A data storage method for a trusted terminal

Abstract

The present invention belongs to the field of information technology application innovation technology, and specifically relates to a data storage method for a trusted terminal, the method comprising: establishing a data circulation model based on the data source and data flow of each data; establishing a data genetic model based on the idea of genetic algorithm and the data circulation model; obtaining the importance of each data set according to the genetic performance of each data in the data genetic model; obtaining the heat change model of each data set based on the retrieval record of each data and the straight line fitting method; obtaining the number of backups of each data set according to the importance of the data set and the heat change model of the data set, and obtaining several backup plans; obtaining the excellence of each backup plan according to the genetic performance of the nodes in each backup plan in the data genetic model; and storing data with the backup plan with the greatest excellence. The present invention improves the storage security of trusted terminal data and reduces storage costs.

Description

Technical Field

[0001] The present invention relates to the field of information technology application innovation technology. More specifically, the present invention relates to a data storage method for an information technology terminal. Background Art

[0002] The ICT industry is a key area for achieving independent and controllable information technology. With the development of operating systems, software, and hardware related to the ICT industry, enterprises have begun to use ICT terminals as tools for business operations and production, for example, using them to monitor workshop production and equipment operation data. At the same time, network security issues are increasing. Therefore, when using ICT terminals for data storage, it is necessary to ensure data security, thereby promoting the healthy development of the ICT industry and the stable operation of enterprise production.

[0003] In order to ensure the security and effective storage of stored data, layered storage and redundant backup methods are currently commonly used. In related technologies, for example, the Chinese patent document with authorization announcement number CN103558991B discloses a data layered storage processing method, device and storage device, which discloses that data is divided into cold data and hot data by detecting the data access frequency of the migration unit, so that the high-level disk only stores hot data, thereby controlling the consumption of data storage resources. For example, the Chinese patent document with authorization announcement number CN11714952B discloses a server data backup and recovery system and method, which discloses that the backup plan can be flexibly adjusted according to factors such as data importance and update frequency, which can improve data backup efficiency and reduce resource usage.

[0004] However, during data storage and backup, data is divided and stored based on access frequency and importance, lacking analysis of the connections between data. This results in slower tracing of important or hot data. Furthermore, while increasing the number of backups improves disaster recovery capabilities, it also increases the likelihood of data leaks. Backing up important or hot data to a single location can lead to the complete leakage of all of it, resulting in significant losses. Summary of the Invention

[0005] In order to solve the above-mentioned technical problems of insufficient security and weak data traceability during data storage, the present invention provides a data storage method for an information-based terminal, comprising:

[0006] Acquire several data sets of the trusted innovation terminal, where the data set is a collection of data of different time periods, and obtain metadata of all data of each data set, where the metadata includes data detection time, data source, data flow direction, and data size; establish a data circulation model based on the data source and data flow direction of each data; establish a data genetic model based on the genetic algorithm idea and the data circulation model; acquire the importance of each data set according to the genetic expression of each data in the data genetic model; acquire the heat change model of each data set based on the retrieval record of each data and the straight line fitting method; acquire the number of each data set that should be backed up according to the importance of the data set and the heat change model of the data set; establish a backup node sequence according to the number of each data set that should be backed up; randomly place all nodes of the backup node sequence in all backup units to obtain several backup schemes; acquire the superiority of each backup scheme according to the genetic expression of the nodes in all backup units in the data genetic model in each backup scheme; store data with the backup scheme with the highest superiority.

[0007] This invention analyzes the flow of various enterprise data through a data circulation model, thereby improving data traceability and increasing the efficiency of stored data extraction. It also uses genetic algorithms to further analyze data flows and establish a data genetic model, thereby capturing the genetic representation of data and its importance. This allows for more accurate determination of the number of data backups required, reducing data storage costs.

[0008] Preferably, the establishment of the data circulation model includes: taking any data as a node, connecting any two nodes according to the data source and data flow relationship between any two nodes; recording the distance between any two nodes as 1, and performing a depth-first search algorithm on the nodes with an in-degree of 0 in the data circulation graph as the starting points, obtaining the shortest distance from each starting point to each node, recording it as the level of each node, and when the level of the node is not unique, recording the minimum value of the node level as the level of the node, connecting all nodes according to the level, and obtaining the data circulation model.

[0009] Preferably, the connecting of any two nodes includes: for the i-th node and the c-th node, if the data corresponding to the c-th node is the data source of the data corresponding to the i-th node, then the i-th node and the c-th node are connected, and the direction points to the i-th node; if the data corresponding to the c-th node is the data flow direction of the data corresponding to the i-th node, then the i-th node and the c-th node are connected, and the direction points to the c-th node.

[0010] Preferably, the establishment of the data inheritance model includes: treating each data as a 01 string of the same size as the data size, recorded as the corresponding chromosome of each data; obtaining the inheritance probability sequence of each data set based on the data circulation of each data set reflected by the data circulation model; obtaining the node-to-node selection operator and the data set-to-data set mutation operator based on the data size of the data corresponding to different nodes and the data circulation model; the data inheritance model is expressed as, the corresponding chromosome of the data corresponding to the e-th level node is inherited with the inheritance probability sequence of the corresponding data set to obtain the node of the e+1th level, and the corresponding chromosome of the data corresponding to the node of the e+1th level is represented by the product of the corresponding chromosome of the data corresponding to the e-th level node and the selection operator and the mutation operator.

[0011] Preferably, the method of obtaining the genetic probability sequence of each data set includes: treating any data set as a target data set, obtaining the frequency distribution sequence of the data flow of all data in the target data set, recording it as the genetic sequence of the target data set, and dividing each value of the genetic sequence of the target data set by the number of data in the target data set to obtain the genetic probability sequence of the target data set.

[0012] The present invention determines the flow of data from a relatively macroscopic perspective by obtaining the genetic probability of a data set, thereby avoiding the impact of adding or deleting data on the data genetic model between data and increasing the scope of application of the data genetic model.

[0013] Preferably, obtaining a node-to-node selection operator includes:

[0014] If the bth node is the data source of the ath node, then the selection operator of the ath node on the bth node satisfies the expression:

[0015] ;

[0016] Where, Represents the selection operator of the a-th node to the b-th node; 、 Indicates the data size of the data corresponding to the a-th node and the b-th node; 、 Represents the data size set of the corresponding data sets of the data corresponding to the a-th node and the b-th node; represents the maximum value function; Represents the normalization function.

[0017] The present invention analyzes the specific impact of the selection operator on the data. When the selection operator between the data is small, although the data is circulated, the changes in the low-level data have little impact on the high-level data.

[0018] Preferably, the step of obtaining a mutation operator of a data set to a data set includes:

[0019] If the bth node is the data source of the ath node, then the mutation operator of the corresponding data set of the ath node's corresponding data to the corresponding data set of the bth node's corresponding data satisfies the expression:

[0020] ;

[0021] Where, represents the mutation operator of the corresponding data set of the data corresponding to the a-th node to the corresponding data set of the data corresponding to the b-th node; 、 Represents the data size set of the corresponding data sets of the data corresponding to the a-th node and the b-th node; represents the maximum value function; represents the edit distance function; Represents the normalization function.

[0022] The present invention measures the relationship between data sets through a mutation operator. When the mutation operator is large, even if the selection operator between data is large, the impact of changes in low-level data on high-level data is small.

[0023] Preferably, the obtaining of the importance of each data set includes: taking each node of the first level as a starting point, traversing the data flow model using a depth-first search algorithm to obtain all possible paths; and calculating the node sequence of the mth possible path between the bth node and the ath node. ,Will The selection operator of any two adjacent nodes is multiplied by the mutation operator of the corresponding data set of the two adjacent nodes, and the cumulative multiplication is obtained. To obtain the importance of the bth node to the ath node, add the importance of the bth node to the ath node in all possible paths between the bth node and the ath node to obtain the importance of the bth node to the ath node; for the dth data set, obtain the importance of the node corresponding to the data contained in the dth data set to all other nodes, and sum them up to obtain the importance of the dth data set.

[0024] Preferably, the number of data sets to be backed up satisfies the expression:

[0025] ;

[0026] Where, Indicates the number of backups that should be made for the dth data set; 、 Indicates the minimum and maximum values of the preset backup quantity range; Indicates the importance of the d-th data set; Represents the slope of the heat change model of the d-th data set; Represents the ceiling function.

[0027] The present invention determines the number of data sets that should be backed up based on the importance of the data sets and the popularity change model of the data sets, thereby making the backup data more effective and reducing the data storage cost.

[0028] Preferably, obtaining the quality of each backup solution includes:

[0029] The normalized negative correlation of the importance of the bth node to the ath node is used as the distance between the bth node and the ath node;

[0030] The goodness of any backup plan satisfies the expression:

[0031] ;

[0032] Where, Indicates the goodness of the t-th backup plan; Indicates the number of backup units; represents the distance between the u-th node and the v-th node in the s-th backup unit of the t-th backup scheme; Indicates the number of nodes in the sth backup unit of the tth backup scheme.

[0033] The beneficial effects of the present invention are:

[0034] (1) The present invention adopts the idea of genetic algorithm to analyze the relationship between the data of the terminal, improves the traceability of the data, can measure the impact of each data, and ensures the analysis of the relationship between the data;

[0035] (2) The present invention selects the backup solution with the highest degree of optimization to avoid storing strongly associated data in the same storage unit, thereby avoiding large losses and reducing the possibility of data leakage. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 1 is a flow chart schematically illustrating a data storage method for an information-based terminal in the present invention;

[0037] Figure 2 is a diagram schematically illustrating a data flow diagram;

[0038] Figure 3 is a schematic diagram schematically illustrating a data circulation model;

[0039] Figure 4 is a node diagram schematically illustrating a data circulation model;

[0040] Figure 5 It is a schematic diagram schematically illustrating the data inheritance process. DETAILED DESCRIPTION

[0041] The embodiment of the present invention discloses a data storage method for a terminal of a computer system. Figure 1 , including steps S1 to S4:

[0042] S1: Obtain several data sets of the trusted computing terminal, where the data set is a collection of data of different time periods for one item, and obtain the metadata of all data in each data set.

[0043] It should be noted that, taking a production-oriented enterprise as an example, the departments involved in data mainly include production, logistics, warehousing, sales, quality, equipment, supply chain, etc. In order to analyze the relationship between data between departments, and to store and back up the data of the enterprise's information and innovation terminals from an overall level, it is necessary to obtain various data in a more comprehensive manner.

[0044] Specifically, set a time period and collect various data from the trusted innovation terminals in the trusted innovation environment according to the time period to obtain several data sets. Any data set is a collection of data from different time periods for one data item. Establish the enterprise metadata dictionary and obtain the metadata of all data in each data set.

[0045] The credible innovation environment is a credible innovation data collection, transmission and storage environment composed of credible innovation chips, credible innovation encryption systems, credible innovation sensors, etc. The metadata dictionary is a standardized management tool for uniformly naming metadata, which provides a basis for data collection personnel or data creation personnel to fill in the metadata of the data. The metadata includes the generation time, data source, data flow, retrieval record, data size, etc. The retrieval record is the credible innovation terminal that retrieves data from the pre-stored database and the corresponding retrieval time. It should be noted that, taking the product quality data of the production process as an example, the metadata of the product quality data includes the detection time, data source, data flow, data size, etc. The data source is the various parameters of the product, and the data flow is the subsequent data produced by the quality, supply chain and other departments with reference to the product quality data.

[0046] At this point, several items of data from the trusted computing terminal and the metadata of each item of data have been obtained.

[0047] S2: Establish a data circulation model based on the data source and data flow of each data.

[0048] It should be noted that, taking a production-oriented enterprise as an example, there are various ways of data circulation between departments. For example, the supply chain and procurement department provide raw material data to the production department, the production department provides product data to the warehousing department, the sales department provides order data to the warehousing department, and the warehousing department completes the order delivery, etc. Therefore, the present invention establishes a data circulation model, so that the logical relationship between the data can be clarified, providing a basis for hierarchical storage and redundant backup of data.

[0049] Specifically, any data is regarded as a node, and any two nodes are connected. The connection process is: for the i-th node and the c-th node, if the data corresponding to the c-th node is the data source of the data corresponding to the i-th node, then the i-th node and the c-th node are connected, and the direction points to the i-th node; if the data corresponding to the c-th node is the data flow direction of the data corresponding to the i-th node, then the i-th node and the c-th node are connected, and the direction points to the c-th node.

[0050] After completing the connection of all nodes, we can get the data flow graph, such as Figure 2 This is a schematic diagram of the data flow diagram.

[0051] It should be noted that since the circulation of nodes is relatively complex and the circulation direction of a node is not unique, the enterprise's data circulation diagram will form a complex network. In order to manage the nodes more clearly, the present invention classifies the nodes in the data circulation diagram.

[0052] Preferably, the distance between any two nodes is recorded as 1, and the nodes with in-degree 0 in the data flow graph are used as starting points to perform a depth-first search algorithm to obtain the shortest distance from each starting point to each node, which is recorded as the level of each node. When the level of a node is not unique, the minimum value of the node level is recorded as the level of the node. All nodes are connected according to the level to obtain a data flow model, such as Figure 3 Schematic diagram of the data flow model.

[0053] At this point, the data flow model is obtained.

[0054] S3: Based on the genetic algorithm concept and the data circulation model, a data genetic model is established; the importance of each data set is obtained based on the genetic performance of each data in the data genetic model; based on the retrieval records of each data set and the straight-line fitting method, the heat change model of each data set is obtained; based on the importance of the data set and the heat change model of the data set, the number of backups of each data set is obtained.

[0055] It should be noted that if the data corresponding to the nodes at the lower level are missing, it will affect the data corresponding to all the nodes to which it flows, thereby causing a greater impact. Therefore, the importance of the data corresponding to the nodes at the lower level is higher. However, the data corresponding to the nodes at the lower level often have a larger amount of data. For example, the various parameters of the products in the production process, in subsequent production, the specific data of a product are often not retrieved again. Instead, the product qualification rate extracted from the various parameters of the product is used to optimize the operation of the production machine and the use of raw materials. Therefore, although the data corresponding to the nodes at the lower level are of higher importance, the retrieval frequency is low. If a higher redundancy is used for backup, a large amount of data storage costs will be generated. Therefore, the present invention combines the importance of the data and the usage of the data to adaptively determine the redundant backup amount for each data.

[0056] It should be noted that if Figure 4 This is a node diagram of the data flow model. Nodes A and B are both third-level nodes, but the data corresponding to node B has more data flows and affects the rest of the data at the same level. Therefore, the importance of node B is higher. Therefore, the importance of each data should also be determined based on the scope of its influence.

[0057] It should be noted that the data flow process is not just a simple selection from low-level data, but also involves a large amount of data processing, such as denoising operations. Therefore, the data flow process can be regarded as a data selection and mutation process. When the data source is not unique, there is also crossover, so the data can be regarded as chromosomes, and the data flow process is the genetic process of chromosomes. Therefore, the present invention analyzes the impact of data flow through the idea of genetic algorithm.

[0058] Specifically, establish a data genetic model:

[0059] Each data is regarded as a 01 string of the same size as the data size, and is recorded as the corresponding chromosome of each data.

[0060] Consider any dataset as the target dataset, obtain the frequency distribution sequence of the data flow of all the data in the target dataset, record it as the genetic sequence of the target dataset, and divide each value of the genetic sequence of the target dataset by the number of data in the target dataset to obtain the genetic probability sequence of the target dataset. It should be noted that, for example, if the number of data in the target dataset is 10, of which 2 data flow to the dth dataset and 3 data flow to the d+1th dataset, then the genetic sequence of the target dataset is , where the serial number corresponding to 2 is d, and the serial number corresponding to 3 is d+1.

[0061] If the bth node is the data source of the ath node, then the selection operator of the ath node on the bth node satisfies the expression:

[0062] ;

[0063] Where, Represents the selection operator of the a-th node to the b-th node; 、 Indicates the data size of the data corresponding to the a-th node and the b-th node; 、 Represents the data size set of the corresponding data sets of the data corresponding to the a-th node and the b-th node; represents the maximum value function; Represents the normalization function.

[0064] Where, 、 Indicates the relative data size of the data corresponding to the a-th node and the b-th node in the corresponding data set. It means taking the difference between the relative data sizes of the data corresponding to the a-th node and the b-th node in the corresponding data sets, and expressing the relative data size difference between the data corresponding to the a-th node and the b-th node. The smaller the value, the less the data corresponding to the a-th node selects from the data corresponding to the b-th node, thus indicating that the selection operator of the a-th node for the b-th node is smaller.

[0065] It should be noted that due to the existence of data variation, the selection operator cannot fully reflect the data flow. Therefore, the data size sets of the corresponding data set of the data corresponding to the a-th node and the corresponding data set of the data corresponding to the b-th node are analyzed. If the data size changes of the two data sets are relatively consistent, the data variation operators of the two data sets are smaller, and the selection operator is more accurate.

[0066] The mutation operator of the corresponding data set of the data corresponding to the ath node to the corresponding data set of the data corresponding to the bth node satisfies the expression:

[0067] ;

[0068] Where, represents the mutation operator of the corresponding data set of the data corresponding to the a-th node to the corresponding data set of the data corresponding to the b-th node; 、 Represents the data size set of the corresponding data sets of the data corresponding to the a-th node and the b-th node; represents the maximum value function; represents the edit distance function; Represents the normalization function.

[0069] Where, Represents the edit distance of the relative data size sets of the corresponding data sets of the data corresponding to the a-th node and the b-th node. The larger the value, the weaker the consistency of the data size changes of the two data sets, and thus the larger the mutation operator of the corresponding data set of the data corresponding to the a-th node to the corresponding data set of the data corresponding to the b-th node.

[0070] The data inheritance model is expressed as follows: the corresponding chromosomes of the data corresponding to the nodes at the e-th level are inherited according to the genetic probability sequence of the corresponding data set to obtain the nodes at the e+1-th level, and the corresponding chromosomes of the data corresponding to the nodes at the e+1-th level are represented by the product of the corresponding chromosomes of the data corresponding to the nodes at the e-th level and the selection operator and the mutation operator.

[0071] At this point, a data genetic model has been established.

[0072] It should be noted that if Figure 5 This is a schematic diagram of the data inheritance process. In the data inheritance model, the corresponding chromosomes of low-level data are continuously inherited through selection and mutation to form the corresponding chromosomes of high-level data. The more inherited parts of the corresponding chromosomes of the data, the wider the influence of the data and the greater the importance of the data.

[0073] Preferably, the importance of each data set is obtained based on the genetic performance of each data in the data genetic model:

[0074] Taking each node of the first level as the starting point, the data flow model is traversed using the depth-first search algorithm to obtain all possible paths; all possible paths between any two nodes are obtained.

[0075] It should be noted that all possible paths between the a-th node and the b-th node reflect the data inheritance of the a-th node and the b-th node. Among the possible paths from the a-th node to the b-th node, the larger the selection operator and the smaller the mutation operator, the stronger the inheritance performance of the b-th node on the a-th node and the greater the importance of the b-th node. When all the data in the data set to which the data corresponding to the b-th node belongs are of great importance, it means that the data set to which the data corresponding to the b-th node belongs is more important.

[0076] The node sequence of the mth possible path between the bth node and the ath node ,Will The selection operator of any two adjacent nodes is multiplied by the mutation operator of the corresponding data set of the two adjacent nodes, and the cumulative multiplication is obtained. The importance of the bth node to the ath node in is obtained by adding the importance of the bth node to the ath node in all possible paths between the bth node and the ath node to obtain the importance of the bth node to the ath node.

[0077] For the dth data set, obtain the importance of the node corresponding to the data contained in the dth data set to all other nodes, and sum them up to obtain the importance of the dth data set.

[0078] So far, the importance of each data set has been obtained.

[0079] Preferably, based on the retrieved records of each data and the straight line fitting method, the heat change model of each data set is obtained:

[0080] Obtain the number of times each data item is retrieved in each time period; take the difference between the generation time of each data item and the start time of each time period as the relative retrieval time of each data item; for all data items in the d-th data set, establish a coordinate system with the relative retrieval time as the horizontal axis and the number of times the data item is retrieved in the corresponding time period as the vertical axis, draw a scatter plot of the relative retrieval time-number of times of retrieval of all data items in the d-th data set, and use the least squares method to perform straight line fitting; use the obtained fitting line as the heat change model of the d-th data set.

[0081] At this point, the heat change model of each data set has been obtained.

[0082] It should be noted that in the dataset popularity change model, as the relative retrieval time increases, the fewer the retrieval times, the fewer backups the dataset requires. At the same time, the weaker the importance of the dataset, the fewer backups the data requires.

[0083] Preferably, the number of backups required for each dataset is obtained based on the importance of the dataset and the popularity change model of the dataset:

[0084] ;

[0085] Where, Indicates the number of backups that should be made for the dth data set; 、 Indicates the minimum and maximum values of the preset backup quantity range; Indicates the importance of the d-th data set; Represents the slope of the heat change model of the d-th data set; It should be noted that the preset backup quantity range is set by the implementation personnel according to the actual implementation situation. For example, 、 Set to 2 and 10.

[0086] At this point, the number of data sets that should be backed up has been obtained.

[0087] S4: Store the data based on the number of backups required for each data set and the data inheritance model.

[0088] It should be noted that in order to avoid bulk data loss caused by storing data with strong genetic performance together, data with strong genetic performance can be stored in different backup units based on the data genetic model, thereby ensuring the safety of the backup units.

[0089] Specifically, a backup node sequence is established. It should be noted that, taking nodes a, b, and c as an example, if the number of backups of the data corresponding to nodes a, b, and c is 2, 3, and 4, then the backup node sequence of nodes a, b, and c is .

[0090] The negative correlation normalized result of the importance of the bth node to the ath node is taken as the distance between the bth node and the ath node.

[0091] It should be noted that in order to maximize the distance between the data of the overall nodes of all backup units, nodes with smaller distances need to be backed up in different backup units.

[0092] All nodes in the backup node sequence are randomly placed in all backup units to obtain several backup schemes. The goodness of any backup scheme satisfies the expression:

[0093] ;

[0094] Where, Indicates the goodness of the t-th backup plan; Indicates the number of backup units; represents the distance between the u-th node and the v-th node in the s-th backup unit of the t-th backup scheme; Indicates the number of nodes in the sth backup unit of the tth backup scheme.

[0095] Where, It means adding the distances of all nodes in the s-th backup unit of the t-th backup scheme as the overall difference of the s-th backup unit of the t-th backup scheme. It represents the average overall difference of all backup units of the t-th backup scheme. The larger the value, the greater the distance between all nodes, which indicates that the superiority of the t-th backup scheme is higher.

[0096] Obtain the most optimal backup plan and back up the data corresponding to all nodes to complete the data storage of the trusted terminal.

[0097] At this point, the data storage of the trusted terminal is completed.

[0098] Although this specification has shown and described several embodiments of the present invention, it is obvious to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and substitutions without departing from the idea and spirit of the present invention.

Claims

1. A data storage method for a trusted terminal, characterized in that: include: Obtain several data sets from the trusted terminal, each of which is a collection of data from different time periods, and obtain metadata for all data in each data set, including the detection time, data source, data flow, and data size of the data; Establish a data circulation model based on the data sources and data flows of each data; Based on the genetic algorithm idea and data circulation model, a data genetic model is established; Obtain the importance of each data set based on the genetic performance of each data in the data genetic model, including: Take each node of the first level as the starting point, use the depth-first search algorithm to traverse the data flow model to obtain all possible paths; for the node sequence of the mth possible path between the bth node and the ath node ,Will The selection operator of any two adjacent nodes is multiplied by the mutation operator of the corresponding data set of the two adjacent nodes, and the cumulative multiplication is obtained. The importance of the bth node to the ath node is obtained by adding the importance of the bth node to the ath node in all possible paths between the bth node and the ath node to obtain the importance of the bth node to the ath node; for the dth data set, the importance of the node corresponding to the data contained in the dth data set to all other nodes is obtained, and the importance of the dth data set is obtained by summing them up; based on the retrieval records of each data set and the straight line fitting method, the heat change model of each data set is obtained; according to the importance of the data set and the heat change model of the data set, the number of backups of each data set is obtained; The number of backups for each data set must satisfy the expression: Where, Indicates the number of backups that should be made for the dth data set; 、 Indicates the minimum and maximum values of the preset backup quantity range; Indicates the importance of the d-th data set; Represents the slope of the heat change model of the d-th data set; represents the ceiling function; The data is stored based on the number of backups required for each data set and the data inheritance model; A backup node sequence is established based on the number of data sets that need to be backed up. All nodes in the backup node sequence are randomly placed in all backup units to obtain several backup schemes. The optimality of each backup scheme is obtained based on the genetic performance of the nodes in all backup units in the data genetic model. The data is stored using the backup scheme with the highest optimality.

2. A data storage method for a trusted terminal according to claim 1, characterized in that: The establishing of the data circulation model includes: Treat any data as a node and connect any two nodes based on the data source and data flow relationship between them; The distance between any two nodes is recorded as 1, and the nodes with in-degree 0 in the data flow graph are used as starting points to perform a depth-first search algorithm to obtain the shortest distance from each starting point to each node, which is recorded as the level of each node. When the level of a node is not unique, the minimum value of the node's level is recorded as the node's level. All nodes are connected according to the level to obtain a data flow model.

3. A data storage method for a credentialed terminal according to claim 2, characterized in that: The connecting of any two nodes comprises: For the i-th node and the c-th node, if the data corresponding to the c-th node is the data source of the data corresponding to the i-th node, then the i-th node and the c-th node are connected, and the direction points to the i-th node; if the data corresponding to the c-th node is the data flow direction of the data corresponding to the i-th node, then the i-th node and the c-th node are connected, and the direction points to the c-th node.

4. A data storage method for a credentialed terminal according to claim 1, characterized in that: The establishing of the data genetic model comprises: Each data is considered as a 01 string of the same size as the data size, and is recorded as the corresponding chromosome of each data; based on the data circulation of each data set reflected by the data circulation model, the genetic probability sequence of each data set is obtained; based on the data size of the data corresponding to different nodes and the data circulation model, the node-to-node selection operator and the data set-to-data set mutation operator are obtained; The data inheritance model is expressed as follows: the corresponding chromosomes of the data corresponding to the nodes at the e-th level are inherited according to the genetic probability sequence of the corresponding data set to obtain the nodes at the e+1-th level, and the corresponding chromosomes of the data corresponding to the nodes at the e+1-th level are represented by the product of the corresponding chromosomes of the data corresponding to the nodes at the e-th level and the selection operator and the mutation operator.

5. A data storage method for a credible terminal according to claim 4, characterized in that: The step of obtaining the genetic probability sequence of each data set includes: Consider any data set as the target data set, obtain the frequency distribution sequence of the data flow of all data in the target data set, record it as the genetic sequence of the target data set, divide each value of the genetic sequence of the target data set by the number of data in the target data set, and obtain the genetic probability sequence of the target data set.

6. A data storage method for a credentialed terminal according to claim 4, characterized in that: The obtaining of a node-to-node selection operator includes: If the bth node is the data source of the ath node, then the selection operator of the ath node on the bth node satisfies the expression: ； Where, Represents the selection operator of the a-th node to the b-th node; 、 Indicates the data size of the data corresponding to the a-th node and the b-th node; 、 Represents the data size set of the corresponding data sets of the data corresponding to the a-th node and the b-th node; represents the maximum value function; Represents the normalization function.

7. A data storage method for a credible terminal according to claim 4, characterized in that: The step of obtaining a mutation operator of a dataset to a dataset includes: If the bth node is the data source of the ath node, then the mutation operator of the corresponding data set of the ath node's corresponding data to the corresponding data set of the bth node's corresponding data satisfies the expression: ； Where, represents the mutation operator of the corresponding data set of the data corresponding to the a-th node to the corresponding data set of the data corresponding to the b-th node; 、 Represents the data size set of the corresponding data sets of the data corresponding to the a-th node and the b-th node; represents the maximum value function; represents the edit distance function; Represents the normalization function.

8. A data storage method for a trusted terminal according to claim 1, characterized in that: The obtaining of the quality of each backup solution includes: The normalized negative correlation of the importance of the bth node to the ath node is used as the distance between the bth node and the ath node; The goodness of any backup plan satisfies the expression: ； Where, Indicates the goodness of the t-th backup plan; Indicates the number of backup units; represents the distance between the u-th node and the v-th node in the s-th backup unit of the t-th backup scheme; Indicates the number of nodes in the sth backup unit of the tth backup scheme.

Images