time point, and then obtain the historical dimensionality reduction information of each data exchange node in the second adjacent node;

[0163] If not, then calculating the dimension reduction information of the first adjacent node that needs to perform data exchange transmission at the current time point in the available data exchange path (the second available data exchange path) determined as the proposed path in step S32;

[0164] For example, suppose the proposed path is Figure 3 The first available path for data exchange expressed in (a) is: Figure 3 The historically used path expressed in (b) of the figure and the first data exchange available path constitute a path combination with the maximum morphological similarity and the minimum urgency value deviation. The specific correspondence between the first adjacent node in the first data exchange available path that needs to perform data exchange transmission at the current time point and the second adjacent node in the historically used path in the path combination has been specifically described in the above description of the scheme of step S2, and will not be repeated here. In addition, under the "if no" judgment condition of step C1, how to calculate the dimensionality reduction information of the first adjacent node has also been specifically described in the above content, and will not be repeated here.

[0165] C2, obtaining random number types bound to encryption links respectively corresponding to encryption complexity requirement values ​​in the encryption complexity requirement value interval obtained by synchronously sliding following the ratio threshold interval into which the networking urgency value of the first adjacent node falls in step L1;

[0166] For example, assuming that after step L1, the encryption complexity requirement value interval obtained by synchronously sliding following the ratio threshold interval into which the networking urgency value of the first adjacent node falls is [ , ], falling into [ , ] The encryption complexity requirement values ​​in the national secret generation algorithm in this encryption complexity requirement value range are assumed to include the values ​​in Table a above. , , , then get , , The random number types bound separately, such as The bound random number type is the node coordinate value of the pointed node among the adjacent nodes after three-dimensional dimensionality reduction.

[0167] C3, extract the random numbers corresponding to the random number types obtained in step C2 from the historical dimensionality reduction information or the calculated dimensionality reduction information obtained in step C1 and add them to the corresponding encryption links. For example, assuming The corresponding encryption link is: the pointed node forwards the authentication request pointing to node 10 to the authentication center, and the encryption method of the authentication request is, for example, to add the random number of the pointed node 20 to the authentication request. For example, the random number of the pointed node 20 includes: the node coordinate value of the pointed node after three-dimensional dimensionality reduction, and the random number is added to Corresponding to the above encryption link.

[0168] The following is a brief description of the process of encrypting data to be exchanged and transmitted, taking a national secret encryption algorithm as an example:

[0169] For example, encrypting the exchanged transmission data between adjacent nodes in a networked state includes the following steps:

[0170] for example, Figure 4 The pointing node 10 and the pointed node 20 shown in FIG. 1 are in a networked state. At this time, the exchange data needs to be transmitted from the node 10 to the node 20. The encryption process is as follows:

[0171] First, the pointed node 20 forwards the authentication request pointing to the node 10 to the authentication center, and adds the random number of the pointed node 20 (one or more specified by the node coordinate value of the pointed node 20 after three-dimensional dimension reduction, the network hierarchy relationship with the adjacent nodes, the coordinate mapping relationship, and the system type of the business system to which it belongs) to the authentication request;

[0172] Then, after receiving the authentication request, the authentication center generates a first authentication password, which includes a random number of the pointed node 20 and is sent to the pointing node 10 via the pointed node 20;

[0173] After receiving the first authentication code, the pointing node verifies the authority to exchange data with the pointed node 20, then generates a second authentication code (including the pointing node's random number), and sends it to the authentication center through the pointed node.

[0174] Finally, the authentication center cross-verifies the correctness of the first authentication password and the second authentication password, and after the verification is passed, notifies the pointing node to transmit the data exchange to the pointed node. It should be noted here that the method of cross-verifying the first authentication password and the second authentication password is not within the scope of the rights protection claimed by this application, nor does it affect the solution of the technical problem of this application, so it is not described in detail.

[0175] In order to further find a balance between encryption complexity and data exchange transmission efficiency, in this embodiment, it is preferred to encrypt the data to be exchanged and transmitted of the next adjacent node that is continuously arranged after the previous adjacent node in the proposed path with the encryption level gradually increased. The specific encryption method includes the steps of:

[0176] D1, determine whether the encryption complexity requirement value interval of the next adjacent node is increased compared with the previous adjacent node,

[0177] If so, then obtain the encryption links that have encrypted the data exchanged and transmitted between the previous adjacent nodes with the dimensionality reduction information, and then filter out the remaining encryption links that have not been encrypted in the national secret generation algorithm;

[0178] If not, then refuse to re-encrypt the secret data that has been exchanged and transmitted at the previous adjacent node and is to be exchanged and transmitted at the next adjacent node;

[0179] For example, by Figure 4 The encryption complexity requirement value interval of the previous adjacent node composed of data exchange nodes 10 and 20 in is assumed to be (0.1, 0.2), Figure 4 The encryption complexity requirement value interval of the next adjacent node after the previous adjacent node formed by the continuous order of the data exchange nodes 20 and 50 is assumed to be (0.3-0.4), then the "if" judgment condition of step D1 is achieved. Assume that the data exchanged and transmitted between the previous adjacent nodes have been encrypted using the first encryption link and the second encryption link in the national secret generation algorithm. Assume that the national secret generation algorithm also has a third encryption link, then the remaining encryption links that have not been encrypted in the national secret generation algorithm are selected as the third encryption link. And suppose that the data exchanged and transmitted between the previous adjacent nodes have been encrypted using all the encryption links in the national secret generation algorithm. When the encrypted data that has been encrypted in the previous adjacent node is continuously transmitted to the pointed node in the next adjacent node, the encrypted data will no longer be encrypted twice to improve the efficiency of data exchange and transmission.

[0180] D2, for the data to be exchanged and transmitted at the next adjacent node, through the above steps C1-C3, the corresponding type of random number is added to each corresponding remaining encryption link.

[0181] In addition, in order to further increase the security of data exchange and transmission, after the data to be exchanged is encrypted with national secrets, this embodiment also performs the following actions: after the data to be exchanged containing the national secrets with random numbers added is exchanged and transmitted to the data exchange node, the network is restored to its original state (the original isolation state, i.e., restored to the network isolation state).

[0182] In addition, in the scenario where the data exchange has an allowable waiting time, this embodiment does not immediately transmit the national secret data to be exchanged with the random number added to the adjacent node. The specific operation method is:

[0183] Calculate an ideal time point, then open up the network between adjacent nodes, and restore the network to its original state after completing the exchange transmission.

[0184] The calculation method of the ideal time point includes the following steps:

[0185] D1, starting the timing with the instruction that the pointing node among the adjacent nodes in the isolated state in the proposed path generates the business data with the national secret added with the random number;

[0186] E2, according to the preset frequency, determines whether the accumulated timing reaches the preset allowed waiting time.

[0187] If so, the network between the adjacent nodes is opened up;

[0188] If not, go to step E3;

[0189] E3, determine whether the adjacent nodes have reached a state of data accumulation and release balance,

[0190] If so, the ideal time point is achieved, and then the network between the adjacent nodes is opened;

[0191] If not, return to step E2.

[0192] The data accumulation and release balance state in step E3 is: at the current time point, the ratio of the difference between the amount of data to be sent from the pointing node to the pointed node and the amount of data to be released from the pointed node among the adjacent nodes to the amount of data to be sent is within the preset ratio threshold range; the amount of data to be released is a negative value.

[0193] In summary, the present invention finds The maximum value point in the curve is found, and the ratio threshold interval is dynamically divided based on this. Through steps D1-D2, each adjacent node in the proposed path is encrypted step by step according to different networking urgency values, and a balance is found for each adjacent node between ensuring that the exchanged data has a sufficient encryption level and ensuring the efficiency of data exchange and transmission. The networking urgency value corresponding to the adjacent node whose distance characteristic value jumps is corrected with the maximum networking urgency value, and the networking urgency values ​​between adjacent nodes in the same proposed path that are disconnected at the current time point are balanced to avoid the situation where the networking urgency value of some adjacent nodes at the current time point has too large deviations, which affects the overall data exchange and transmission efficiency of the proposed path.

[0194] It should be noted that the above specific implementations are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art should understand that various modifications, equivalent substitutions, changes, etc. can be made to the present invention. However, as long as these changes do not deviate from the spirit of the present invention, they should be within the scope of protection of the present invention. In addition, some terms used in the specification and claims of this application are not restrictive, but are only for the convenience of description.

Claims

1. A data dynamic fusion iterative training modeling and application method, characterized in that: Includes steps: L1, for each group of adjacent nodes in the proposed path for data exchange transmission identified at the current time point and in a disconnected state, a network urgency value is calculated, and then, in the form of a sliding "delimited window", an encryption complexity requirement value interval that slides synchronously with the dynamically determined ratio threshold interval into which the network urgency value falls is obtained; L2, solving the first mean value of the encryption complexity requirement value corresponding to each encryption link in the national secret generation algorithm whose encryption complexity requirement value falls within the encryption complexity requirement value interval; L3, solving the distance characteristic value corresponding to the adjacent nodes in the proposed path , and obtain the latest updated online urgency value correction dataset; L4, judge Is there a jump? If yes, the networking urgency value of the adjacent node is corrected by using the second mean of the networking urgency fitting values ​​corresponding to each maximum point in the networking urgency value correction data set; if no, no correction is performed; L5, using the correction data as the newly added samples to update the training correction data set to build a model, so as to obtain the updated correction data set of the network urgency value; The networking urgency value is: at the current time point, the ratio of the difference between the amount of data to be sent from the pointing node to the pointed node and the amount of data to be released from the pointed node to the amount of data to be sent.

2. According to the data dynamic fusion iterative training modeling and application method of claim 1, it is characterized in that: In step L1, the method for identifying the proposed path comprises the steps of: S1, by constructing hierarchical networks of different types of business systems and reducing the dimensionality of node space, in order to find data exchange nodes with cross-layer relationships that have the ability to execute specified business services at the current time point in one-dimensional space, and form several available data exchange paths from the data exchange start node to the data exchange target node; S2, calculating the comprehensive urgency value of changing the interconnection state between adjacent nodes at different levels of each of the available data exchange paths at the current time point, and obtaining the historically used path with the maximum morphological similarity to each of the available data exchange paths, If the acquisition is successful, then the historical comprehensive urgency value of each of the historically used paths when forming is further acquired, and the data exchange available path with the maximum morphological similarity and the historically used path are formed into a corresponding path combination, and then the process proceeds to step S31; If the acquisition fails, go to step S32; S31, solving the urgency value deviation of each of the path combinations, and taking the available data exchange path in the path combination with the minimum urgency value deviation as the proposed path; S32, taking the available data exchange path with the largest comprehensive urgency value as the proposed path.

3. A data dynamic fusion iterative training modeling and application method according to claim 2, characterized in that: The method for encrypting service data exchanged from a pointing node among the adjacent nodes to a pointed node comprises the steps of: C1, determining whether the proposed path is the first available data exchange path in the path combination having the maximum morphological similarity and the minimum urgency value deviation obtained by step S31, If so, in the historically used paths in the path combination, identify a second adjacent node having a corresponding relationship with a first adjacent node in the first data exchange available path that needs to perform data exchange transmission at the current time point, and then obtain historical dimensionality reduction information of each data exchange node in the second adjacent node; If not, then calculating the dimension reduction information of the first adjacent node that needs to perform data exchange transmission at the current time point in the second data exchange available path determined as the proposed path in step S32; C2, obtaining random number types bound to encryption links respectively corresponding to the encryption complexity requirement values ​​of the encryption complexity requirement value interval obtained by synchronously sliding following the ratio threshold interval into which the networking urgency value of the first neighboring node falls in step L1; C3, extracting random numbers corresponding to each of the random number types obtained in step C2 from the historical dimensionality reduction information or the dimensionality reduction information and adding them to the corresponding encryption links.

4. A data dynamic fusion iterative training modeling and application method according to claim 3, characterized in that: The method for encrypting data to be exchanged and transmitted of the next adjacent node which is continuously arranged after the previous adjacent node in the proposed path with the encryption level gradually increased comprises the steps of: D1, determining whether the encryption complexity requirement value interval of the next adjacent node is increased compared with the previous adjacent node, If so, then obtaining each encryption link in which the data exchanged and transmitted between the previous adjacent nodes is encrypted with the dimensionality reduction information, and then screening out the remaining encryption links that have not been encrypted in the national encryption generation algorithm; If not, refusing to re-encrypt the encrypted data that has been exchanged and transmitted at the previous adjacent node and is to be exchanged and transmitted at the next adjacent node; D2, for the data to be exchanged and transmitted at the next adjacent node, through steps C1-C3, the random number of the corresponding type is added to each of the corresponding remaining encryption links.

5. The data dynamic fusion iterative training modeling and application method according to claim 1 is characterized in that: The method for dynamically determining the ratio threshold interval comprises the steps of: L11, the networking urgency value associated with each adjacent node in each historically used path having the same data exchange start node and the same data exchange target node as the proposed path and the mean of the encryption complexity requirement , drawn in the same plane coordinate system Curve and curve; Indicates group the adjacent nodes; L12, calculate the Curve and The distance characteristic value between points with positional symmetry in the curve , and then and The corresponding networking urgency value For fitting points, a high-order function curve is fitted to the fitting points associated with each of the adjacent nodes to obtain a curve representing the complex trend of encryption under different network urgency values. curve; L13, find and obtain the The networking urgency fitting value corresponding to the extreme point in the curve, and the networking urgency fitting value corresponding to each of the extreme points are used as dividing points, and the historical networking urgency value interval composed of the minimum historical networking urgency value and the maximum historical networking urgency value of each data exchange node in each of the historically used paths is divided into a number of ratio threshold intervals.

6. A data dynamic fusion iterative training modeling and application method according to claim 5, characterized in that: In step L1, the method of obtaining the encryption complexity requirement value interval that slides synchronously with the dynamically determined ratio threshold interval into which the networking urgency value falls in the form of a sliding "delimited window" is: Calculating the ratio of the first width of each ratio threshold interval divided by the latest updated historical networking urgency value interval to the second width of the historical networking urgency value interval; The overall interval of encryption complexity requirement values ​​composed of the minimum complexity requirement value and the maximum complexity requirement value in each encryption link in the national secret generation algorithm is taken as the demarcation window, and each of the proportions is taken as a sliding stop position. The encryption complexity requirement value interval having the proportions corresponding to the width of the demarcation window is obtained by sliding in the demarcation window.

7. A data dynamic fusion iterative training modeling and application method according to claim 5, characterized in that: In step L3, the method for solving the distance characteristic value corresponding to the adjacent nodes in the proposed path is: The proposed path is drawn in the same plane coordinate system by the method described in step L11. Curve and curve, and then calculate the Curve and The distance between points with symmetrical position in the curve is taken as the distance characteristic value ; The acquired latest updated network urgency value correction data set includes the latest updated The coordinate information of each of the maximum points in the curve; In step L4, when and or and When There is a jump; Indicates that the proposed path is sorted in the order of data exchange transmission. The node after the adjacent node of the group and with the closest sorting distance The distance characteristic value of the adjacent nodes of the group; Indicates that the proposed path is sorted in the order of data exchange transmission. The node before the adjacent node of the group and with the closest sorting distance The distance characteristic value of the adjacent nodes of the group.

8. A data dynamic fusion iterative training modeling and application method according to claim 2 or 3, characterized in that: In step S1, the method for constructing the hierarchical network of different types of business systems includes the following steps: A1, for the data exchange nodes distributed at various locations under each of the business systems, a rectangular area is selected with a minimum rectangular selection area, and then the plane coordinates of any vertex of the rectangular area corresponding to each of the business systems are solved; A2, taking the first vertex of the solved plane coordinates in the first rectangular area containing the data exchange start node as the origin of the three-dimensional space coordinate system, and for each of the remaining second rectangular areas, calculating the distance between the second vertex of the solved plane coordinates and the origin as the z-axis coordinate value of the corresponding second rectangular area in the three-dimensional space coordinate system; A3, taking the first vertex and each of the second vertices as the z-axis positions of the corresponding rectangular areas in the three-dimensional space coordinate system, loading each rectangular area into the three-dimensional space coordinate system to form the hierarchical network that is parallel to the upper and lower parts in the three-dimensional space.

9. The data dynamic fusion iterative training modeling and application method according to claim 3 is characterized in that: The dimension reduction information includes: at least one of the node coordinate value of the data exchange node after three-dimensional dimension reduction, the z-axis coordinate value feature, the network hierarchy relationship with the adjacent nodes, the coordinate mapping relationship between dimension increase and dimension reduction, and the system type of the business system to which the data exchange node belongs; The method for reducing the dimension of the data exchange nodes in the three-dimensional space is: using the dimension reduction condition that the number of coordinate values ​​with corresponding axes and the same three-dimensional coordinate values ​​is at least "2", the data exchange nodes under each network level are reduced to a straight line on the xz axis or yz axis plane, so as to retain the z-axis coordinate value characteristics of the data exchange nodes after the dimension reduction.

Images