A distributed storage-based encryption method and system

Abstract

The present invention discloses an encryption method and system based on distributed storage, which relates to the technical field of storage encryption. By dividing the original data set into several groups of fragments and distributing and storing them on different nodes, the degree of data dispersion is increased. Even if some nodes are attacked, it is difficult for the attacker to obtain enough fragments to reconstruct the original data. At the same time, the cracking probability Pc is lower than the maximum tolerance threshold Pr max The system can detect the situation and extract the number of shards corresponding to the minimum cracking probability, further ensuring that the data security is guaranteed to the greatest extent under the set reorganization threshold. By monitoring and recording the system performance data under different reorganization thresholds, the performance fluctuation coefficient Xbxs can be constructed to accurately evaluate the performance of the system under different configurations. If the performance fluctuation coefficient Xbxs exceeds the preset fluctuation threshold, the system will automatically issue an exclusion instruction to promptly find the reorganization threshold that does not meet the performance requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of storage encryption, and in particular to an encryption method and system based on distributed storage. Background Art

[0002] As the demand for data storage and processing continues to increase, distributed storage, as an important technical field, has received more and more attention. Specifically, distributed storage systems can further improve data security and access speed by storing data in multiple nodes.

[0003] However, when the original data set is sharded, although sharding can significantly improve data security, it also brings new challenges. If the attacker can obtain enough shards, it is still possible to reconstruct the original data, which puts higher requirements on the security of the system. In addition, under different reorganization thresholds, the system will fluctuate in data security and performance indicators, which will bring challenges to finding a balance point. If the reorganization threshold is too low, the system security may not be guaranteed; if the reorganization threshold is too high, the system performance will be affected, resulting in longer response time and increased resource consumption. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides an encryption method and system based on distributed storage, which solves the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: an encryption method based on distributed storage, comprising the following steps:

[0006] S1. The user transfers the original data set to be stored to the distributed storage system in advance, divides the original data set into several groups of fragments, and then stores the fragmented data blocks in different nodes of the distributed storage system in a dispersed manner;

[0007] S2. Use Gatling to simulate the access requests of concurrent users, monitor and record the relevant performance test data sets in the system under different reorganization thresholds, and construct the performance fluctuation coefficient Xbxs under the corresponding situation based on the relevant performance test data sets. If the performance fluctuation coefficient Xbxs exceeds the fluctuation threshold, an exclusion instruction is issued externally;

[0008] S3, after receiving the exclusion instruction, the total number of shards N is determined according to the shards divided in step S1, and according to the total number of shards N, the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards is analyzed, and the cracking probability Pc is lower than the maximum tolerance threshold Pr is selected. max The situation, and extract the minimum cracking probability Pcmin , the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial reorganization threshold number Cxs;

[0009] S4. Use convolutional neural network technology to build an encryption prediction analysis model, and calculate the minimum cracking probability Pc min It is associated with the performance fluctuation coefficient Xbxs and after linear normalization, the risk assessment index Fpzs is constructed by fitting;

[0010] S5. Pre-set the evaluation threshold G, and compare and analyze the risk assessment index Fpzs with the evaluation threshold G to comprehensively analyze the balance status of the current original data set in terms of simultaneously satisfying encryption security and system performance indicators, and take corresponding maintenance measures according to the corresponding balance status.

[0011] Preferably, S1 specifically comprises the following steps:

[0012] S11. The user uploads the original data set to be stored to the entry node of the distributed storage system. The received original data set will be divided into several groups of shards at the entry node according to the sharding mechanism of the blockchain, and the data volume of each group of shards is the same;

[0013] S12. Based on the groups of shards obtained in step S11, the divided shards are dispersedly stored on different nodes of the distributed storage system in combination with a distributed hash table.

[0014] Preferably, S2 specifically includes the following steps:

[0015] S21, based on the access requests of concurrent users, respectively monitoring and recording the relevant performance test data sets in the system under different reorganization threshold numbers;

[0016] S211. The relevant performance test data set includes the network bandwidth Wdk and the reassembly time Czsc of the system under different reassembly threshold numbers.

[0017] Preferably, S2 also includes:

[0018] S22, based on the relevant performance test data set in step S211, by associating the reorganization time Czsc with the network bandwidth Wdk under the same reorganization threshold number, and after linear normalization processing, obtaining the performance fluctuation coefficient Xbxs under the corresponding situation, the performance fluctuation coefficient Xbxs is obtained by the following formula:

[0019]

[0020] Where n is the number of reorganization thresholds, i = 1, 2, 3, ..., n, Czsc i It is expressed as the reorganization time when the number of reorganization thresholds is i. It is expressed as the mean of the reorganization time under different reorganization thresholds, Wdk i It is expressed as the network bandwidth when the number of reorganization thresholds is i, It is expressed as the mean value of the network bandwidth under different reassembly threshold numbers. α and β are weight coefficients, where the reassembly threshold number n≤the total number of fragments N.

[0021] Preferably, S2 also includes:

[0022] S23, pre-set the fluctuation threshold, and compare and analyze the fluctuation threshold with the performance fluctuation coefficient Xbxs to evaluate the performance stability of the system under different reorganization threshold numbers. The specific evaluation contents are as follows:

[0023] If the performance fluctuation coefficient Xbxs in the corresponding case exceeds the fluctuation threshold, it means that the stability of system performance is affected under the current reorganization threshold number, and an exclusion instruction will be issued externally;

[0024] If the performance fluctuation coefficient Xbxs in the corresponding case does not exceed the fluctuation threshold, it means that the stability of system performance is not affected under the current reorganization threshold number, and no additional exclusion instructions will be issued temporarily;

[0025] S231. After receiving the exclusion instruction, exclude the reorganization threshold quantity in the corresponding case, and retain the reorganization threshold quantity in the case where no exclusion instruction is issued for analysis and screening.

[0026] Preferably, S3 specifically includes the following steps:

[0027] S31. Based on the total number of shards N, analyze the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards. The specific analysis content is as follows:

[0028] S311. Based on a statistical algorithm, the number of different fractions of n slices selected from the total number of slices N is analyzed and calculated to construct a combination coefficient C(N, n), which is obtained specifically in the following manner:

[0029]

[0030] In the formula, N is the total number of shards, N! is the product of all positive integers from 1 to N, that is, the factorial of N; n! is the product of all positive integers from 1 to n, that is, the factorial of n;

[0031] S312, according to the combination coefficient C (N, n) obtained in S311, combined with the total number of slices N, to obtain the cracking probability Pc, the cracking probability Pc is obtained by the following formula:

[0032]

[0033] Where Y represents the probability of an attacker obtaining a single shard, Y n It represents the joint probability of an attacker obtaining n fragments; n is a variable, which represents the number of reorganization thresholds and is in the range of [k, N].

[0034] Preferably, S3 also includes:

[0035] S32. Preset the maximum tolerance threshold Pr max , and based on the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards, several groups of reorganization thresholds that meet security requirements are preliminarily screened out. The specific screening contents are as follows:

[0036] S321, if the cracking probability Pc in the corresponding situation is less than the maximum tolerance threshold Pr max , then the number of fragments in the corresponding case is regarded as the number of selective reorganization thresholds and included in the screening range;

[0037] S322, if the cracking probability Pc in the corresponding case ≥ the maximum tolerance threshold Pr max , the number of fragments in the corresponding case is not regarded as the selective reorganization threshold number and is not included in the screening range;

[0038] S33, according to the number of selective recombination thresholds in the screening range, extract the minimum cracking probability Pc under the corresponding recombination threshold number. min , and the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial reorganization threshold number Cxs.

[0039] Preferably, S4, by setting the minimum cracking probability Pc min The performance fluctuation coefficient Xbxs input value is encrypted in the prediction analysis model, and the minimum cracking probability Pc min After linear normalization with the performance fluctuation coefficient Xbxs, the risk assessment index Fpzs is constructed by fitting. The risk assessment index Fpzs is obtained by the following formula:

[0040]

[0041] Where W1 and W2 represent the performance fluctuation coefficient Xbxs and the minimum cracking probability Pc respectively. minThe weight coefficient of , R is represented as the correction constant.

[0042] Preferably, S5 specifically includes the following steps:

[0043] S51. By comparing and analyzing the risk assessment index Fpzs with the assessment threshold G, a comprehensive analysis is performed on the balance of the current original data set in terms of satisfying both encryption security and system performance indicators. The specific analysis content is as follows:

[0044] If the risk assessment index Fpzs ≥ the assessment threshold G, it means that the current original data set is in a balanced state in terms of satisfying both encryption security and system performance indicators;

[0045] If the risk assessment index Fpzs is less than the assessment threshold G, it means that the current original data set is not in a balanced state in terms of simultaneously satisfying the encryption security and system performance indicators, and a troubleshooting instruction will be issued at this time;

[0046] S52. After receiving the troubleshooting instruction, the original data set is re-divided, the number of shards is increased to 10% of the original, and the shard data is doubly encrypted using different keys for the two encryptions.

[0047] An encryption system based on distributed storage, comprising a partitioning module, a performance testing module, a reconstruction analysis module, a comprehensive evaluation module and a maintenance management module;

[0048] In the partitioning module, the user transfers the original data set to be stored to the distributed storage system in advance, divides the original data set into several groups of fragments, and then stores the data blocks after the several groups of fragments in different nodes of the distributed storage system;

[0049] The performance testing module uses Gatling to simulate the access requests of concurrent users, and monitors and records the relevant performance test data sets in the system under different reorganization thresholds. Based on the relevant performance test data sets, the performance fluctuation coefficient Xbxs under the corresponding situation is constructed. If the performance fluctuation coefficient Xbxs exceeds the fluctuation threshold, an exclusion instruction is issued externally;

[0050] After receiving the exclusion instruction, the reconstruction analysis module will determine the total number of shards N according to the shards divided in step S1, and analyze the cracking probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards based on the total number of shards N, and screen out the cracking probability Pc below the maximum tolerance threshold Pr max The situation, and extract the minimum cracking probability Pc min , the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial small reassembly threshold number Cxs;

[0051] The comprehensive evaluation module uses convolutional neural network technology to build an encryption prediction analysis model. min It is associated with the performance fluctuation coefficient Xbxs and after linear normalization, the risk assessment index Fpzs is constructed by fitting;

[0052] The maintenance management module pre-sets the evaluation threshold G, and compares and analyzes the risk assessment index Fpzs with the evaluation threshold G to comprehensively analyze the balance status of the current original data set in terms of simultaneously satisfying encryption security and system performance indicators, and takes corresponding maintenance measures according to the corresponding balance status.

[0053] The present invention provides an encryption method and system based on distributed storage, which has the following beneficial effects:

[0054] (1) By dividing the original data set into several groups of fragments and distributing them on different nodes, the degree of data dispersion is increased. Even if some nodes are attacked, it is difficult for the attacker to obtain enough fragments to reconstruct the original data. At the same time, the cracking probability Pc is lower than the maximum tolerance threshold Pr max The situation is analyzed, and the number of shards corresponding to the minimum cracking probability is extracted, further ensuring that the security of the data is guaranteed to the greatest extent under the set reorganization threshold number. By monitoring and recording the system performance data under different reorganization threshold numbers, the performance fluctuation coefficient Xbxs is constructed, which can accurately evaluate the performance of the system under different configurations. If the performance fluctuation coefficient Xbxs exceeds the preset fluctuation threshold, the system will automatically issue an exclusion instruction to find out the reorganization threshold that does not meet the performance requirements in time, further reducing the possibility of performance degradation in the system. The encryption prediction analysis model is constructed by using convolutional neural network technology, and the risk assessment index Fpzs is constructed by fitting. The security and performance balance of the system under the current configuration can be intelligently evaluated. By comparing and analyzing the risk assessment index Fpzs with the assessment threshold, the system's balance can be monitored and evaluated in real time. When the system is in an unbalanced state, an inspection instruction is issued and corresponding maintenance measures are taken to achieve dynamic adjustment and optimization of the system. In short, this method can not only significantly improve the security of data during data storage, but also ensure the stability and efficient operation of system performance under different reorganization threshold numbers, further resolving the contradiction between data security and system performance.

[0055] (2) By considering the mean and weight coefficients of the reassembly time Czsc and network bandwidth Wdk under different reassembly thresholds, the fluctuation of system performance is quantified. The performance fluctuation coefficient Xbxs is compared and analyzed with the threshold to evaluate the performance stability of the system under different reassembly thresholds. This can dynamically adjust and optimize system performance and ensure stable operation of the system under different reassembly thresholds.

[0056] (3) By calculating the cracking probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards based on statistics and probability theory, this process enables the system to select the appropriate number of reconstruction thresholds according to actual conditions, so as to optimize system performance and resource utilization while ensuring data security. The present invention ensures that the original data set is still difficult to reconstruct when the attacker obtains some shards by accurately calculating and evaluating the cracking probability, thereby greatly improving the security of the system. This method effectively reduces the risk of data leakage and improves the overall defense capability of the system. Through detailed analysis of the combination coefficient and the cracking probability, the system can dynamically adjust and optimize the number of reconstruction thresholds to ensure that the best security configuration can be achieved under different circumstances. This dynamic adjustment capability makes the system more flexible and efficient, and can better cope with various potential security threats.

[0057] (4) By judging whether the cracking probability Pc is less than the maximum tolerance threshold Pr max The system can accurately determine which number of fragments can be used as the selective reorganization threshold number. This precise screening method can help the system avoid some unnecessary security risks while improving the performance and efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 This is a schematic diagram of the steps of an encryption method based on distributed storage of the present invention;

[0059] Figure 2 This is a schematic diagram of an encryption system framework based on distributed storage in the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] Example 1

[0062] See also Figure 1The present invention provides an encryption method based on distributed storage, comprising the following steps:

[0063] S1. The user transfers the original data set to be stored to the distributed storage system in advance, divides the original data set into several groups of fragments, and then stores the fragmented data blocks in different nodes of the distributed storage system in a dispersed manner;

[0064] S2. Use Gatling to simulate the access requests of concurrent users, monitor and record the relevant performance test data sets in the system under different reorganization thresholds, and construct the performance fluctuation coefficient Xbxs under the corresponding situation based on the relevant performance test data sets. If the performance fluctuation coefficient Xbxs exceeds the fluctuation threshold, an exclusion instruction is issued externally;

[0065] S3, after receiving the exclusion instruction, the total number of shards N is determined according to the shards divided in step S1, and according to the total number of shards N, the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards is analyzed, and the cracking probability Pc is lower than the maximum tolerance threshold Pr is selected. max The situation, and extract the minimum cracking probability Pc min , the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial reorganization threshold number Cxs;

[0066] S4. Use convolutional neural network technology to build an encryption prediction analysis model, and calculate the minimum cracking probability Pc min It is associated with the performance fluctuation coefficient Xbxs and after linear normalization, the risk assessment index Fpzs is constructed by fitting;

[0067] S5. Pre-set the evaluation threshold G, and compare and analyze the risk assessment index Fpzs with the evaluation threshold G to comprehensively analyze the balance status of the current original data set in terms of simultaneously satisfying encryption security and system performance indicators, and take corresponding maintenance measures according to the corresponding balance status.

[0068] It should be noted that Gatling is an open source performance testing tool based on Scala, which is mainly used for load testing and performance testing of Web applications. It can simulate a large number of concurrent user requests, monitor and record the performance of the system under high load, so as to find system bottlenecks and performance problems.

[0069] In this embodiment, by dividing the original data set into several groups of fragments and distributing them to different nodes, the data security is further effectively improved, the data leakage risk caused by single point failure is reduced, Gatling is used to simulate concurrent user access requests, and the system performance under different reorganization thresholds is monitored to prevent the system from being unable to effectively protect and transmit data under high load. After receiving the exclusion instruction, the total number of fragments N is analyzed to determine the cracking probability Pc of the attacker reconstructing the original data after obtaining different numbers of fragments, and the situation with the lowest cracking probability is screened out to further optimize data security. The performance fluctuation coefficient Xbxs is constructed through the performance test data set, and an exclusion instruction is issued when the performance fluctuation coefficient Xbxs exceeds the threshold, further ensuring the stability of the system performance under the appropriate number of reorganization thresholds. The encryption prediction analysis model is constructed using convolutional neural network technology, and the risk assessment index Fpzs is constructed by fitting, and the security and performance indicators of the system are comprehensively evaluated. Based on the comparison results, the encryption strategy of the system is dynamically tuned to ensure that the system can automatically adjust to reach a relatively appropriate balance point under different loads and security requirements; according to the comprehensive analysis results, corresponding maintenance measures are taken to timely eliminate the imbalance between system performance and security, so as to further improve the system's adaptive ability and stability. In summary, this method ensures that the system can operate stably in various complex environments through risk assessment and dynamic tuning mechanisms, enhances the system's reliability and fault tolerance, and the automatic error correction and adjustment of system performance and security reduce the need for manual intervention, improving the system's autonomous operation capabilities and management efficiency.

[0070] Example 2

[0071] Please refer to Figure 1 Specifically: S1 specifically includes the following steps:

[0072] S11. The user uploads the original data set to be stored to the entry node of the distributed storage system. This node is responsible for receiving, verifying and preprocessing the data. Data verification and integrity check can be achieved through verification, such as MD5 or SHA-256. The received original data set will be divided into several groups of shards at the entry node according to the sharding mechanism of the blockchain, and the data volume of each group of shards is the same;

[0073] S12. Based on the several groups of shards obtained in step S11, the divided shards are dispersedly stored on different nodes of the distributed storage system in combination with a distributed hash table to ensure that each shard or data block is stored on multiple nodes to improve fault tolerance and data availability.

[0074] In this embodiment, in step S11, the user uploads the original data set to the entry node of the distributed storage system, and ensures that the uploaded data is not tampered with or damaged during the transmission process, further enhancing the security and integrity of the data. The original data set is divided into several groups of fragments using the sharding mechanism of the blockchain, and the data volume of each group of fragments is the same, ensuring that the data will not be lost or misplaced during the sharding process, so as to improve the stability and security of the data sharding. In step S12, the fragmented data is dispersed and stored on different nodes through a distributed hash table, ensuring that each fragment or data block is stored on multiple nodes, which not only improves the fault tolerance of the data, but also enhances the availability of the data. Since each fragment or data block is stored on multiple nodes, even if some nodes fail, the data can still be recovered from other nodes, ensuring the high availability and reliability of the system. By using a distributed hash table to efficiently manage and find the location of the data fragments, the delay of data access is further reduced, and the response speed and overall performance of the system are improved. Multiple nodes in the distributed storage system share the storage and processing tasks, avoiding single-point bottlenecks, and further improving the scalability and processing capabilities of the system. Data shards are stored on multiple nodes. Once a node fails, the corresponding data shards can be restored from other nodes to ensure data integrity and recoverability, improving the system's fault tolerance. This multi-node storage mechanism can also effectively resist data loss and malicious attacks, further improving the security and reliability of the system.

[0075] Example 3

[0076] Please refer to Figure 1 Specifically: S2 specifically includes the following steps:

[0077] S21, based on the access requests of concurrent users, respectively monitoring and recording the relevant performance test data sets in the system under different reorganization threshold numbers;

[0078] S211. The relevant performance test data set includes the network bandwidth Wdk and the reassembly time Czsc of the system under different reassembly threshold numbers.

[0079] S2 also includes:

[0080] S22, based on the relevant performance test data set in step S211, by associating the reorganization time Czsc with the network bandwidth Wdk under the same reorganization threshold number, and after linear normalization processing, obtaining the performance fluctuation coefficient Xbxs under the corresponding situation, the performance fluctuation coefficient Xbxs is obtained by the following formula:

[0081]

[0082] Where n is the number of reorganization thresholds, i = 1, 2, 3, ..., n, Czsc i It is expressed as the reorganization time when the number of reorganization thresholds is i. It is expressed as the mean of the reorganization time under different reorganization thresholds, Wdk i It is expressed as the network bandwidth when the number of reorganization thresholds is i, It is expressed as the mean value of the network bandwidth under different reassembly threshold numbers, α and β are weight coefficients, where the reassembly threshold number n≤the total number of fragments N, 0<α≤1, 0<β≤1, and α+β=1.

[0083] The above-mentioned reorganization time Czsc refers to the time required for reorganizing the fragmented data into complete data. It can be monitored and obtained using the system's built-in logging and performance monitoring tools, such as dmesg and syslog in Linux systems, and event viewer in Windows systems.

[0084] The network bandwidth Wdk is used to represent the network resource consumption during the transmission and reassembly of fragmented data. The network bandwidth usage can be obtained by using the built-in network monitoring functions of routers and switches, such as SNMP or NetFlow protocol. Of course, network traffic monitors, such as network probes or network traffic analyzers, can also be used to monitor and analyze network traffic and bandwidth usage.

[0085] S2 also includes:

[0086] S23, pre-set the fluctuation threshold, and compare and analyze the fluctuation threshold with the performance fluctuation coefficient Xbxs to evaluate the performance stability of the system under different reorganization threshold numbers. The specific evaluation contents are as follows:

[0087] If the performance fluctuation coefficient Xbxs in the corresponding case exceeds the fluctuation threshold, it means that the stability of system performance is affected under the current reorganization threshold number, and an exclusion instruction will be issued externally;

[0088] If the performance fluctuation coefficient Xbxs in the corresponding case does not exceed the fluctuation threshold, it means that the stability of system performance is not affected under the current reorganization threshold number, and no additional exclusion instructions will be issued temporarily;

[0089] S231. After receiving the exclusion instruction, exclude the reorganization threshold quantity in the corresponding case, and retain the reorganization threshold quantity in the case where no exclusion instruction is issued for analysis and screening.

[0090] In this embodiment, by monitoring and recording the system performance data set under different reorganization threshold numbers, the performance fluctuation coefficient Xbxs under the corresponding situation is obtained. By introducing the mean of the reorganization time and the network bandwidth and the calculation formula of the weight coefficient, the accurate analysis of the system performance fluctuation is further realized, which is helpful to find the changing trend of the system performance under different reorganization threshold numbers; in step S23, by presetting the fluctuation threshold and comparing it with the performance fluctuation coefficient Xbxs, the performance stability of the system under different reorganization threshold numbers can be evaluated. Such a dynamic evaluation method can timely find the instability of the system performance and issue an exclusion instruction to ensure the stability and reliability of the system. In step S231, after receiving the exclusion instruction, the reorganization threshold number under the corresponding situation is excluded, and the reorganization threshold number under the situation where the exclusion instruction is not issued is retained for analysis and screening. Such a screening mechanism can effectively exclude the reorganization threshold number that makes the system unstable, retain the reorganization threshold for stable operation of the system performance, and optimize the system performance. By accurately monitoring and evaluating the system performance fluctuation under different reorganization threshold numbers, the stability and security of the system in actual operation are ensured. Especially under high concurrent access requests, the reorganization threshold with unstable performance can be discovered and eliminated in time, ensuring that the system can maintain good performance and security under different load conditions. In short, this method not only improves the monitoring and evaluation accuracy of system performance through detailed steps and technical means, but also ensures the stability and security of the system through dynamic performance stability evaluation and optimization screening mechanism, providing an efficient and reliable solution for distributed storage systems.

[0091] Example 4

[0092] Please refer to Figure 1 Specifically: S3 specifically includes the following steps:

[0093] S31. Based on the total number of shards N, analyze the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards. The specific analysis content is as follows:

[0094] S311. Based on a statistical algorithm, the number of different fractions of n slices selected from the total number of slices N is analyzed and calculated to construct a combination coefficient C(N, n), which is obtained specifically in the following manner:

[0095]

[0096] In the formula, N is the total number of shards, N! is the product of all positive integers from 1 to N, that is, the factorial of N; n! is the product of all positive integers from 1 to n, that is, the factorial of n;

[0097] S312, according to the combination coefficient C (N, n) obtained in S311, combined with the total number of slices N, to obtain the cracking probability Pc, the cracking probability Pc is obtained by the following formula:

[0098]

[0099] Where Y represents the probability of an attacker obtaining a single shard, Y n It represents the joint probability of an attacker obtaining n fragments; n is a variable, which represents the number of reorganization thresholds and is in the range of [k, N].

[0100] In this embodiment, in step S311, the number of different fractions of n shards selected from the total number of shards N is analyzed and calculated by a statistical algorithm to construct a combination coefficient C (N, n). This precise calculation of the number of combinations can comprehensively consider all possible shard combinations, thereby providing a solid foundation for the calculation of the cracking probability Pc. In step S312, the cracking probability Pc is obtained based on the combination coefficient C (N, n) and the total number of shards N, combined with the probability Y of the attacker obtaining a single shard. This probability algorithm in statistics can accurately evaluate the possibility of the attacker successfully reconstructing the original data set under different numbers of shards, providing a scientific basis for data security analysis. And by analyzing and calculating the cracking probability Pc in detail, this method can identify data security risks under different numbers of reorganization thresholds.

[0101] Specifically, by comparing the cracking probability Pc under different numbers of shards, the number of security thresholds is further determined to ensure that system performance is optimized while ensuring data security. This can prevent attackers from reconstructing the original data by obtaining some shards and indirectly improve the operating efficiency of the system. This method provides a scientific method based on statistics and probability theory to comprehensively evaluate and optimize data security and system performance. In short, this method ensures the high security and high performance of the distributed storage system through accurate combination number calculation and cracking probability evaluation.

[0102] Example 5

[0103] Please refer to Figure 1 , specifically: S3 also includes:

[0104] S32. Preset the maximum tolerance threshold Pr max , and based on the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards, several groups of reorganization thresholds that meet security requirements are preliminarily screened out. The specific screening contents are as follows:

[0105] S321, if the cracking probability Pc in the corresponding situation is less than the maximum tolerance threshold Pr max, then the number of fragments in the corresponding case is regarded as the number of selective reorganization thresholds and included in the screening range;

[0106] S322, if the cracking probability Pc in the corresponding case ≥ the maximum tolerance threshold Pr max , the number of fragments in the corresponding case is not regarded as the selective reorganization threshold number and is not included in the screening range;

[0107] S33, according to the number of selective recombination thresholds in the screening range, extract the minimum cracking probability Pc under the corresponding recombination threshold number. min , and the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial reorganization threshold number Cxs.

[0108] In this embodiment, the number of reconstruction thresholds is preliminarily screened based on the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of fragments. If the probability Pc is lower than the maximum tolerance threshold Pr max , the corresponding number of fragments is regarded as the number of selective reorganization thresholds. This strategy helps ensure that the system has sufficient security when attackers try to reconstruct data. max The comparison can accurately screen out the number of reorganization thresholds that meet security requirements. This accurate evaluation based on actual data and preset thresholds ensures that the system can make decisions based on scientific basis when selecting the number of reorganization thresholds, further avoiding excessive or insufficient security protection measures. By extracting the minimum cracking probability Pc in step S33, min The corresponding number of preliminary reorganization thresholds Cxs, the system can optimize system performance and resource utilization under the premise of ensuring data security. This strategy of balancing performance and security enables the system to flexibly adjust under different circumstances, improving the overall efficiency and responsiveness of the system. Through the screening and analysis steps, the method can dynamically adjust and optimize the selection of the number of reorganization thresholds. This flexibility enables the system to make timely adjustments based on real-time security threats and performance requirements, thereby effectively responding to changing security challenges. Finally, the number of preliminary reorganization thresholds Cxs extracted in step S33 can not only ensure the security and performance of the system, but also reduce unnecessary resource consumption and system maintenance costs. This efficient system management method provides reliable guarantee for the long-term stable operation of the distributed storage system.

[0109] Example 6

[0110] Please refer to Figure 1 , specifically: S4, by setting the minimum cracking probability Pc minThe performance fluctuation coefficient Xbxs input value is encrypted in the prediction analysis model, and the minimum cracking probability Pc min After linear normalization with the performance fluctuation coefficient Xbxs, the risk assessment index Fpzs is constructed by fitting. The risk assessment index Fpzs is obtained by the following formula:

[0111]

[0112] Where W1 and W2 represent the performance fluctuation coefficient Xbxs and the minimum cracking probability Pc respectively. min The weight coefficient, R is represented as a correction constant, where 0<W1≤1, 0<W2≤1, and W1+W2=1.

[0113] S5 specifically includes the following steps:

[0114] S51. By comparing and analyzing the risk assessment index Fpzs with the assessment threshold G, a comprehensive analysis is performed on the balance of the current original data set in terms of satisfying both encryption security and system performance indicators. The specific analysis content is as follows:

[0115] If the risk assessment index Fpzs ≥ the assessment threshold G, it means that the current original data set is in a balanced state in terms of satisfying both encryption security and system performance indicators;

[0116] If the risk assessment index Fpzs is less than the assessment threshold G, it means that the current original data set is not in a balanced state in terms of simultaneously satisfying the encryption security and system performance indicators, and a troubleshooting instruction will be issued at this time;

[0117] S52. After receiving the troubleshooting instruction, the original data set is re-divided and the number of shards is increased to 10% of the original to ensure data security and optimize system performance. The shard data is doubly encrypted using different keys.

[0118] In this embodiment, by constructing the risk assessment index Fpzs, the system's balance between security and performance can be further comprehensively evaluated to provide a scientific basis for subsequent decision-making; by comparing the risk assessment index Fpzs with the preset assessment threshold G, the current original data set is comprehensively analyzed in terms of encryption security and system performance indicators. If the risk assessment index Fpzs is greater than or equal to the assessment threshold G, the system considers that it is currently in a state of balance between security and performance; otherwise, a troubleshooting instruction will be triggered to ensure that the system operates in the best state. In short, the dynamic adjustment strategy and dual encryption technology adopted by this method enable the system to respond to and adapt to changing security threats and performance requirements in real time. This real-time and flexibility enable the system to maximize performance and efficiency while ensuring safety. Through comprehensive analysis and dynamic adjustment.

[0119] Example 7

[0120] Please refer to Figure 1 and Figure 2 ,Specifically: An encryption system based on distributed storage, including a partitioning module, a performance testing module, a reconstruction analysis module, a comprehensive evaluation module and a maintenance management module;

[0121] In the partitioning module, the user transfers the original data set to be stored to the distributed storage system in advance, divides the original data set into several groups of fragments, and then stores the data blocks after the several groups of fragments in different nodes of the distributed storage system;

[0122] The performance testing module uses Gatling to simulate the access requests of concurrent users, and monitors and records the relevant performance test data sets in the system under different reorganization thresholds. Based on the relevant performance test data sets, the performance fluctuation coefficient Xbxs under the corresponding situation is constructed. If the performance fluctuation coefficient Xbxs exceeds the fluctuation threshold, an exclusion instruction is issued externally;

[0123] After receiving the exclusion instruction, the reconstruction analysis module will determine the total number of shards N according to the shards divided in step S1, and analyze the cracking probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards based on the total number of shards N, and screen out the cracking probability Pc below the maximum tolerance threshold Pr max The situation, and extract the minimum cracking probability Pc min , the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial small reassembly threshold number Cxs;

[0124] The comprehensive evaluation module uses convolutional neural network technology to build an encryption prediction analysis model. min It is associated with the performance fluctuation coefficient Xbxs and after linear normalization, the risk assessment index Fpzs is constructed by fitting;

[0125] The maintenance management module pre-sets the evaluation threshold G, and compares and analyzes the risk assessment index Fpzs with the evaluation threshold G to comprehensively analyze the balance status of the current original data set in terms of simultaneously satisfying encryption security and system performance indicators, and takes corresponding maintenance measures according to the corresponding balance status.

[0126] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An encryption method based on distributed storage, characterized in that: The following steps are involved: S1. The user transfers the original data set to be stored to the distributed storage system in advance, divides the original data set into several groups of fragments, and then stores the fragmented data blocks in different nodes of the distributed storage system in a dispersed manner; S2. Use Gatling to simulate the access requests of concurrent users, monitor and record the relevant performance test data sets in the system under different reorganization thresholds, and construct the performance fluctuation coefficient Xbxs under the corresponding situation based on the relevant performance test data sets. If the performance fluctuation coefficient Xbxs exceeds the fluctuation threshold, an exclusion instruction is issued externally; S2 specifically includes the following steps: S21, based on the access requests of concurrent users, respectively monitoring and recording the relevant performance test data sets in the system under different reorganization threshold numbers; S211, the relevant performance test data set includes the network bandwidth Wdk and the reassembly time Czsc of the system under different reassembly threshold numbers; S2 also includes: S22, based on the relevant performance test data set in step S211, by associating the reorganization time Czsc with the network bandwidth Wdk under the same reorganization threshold number, and after linear normalization processing, obtaining the performance fluctuation coefficient Xbxs under the corresponding situation, the performance fluctuation coefficient Xbxs is obtained by the following formula: ; Where n is the number of reorganization thresholds, i=1, 2, 3, ..., n, It is expressed as the reorganization time when the number of reorganization thresholds is i. It is expressed as the mean of the reorganization time under different reorganization threshold numbers, It is expressed as the network bandwidth when the number of reorganization thresholds is i, It is expressed as the average value of network bandwidth under different reassembly threshold numbers, and They are all weight coefficients, where the reassembly threshold number n≤the total number of fragments N; S3, after receiving the exclusion instruction, the total number of shards N is determined according to the shards divided in step S1, and according to the total number of shards N, the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards is analyzed, and the cracking probability Pc is lower than the maximum tolerance threshold Pr is selected. max The situation, and extract the minimum cracking probability Pc min , the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial reorganization threshold number Cxs; S3 specifically includes the following steps: S31. Based on the total number of shards N, analyze the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards. The specific analysis content is as follows: S311. Based on a statistical algorithm, the number of different fractions of n shards selected from the total number of shards N is analyzed and calculated to construct a combination coefficient C (N, n), which is obtained specifically in the following manner: ; In the formula, N is the total number of shards, N! is the product of all positive integers from 1 to N, that is, the factorial of N; n! is the product of all positive integers from 1 to n, that is, the factorial of n; S312, according to the combination coefficient C (N, n) obtained in S311, combined with the total number of slices N, to obtain the cracking probability Pc, the cracking probability Pc is obtained by the following formula: ; Where Y represents the probability of an attacker obtaining a single shard. It represents the joint probability of the attacker obtaining n fragments; n is a variable, which represents the number of reorganization thresholds, and its range is Inside; S4. Use convolutional neural network technology to build an encryption prediction analysis model, and calculate the minimum cracking probability Pc min It is associated with the performance fluctuation coefficient Xbxs and after linear normalization, the risk assessment index Fpzs is constructed by fitting; S5. Pre-set the evaluation threshold G, and compare and analyze the risk assessment index Fpzs with the evaluation threshold G to comprehensively analyze the balance status of the current original data set in terms of simultaneously satisfying encryption security and system performance indicators, and take corresponding maintenance measures according to the corresponding balance status.

2. The encryption method based on distributed storage according to claim 1, characterized in that: S1 specifically includes the following steps: S11. The user uploads the original data set to be stored to the entry node of the distributed storage system. The received original data set will be divided into several groups of shards at the entry node according to the sharding mechanism of the blockchain, and the data volume of each group of shards is the same; S12. Based on the groups of shards obtained in step S11, the divided shards are dispersedly stored on different nodes of the distributed storage system in combination with a distributed hash table.

3. The encryption method based on distributed storage according to claim 1, characterized in that: S2 also includes: S23, pre-set the fluctuation threshold, and compare and analyze the fluctuation threshold with the performance fluctuation coefficient Xbxs to evaluate the performance stability of the system under different reorganization threshold numbers. The specific evaluation contents are as follows: If the performance fluctuation coefficient Xbxs in the corresponding case exceeds the fluctuation threshold, it means that the stability of system performance is affected under the current reorganization threshold number, and an exclusion instruction will be issued externally; If the performance fluctuation coefficient Xbxs in the corresponding case does not exceed the fluctuation threshold, it means that the stability of system performance is not affected under the current reorganization threshold number, and no additional exclusion instructions will be issued temporarily; S231. After receiving the exclusion instruction, exclude the reorganization threshold quantity in the corresponding case, and retain the reorganization threshold quantity in the case where no exclusion instruction is issued for analysis and screening.

4. The encryption method based on distributed storage according to claim 1, characterized in that: S3 also includes: S32. Preset the maximum tolerance threshold Pr max , and based on the probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards, several groups of reorganization thresholds that meet security requirements are preliminarily screened out. The specific screening contents are as follows: S321, if the cracking probability Pc in the corresponding situation is less than the maximum tolerance threshold Pr max , then the number of fragments in the corresponding case is regarded as the number of selective reorganization thresholds and included in the screening range; S322, if the cracking probability Pc in the corresponding case ≥ the maximum tolerance threshold Pr max , the number of fragments in the corresponding case is not regarded as the selective reorganization threshold number and is not included in the screening range; S33, according to the number of selective recombination thresholds in the screening range, extract the minimum cracking probability Pc under the corresponding recombination threshold number. min , and the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial reorganization threshold number Cxs.

5. The encryption method based on distributed storage according to claim 1, characterized in that: S4, by the minimum crack probability Pc min The performance fluctuation coefficient Xbxs input value is encrypted in the prediction analysis model, and the minimum cracking probability Pc min After linear normalization with the performance fluctuation coefficient Xbxs, the risk assessment index Fpzs is constructed by fitting. The risk assessment index Fpzs is obtained by the following formula: ; Where W1 and W2 represent the performance fluctuation coefficient Xbxs and the minimum cracking probability Pc respectively. min The weight coefficient of , R is represented as the correction constant.

6. The encryption method based on distributed storage according to claim 1, characterized in that: S5 specifically includes the following steps: S51. By comparing and analyzing the risk assessment index Fpzs with the assessment threshold G, a comprehensive analysis is performed on the balance of the current original data set in terms of satisfying both encryption security and system performance indicators. The specific analysis content is as follows: If the risk assessment index Fpzs ≥ the assessment threshold G, it means that the current original data set is in a balanced state in terms of satisfying both encryption security and system performance indicators; If the risk assessment index Fpzs is less than the assessment threshold G, it means that the current original data set is not in a balanced state in terms of simultaneously satisfying the encryption security and system performance indicators, and a troubleshooting instruction will be issued at this time; S52. After receiving the troubleshooting instruction, the original data set is re-divided, the number of shards is increased to 10% of the original, and the shard data is doubly encrypted using different keys for the two encryptions.

7. An encryption system based on distributed storage, applying the encryption method based on distributed storage as described in any one of claims 1 to 6 above, characterized in that: It includes partition module, performance test module, reconstruction analysis module, comprehensive evaluation module and maintenance management module; In the partitioning module, the user transfers the original data set to be stored to the distributed storage system in advance, divides the original data set into several groups of fragments, and then stores the data blocks after the several groups of fragments in different nodes of the distributed storage system; The performance testing module uses Gatling to simulate the access requests of concurrent users, and monitors and records the relevant performance test data sets in the system under different reorganization thresholds. Based on the relevant performance test data sets, the performance fluctuation coefficient Xbxs under the corresponding situation is constructed. If the performance fluctuation coefficient Xbxs exceeds the fluctuation threshold, an exclusion instruction is issued externally; After receiving the exclusion instruction, the reconstruction analysis module will determine the total number of shards N according to the shards divided in step S1, and analyze the cracking probability Pc of the attacker successfully reconstructing the original data set after obtaining different numbers of shards based on the total number of shards N, and screen out the cracking probability Pc below the maximum tolerance threshold Pr max The situation, and extract the minimum cracking probability Pc min , the minimum cracking probability Pc min The corresponding number of fragments obtained is recorded as the initial small reassembly threshold number Cxs; The comprehensive evaluation module uses convolutional neural network technology to build an encryption prediction analysis model. min It is associated with the performance fluctuation coefficient Xbxs and after linear normalization, the risk assessment index Fpzs is constructed by fitting; The maintenance management module pre-sets the evaluation threshold G, and compares and analyzes the risk assessment index Fpzs with the evaluation threshold G to comprehensively analyze the balance status of the current original data set in terms of simultaneously satisfying encryption security and system performance indicators, and takes corresponding maintenance measures according to the corresponding balance status.

Images